US inflation slowed more sharply than expected in June, although the reading is far from a definitive victory for the Federal Reserve. The CPI fell 0.4% month over month—its first decline since April 2020—and eased to 3.5% year over year, down from 4.2% in May. Core inflation also provided some relief, rising 2.6% year over year while remaining flat on a monthly basis.

At first glance, the data looks clearly dovish. The problem is that much of the improvement came from energy, where prices fell 5.7% during the month, with gasoline plunging by almost 10%. This decline reflected the temporary truce between the United States and Iran, but the backdrop has already changed. The resumption of attacks in the Strait of Hormuz, the new naval blockade, and the rebound in oil prices make it likely that much of this relief will disappear in July.

In other words, the June report is backward-looking. Inflation improved because energy prices fell, but the main catalyst behind that decline has already reversed. If crude oil and gasoline continue to rise, the impact will eventually feed through not only to the energy component, but also to core inflation through transportation, logistics, and production costs. For now, the market expects the Fed to keep interest rates unchanged in the 3.50%–3.75% range at its next meeting, although it is still pricing in close to a 60% probability of a rate hike in September. Kevin Warsh reinforced this interpretation by insisting that the central bank will not tolerate persistently elevated inflation.

There were also positive signals in services and housing. Owners’ equivalent rent increased by just 0.2%, shelter recorded its smallest monthly rise since 2021, and healthcare prices declined slightly. Core goods prices fell for a second consecutive month, which could suggest that the pass-through from tariffs to consumer prices is losing momentum. However, there are still reasons to remain cautious. Part of the weakness came from volatile or potentially unsustainable components, such as motor vehicle insurance, tobacco, and certain medical services. In addition, estimates suggest that core PCE inflation may still have risen by around 3.3% year over year in June, well above the Fed’s 2% target.

Overall, the report represents tactical relief for the Fed, not evidence that the inflation problem has been resolved. Core inflation is cooling, but the energy shock is once again moving to the forefront and threatens to reignite price pressures in the coming months. While the market debates whether June marks the beginning of a broader trend, geopolitics is once again reminding investors that energy-driven disinflation can disappear just as quickly as it arrived.

South Korea is beginning to display some of the classic signs of a market dominated by retail leverage. Margin loans against KOSPI and KOSDAQ stocks have surpassed 35 trillion won, well above the peaks reached during the previous speculative cycle in 2021. The increase has been particularly steep throughout 2026, suggesting that a growing share of the rally was supported not only by fundamentals or long-term capital flows, but also by debt.

The consequences have emerged during the recent correction. Sharp declines in leveraged products have caused massive losses and forced regulators to call an emergency meeting. What is most concerning is the profile of those affected: investors aged between 20 and 39 account for 62% of the accounts that were effectively wiped out. In other words, the group with the least experience and probably the lowest level of personal wealth assumed the largest share of the risk. In fact, it is estimated that one in every 30 South Koreans has received a margin call this month.

This is not a problem unique to South Korea, but another example of how the democratisation of market access has also democratised leverage. When markets rise, margin amplifies returns and reinforces the perception that risk has disappeared. When the market turns, the opposite occurs, with forced selling, liquidations, and further downward pressure on prices. The chart does not necessarily signal an immediate market top, but it does reveal a much more fragile market structure. The greater the proportion of borrowed money, the less it takes to turn a normal correction into a violent deleveraging event. And, as is often the case, those who arrive last at the party are also the ones who end up paying most of the bill.

In fact, much of the problem stems from a failure to understand the capital cycle, into which semiconductor companies fit perfectly. In recent weeks, a table like the one attached has become increasingly popular, showing how optically “cheap” the leading memory-chip companies appear, despite some of them having risen close to 1,000% this year. What this interpretation fails to capture are the real implications of that apparent bargain. At the peak of the cycle, when memory is in short supply and prices are artificially elevated, earnings surge. However, new capacity is already being built and is expected to bring this period of extraordinary margins to an end. Of course, it remains to be seen whether the next cyclical trough will be higher than the previous one and whether profitability will stabilise at more constructive levels. But buying cyclical companies when they appear cheapest is a strategy that has rarely worked.

A stock’s performance depends far less on the specific company than we tend to think. Historically, more than half of returns have been explained by the sector, industry, or subsector to which a company belongs, compared with roughly 26% attributable to company-specific factors. In cyclical businesses, therefore, getting the sector positioning right is often even more important than finding the perfect operator.

This is where gold miners start to look particularly attractive. Based on 2026 estimates, the sector offers a free cash flow yield of around 7.7% and trades at approximately 7.3x EV/EBITDA, below most sectors in the market. What makes this especially interesting is that this depressed valuation coexists with meaningful EBITDA growth, something that is difficult to reconcile with the sector’s current discount. The market continues to treat miners solely as low-quality, capital-intensive businesses—which they are—with a long history of poor capital allocation. To some extent, that mistrust is justified. For years, many companies destroyed value through acquisitions made at the top of the cycle, cost overruns, weak balance sheets, and constant shareholder dilution.

However, the sector is entering this phase with healthier financial structures, greater capital discipline, and, in many cases, a clear focus on returning cash to shareholders.

The optionality is also enormous. If gold continues to rise while costs remain relatively contained, a large part of every additional increase in the gold price flows directly into earnings and free cash flow. At $5,000 per ounce, the sector’s FCF yield would approach 8%; at $6,000 per ounce, it would exceed 11%; and at $7,000 per ounce, it would be close to 14%. The thesis does not require any of these scenarios to materialise, but they help illustrate the significant operating leverage embedded in the business. Of course, not all mining companies are created equal. Jurisdiction, asset life, ore grade, costs, balance-sheet strength, and capital discipline remain decisive factors.

In fact, this is not an issue specific to gold miners, but one that applies to the entire commodities complex. Despite having rebounded strongly in relative terms from the 2020 lows, commodities remain at historically depressed levels. That becomes particularly intriguing and attractive in an era of US dollar debasement and record infrastructure investment.

The war between the United States and Iran has once again placed oil at the center of the geopolitical chessboard, yet the market’s reaction has been surprisingly restrained. Under almost any other circumstances, a combination of attacks on oil tankers in the Strait of Hormuz, production shutdowns across the Middle East, direct clashes between Washington and Tehran, and the disruption of one of the world’s most important energy arteries would have triggered a vertical surge in crude prices. Instead, Brent is hovering at around $85 per barrel, while WTI remains near the $80 range.

The physical situation is far more fragile than prices suggest. Global oil inventories are approximately 900 million barrels below their late-February levels, refined product stocks are so depleted that refining margins are trading at record highs, and U.S. commercial inventories are approaching minimum operational levels. China has also lifted restrictions on product exports and is beginning to absorb a significant portion of the available floating storage. In other words, there is no substantial buffer available to withstand a prolonged disruption.

At the same time, shut-in production across the Middle East is once again approaching 7.5 million barrels per day and could rise towards 11 million if Fujairah is restricted and Oman’s maritime corridor becomes unusable. This risk is now being compounded by the Red Sea front. Iran has reportedly asked the Houthis to prepare to disrupt traffic through Bab el-Mandeb if the United States attacks its electricity infrastructure. The group is said to have already deployed drones and missiles near the strait, a route through which approximately 4 million barrels per day of crude oil and another 1 million barrels of refined products pass. The problem is not only the possibility of Bab el-Mandeb being closed, but that this threat is emerging while Hormuz is already blocked. The region therefore risks losing its two main energy export routes simultaneously. Saudi Arabia has redirected roughly 70% of its exports towards Yanbu, on the Red Sea, specifically to bypass the crisis in the Gulf. If the Houthis attack that infrastructure or resume targeting maritime traffic, the principal alternative route would also be compromised.

Iran’s strategy appears clear. Tehran is seeking to raise the economic cost of any further U.S. escalation and turn global energy supply into its primary source of leverage. It does not need to sink every vessel or physically close the straits for months. It only needs to increase the perceived risk, drive up insurance premiums, and force shipping companies and traders to avoid the region. In a market with inventories at minimal levels, every barrel that takes longer to arrive matters.

What is truly unusual, therefore, is not that the market is under pressure, but that prices have yet to reflect it. The most likely explanation is that investors continue to treat the conflict as a temporary disruption, relying on a rapid de-escalation or on the ability of other producers to replace the lost barrels. But that confidence rests on an increasingly fragile foundation. Effective spare capacity is limited, inventories have been drained, and alternative routes are also under threat. Oil does not need 11 million barrels per day to disappear for an extended period in order to rise. It only needs the market to stop believing that those barrels will return tomorrow. The current complacency can persist for as long as traders continue to view the conflict as a political episode. The moment they begin to see it as a physical supply problem, the adjustment could be extremely rapid. As always, the barrel will ultimately impose its own arithmetic.

Samsung reported seemingly spectacular results, with second-quarter operating profit multiplying 19x year-on-year and even exceeding the combined earnings generated over the previous three years. In fact, it became the most profitable company in the world, ahead of Nvidia. On paper, the quarter confirms that the memory cycle remains firmly in expansion mode, driven by demand from AI data centers, aggressive price increases in DRAM and NAND, and the spillover effect of HBM into more conventional products. However, the market did not buy the narrative. The stock fell sharply, dragging SK Hynix and the KOSPI down with it, wiping out more than $80 billion in market capitalization.

As always, the excuse is that the good news was already priced in. Samsung had come off a strong rally, and the market did not just want a good quarter; it wanted confirmation that the memory supercycle can be sustained beyond 2025. And that is where the doubts begin, especially in such a cyclical sector. People often say that this time is different, but it never is. And in a sector this cyclical, where multiples depend more on the next turn in inventories than on the current quarter, a small nuance in pricing can matter more than a big earnings headline.

The underlying issue is not Samsung, but the sustainability of AI capex. The market is starting to question whether Meta, Microsoft, Amazon or Alphabet can continue increasing infrastructure investment at this pace, especially when much of the economic return remains uncertain. The bullish thesis in semiconductors has rested on a very powerful idea: hyperscalers have no alternative and must keep building capacity almost regardless of price. But if capital starts demanding discipline, the trade changes completely. Morgan Stanley is already talking about greater caution in capex, and that is enough to puncture a sector that had been riding a historic rally.

Another classic cycle risk is also emerging: supply. Samsung and SK Hynix have announced massive investments to capture AI demand, but every new dollar of capacity planted today can turn into tomorrow’s excess if demand cools. It is the old semiconductor story: when margins are extraordinary, everyone invests; when the new capacity arrives, the cycle may already have turned. For now, shortages in conventional memory continue to support prices, but the market is beginning to look beyond the quarter and ask whether this is a structural shortage or simply a very profitable cyclical peak.

China is beginning to treat its artificial intelligence models as strategic assets. Beijing has held meetings with Alibaba, ByteDance, and Z.ai to explore potential restrictions on international access to their most advanced models, including some that have not yet been released. This is not merely about limiting downloads or partially closing open-weight models, but also about treating any leak or theft of proprietary technology as a national security matter, while controlling who is allowed to finance domestic AI startups.

The move fits perfectly with the sector’s new geopolitical logic. Artificial intelligence is no longer a purely technological industry; it is becoming critical infrastructure, much like semiconductors, enriched uranium, or rare earths. The United States had already taken similar steps by restricting foreign access to Anthropic’s most sensitive models, particularly Mythos, which was designed for cybersecurity professionals and is viewed in China as a potential offensive tool against its interests. Beijing is now responding with its own version of export controls, applied not to hardware, but to the model itself.

The key point is that Chinese models had begun to gain global traction because of a combination that was deeply uncomfortable for the West: capabilities increasingly close to those of American models, but at much lower costs. Qwen, Doubao, and GLM-5.2 were becoming genuine alternatives for companies unwilling to pay the prices charged by OpenAI, Anthropic, or Google. If China limits international access to these tools, the effect could be twofold: on the one hand, it would protect its domestic technological advantage; on the other, it could raise the cost of AI adoption outside China by reducing competitive pressure on US models.

What is particularly interesting is that this also marks a shift in the open-source narrative. For years, model openness was presented as an inevitable, almost ideological force. But once the technology becomes strategically sensitive enough, the rhetoric changes. According to a Chinese legal debate published in a journal linked to the Supreme People’s Court, a tiered system is already being considered, with basic tools subject only to registration, advanced technologies requiring security reviews, and frontier models either banned from public release altogether or restricted to domestic use. In other words, open source is acceptable only as long as it does not touch the nerve of national power. At its core, we are seeing the same dynamic that unfolded in semiconductors, but at a higher layer of the value chain. First, controls were imposed on chips, lithography machines, and computing capacity. Now the output of that computing power—the models themselves—is beginning to be controlled. Technological globalisation continues to unravel layer by layer, and each bloc is trying to lock down its best assets before the other side can acquire them, copy them, or use them against it.

For companies, this introduces a new source of risk. Building products on cheap foreign models may be attractive in the short term, but if access ultimately depends on political decisions in Beijing or Washington, the real cost lies not in today’s pricing, but in the fragility of the technology supply chain.

AI is being nationalised at remarkable speed. And, as is often the case, the market continues to value many companies as though the digital world were infinitely scalable, open, and borderless. Every week brings more evidence that this era is over.

Russia has temporarily banned diesel exports until July 31, after Ukrainian attacks on refineries began to have a visible impact on the domestic market, leading to gasoline shortages, queues at filling stations, and growing pressure on prices. Moscow is now trying to redirect fuel toward domestic consumption, while opening the door to imports to offset the deterioration in its refining capacity. The measure also comes at a time when Russian exports had already collapsed. In June, seaborne shipments of diesel and gasoil fell by 39% month-on-month and 46% year-on-year, to approximately 1.8 million tonnes. During the first eight days of July, exports stood at just 214,000 barrels per day, well below the normal levels seen before the war. In practice, Russia was already operating under a de facto export ban.

The impact of Ukrainian attacks on Russian refineries is beginning to spread beyond the country’s borders and threatens to become a regional problem for Central Asia. The diesel export ban has exposed the enormous energy dependence of several countries in the region, which for years had assumed that Russian supplies were virtually guaranteed.

Kyrgyzstan is probably the most vulnerable case. The country depends almost entirely on Russia to supply its domestic market and has already removed price controls on AI-95 gasoline, while seeking alternatives from Belarus, China, Uzbekistan, Turkmenistan, Azerbaijan, and Kazakhstan. The problem is that many of these potential suppliers do not have significant spare refining capacity either. Kazakhstan, despite being a major oil producer, has made it clear that it will only export fuel after securing its own domestic consumption.

The situation is similar in Tajikistan, which imported more than 80% of its petroleum products from Russia over the past year. The government claims to have reserves covering around 60 days, but it is already seeking aviation fuel from Kazakhstan and Turkmenistan while tightening price controls. Beyond the immediate shock, the episode once again highlights one of Central Asia’s major structural problems. The region produces abundant energy resources, but lacks sufficient refining capacity, logistical coordination, and trade integration to guarantee its own supply. For decades, governments have prioritised raw-material exports and short-term political stability, while placing the development of a more resilient regional infrastructure on the back burner.

Despite the critical importance of these products and value chains, no one appears genuinely interested in maintaining them or investing in them. Not even during the greatest energy crisis in history—so far concealed by China’s buyers’ strike—have investment flows into the sector recovered. In fact, the industry has experienced the largest net capital outflow ever recorded. It certainly raises questions about the efficient-market hypothesis.

The ceasefire between the United States and Iran appears to have died before it even became anything resembling an agreement. Trump has declared the memorandum signed on June 17, under Pakistani mediation, effectively over, following the failure of indirect talks in Qatar and a new wave of U.S. strikes against Iranian positions. The 60-day window to negotiate a permanent solution has, in practice, closed, and the language has reverted to the usual one: military pressure, sanctions and controlled escalation… if such a thing even exists in the Middle East.

The immediate trigger was the attack on three cargo ships in the Strait of Hormuz, which Washington attributes to Iran, although Tehran has not formally claimed responsibility. The U.S. response has been twofold. On the one hand, new airstrikes on Iran’s southern coast, including strategic areas such as Bandar Abbas, Chabahar and Konarak, all relevant to the country’s naval and logistics infrastructure. On the other, the revocation of the license that allowed Iranian oil to be sold until August 21, shortening the settlement window for transactions to July 17. In other words, the tap is being turned off again exactly where it hurts most: energy revenues and the ability to finance the war. To be fair, these sales were highly unlikely anyway, as compliance mechanisms and caution worked against such a short window, and the most likely outcome is that this oil ends up in China, as usual.

Iran, as expected, has responded by raising the regional cost of the conflict. The attacks on Kuwait and Bahrain — both of which host a U.S. military presence — show that Tehran does not need to defeat the United States to make life difficult for it. It only needs to ensure that every move carries consequences in the Gulf, across maritime routes and among regional allies. In that sense, the Strait of Hormuz remains its major strategic card. Before the war, nearly one-fifth of global oil supply passed through it, and although the market has grown dangerously accustomed to geopolitical risk, the physical reality has not changed: if Hormuz is closed or partially disabled, there is no immediate substitute.

What is interesting is that oil barely reacted. Brent rose by around 1% to $78.8, far from the more than $120 seen in April. This tells us two things. First, that the market is still pricing in a limited escalation, not a full-blown war. Second, that there is enormous complacency around the ability to keep global energy flows open even in a direct conflict between the United States and Iran. The market may be right in the short term, but the asymmetry is clear: if the conflict remains contained, the impact is limited; if Hormuz breaks, the shock would be immediate.

The U.S. labor market continues to send mixed signals, allowing almost any narrative to be defended depending on which data point one looks at. Job openings rose slightly in May to 7.59 million, the highest level in two years, while the ratio of available jobs to unemployed workers remained at 1.04, broadly stable. At first glance, this does not look like a labor market deteriorating aggressively. But beneath the surface, the details are considerably less reassuring.

Hiring fell for the second consecutive month, to 5.17 million, and the pace of private-sector hiring weakened again, even though May’s nonfarm payrolls had shown solid job creation. The most uncomfortable data point comes from consumer perception. The share of respondents who consider jobs “hard to get” rose to 22.5%, the highest level since January 2021. At the same time, the Conference Board’s labor differential — the gap between those who see jobs as plentiful and those who see them as scarce — narrowed to 2.4 points, a level that has historically tended to correlate with a rise in the unemployment rate. In other words, even if companies are not laying off workers en masse, households are beginning to feel that the labor market no longer has the same momentum.

The problem with the current data is that there is increasingly more noise. The JOLTS report, already volatile by nature, has a very low response rate: barely 24% of contacted firms take part, compared with 35% two years ago and around 70% at the end of the previous decade. This increases the risk of bias and means the figures need to be treated with more caution than usual. Even so, the overall direction seems clear: for now, there is no collapse, but there is cooling. Job openings are holding up, layoffs remain low by historical standards, but hiring is moderating and quits remain depressed, a sign that workers no longer see it as so easy to switch jobs in order to improve their conditions.

For the Fed, the reading is relatively comfortable in the short term. As long as the labor market does not break, the focus can remain on inflation. In fact, with the conflict between the United States, Israel, and Iran apparently contained by a fragile truce and oil prices correcting, the central bank has somewhat more room to prioritize price stability. The market is already pricing in the possibility that the Fed could raise rates again this year, after keeping them in the 3.50%–3.75% range.

The institutional and moral deterioration in the United States is accelerating just as the country celebrates its 250th birthday. Trump has reported more than $1.4 billion in income from his crypto businesses in 2025, according to his latest financial disclosures filed with the Office of Government Ethics. The figure is extraordinary not only because of its size, but because of what it represents: the bulk of the president’s income no longer comes from hotels, golf courses, real estate licensing, or resorts, but from digital assets that have directly benefited from the regulatory shift of his own administration and from a pump-and-dump scheme.

The main driver has been World Liberty Financial, the crypto project co-founded by Trump and his sons, from which his companies received nearly $800 million. Within that figure, more than $520 million came from token sales and another $250 million from the sale of stakes in the business itself. On top of that, he earned another $635 million from the sale of his memecoins. Altogether, Reuters estimates that the Trump family has generated at least $2.3 billion from crypto-linked projects since his return to the White House.

Trump has pushed an openly pro-crypto agenda: federal rules for stablecoins, reduced regulatory pressure from the SEC and the Department of Justice, and an explicit narrative of turning the United States into the “crypto capital of the world.” The fact that, at the same time, a large part of his personal wealth has become dependent on that very same sector creates a conflict of interest that is hard to disguise, even if the president and vice president are legally exempt from many of the ethics rules that apply to the rest of the executive branch.

The White House denies any conflict and argues that the businesses are managed by his sons. But from an economic standpoint, that distinction matters little, since Trump remains the beneficiary of the trust that receives those revenues. This is not a marginal or symbolic exposure, but a complete transformation of his personal balance sheet. Just a year ago, World Liberty generated around $57 million for him in token sales. In the new disclosure, that figure has multiplied by nine.

Meanwhile, his traditional businesses continue to operate, but they have been pushed into the background. His golf courses and resorts brought in just over $500 million, with Mar-a-Lago jumping from $50 million to $77 million after consolidating itself as a kind of “Winter White House.” The Trump brand also continues to be monetized overseas, especially in the Middle East, with more than $50 million in real estate licensing income. The case illustrates a broader and increasingly uncomfortable trend: the line between political power, industrial policy, and private enrichment is becoming blurred. In other strategic sectors, the state takes stakes or attaches conditions to public support. Here, the movement is almost the reverse: a president with direct power to shape the regulation of an industry personally captures a huge share of the upside from that same industry.

The question is not whether crypto has gained institutional legitimacy under Trump. That is already clear. The question is who has benefited from that legitimacy. And in this case, the answer seems fairly obvious. The presidency has not only changed the regulation of the sector; it has also changed the personal fortune of the person leading it.

OpenAI has reportedly proposed that the U.S. government take a 5% equity stake in the company, in a structure inspired by the Alaska Permanent Fund, which would use the future capital gains of major AI companies to finance dividends or public benefits. The idea would not be limited to OpenAI, but would seek to extend to the rest of the major U.S. labs, although it is not clear whether companies such as Anthropic, Google, or xAI would be willing to accept something similar.

The news fits perfectly with a trend we have been seeing for months, in which the U.S. government is no longer acting only as a regulator or financier, but is beginning to behave like a strategic shareholder. We have already seen this in semiconductors, critical minerals, and sensitive supply chains, where subsidies and public support are increasingly being converted into equity stakes. Now that logic is being applied to artificial intelligence, probably the most important — and politically sensitive — sector of the next decade.

AI concentrates enormous valuations, requires a massive amount of capital and energy, and at the same time threatens to create highly visible labor-market disruptions. According to recent surveys, a significant share of Americans fear that AI could destroy jobs in their household. In that context, the idea that a handful of private companies should capture almost all of the economic upside from a technology built, in part, on infrastructure, data, collective knowledge, and public support is becoming politically difficult to defend. For OpenAI, the move may be a way to buy regulatory stability ahead of a future IPO. Giving 5% to the government may look expensive, but if in return it reduces the risk of more aggressive intervention, operational restrictions, or targeted taxes, it could end up being a reasonable insurance policy. Especially in an environment where Washington has already intervened in the launch timeline of advanced models, and where the line between national security, technology regulation, and industrial policy is increasingly blurred. In fact, this is one of the most discussed mechanisms for establishing a universal basic income.

If the government takes equity stakes in the major AI companies, even through a public vehicle, it would completely change the nature of the sector. We would no longer be talking only about private technology companies, but about semi-public strategic infrastructure, similar in importance to energy, defense, or telecommunications. That could reduce certain regulatory risks in the United States, but it also opens an international Pandora’s box: if Washington demands a stake so that its citizens can participate in the upside of AI, why wouldn’t Europe, India, or any other relevant market do the same? At its core, what we are seeing is the natural continuation of resource nationalism, but applied to intellectual and computational capital. In the past, governments wanted to secure oil, uranium, copper, or semiconductors. Now they also want to secure a share of foundation models. AI is becoming a strategic commodity — only instead of being extracted from a mine, it is trained in data centers.

The S&P 500 closed the first half of the year up 8.7%. That sounds like a calm, almost boring year, yet it does not capture what has actually happened. Just look at the Philadelphia Semiconductor Index, which rose 93.6% over those same six months, to understand where almost all of it came from. If you take the S&P itself and, instead of weighting it by size, give every company the same weight, the gain is just 11.2%. That gap between one version and the other is, essentially, one single sector pulling the cart. The Russell 2000 rose 21.3% and transports gained 26.3%, but the headline belongs to chips. What we call “the U.S. stock market” is today a leveraged bet on semiconductors, which in turn are a leveraged bet on hyperscaler spending.

And that spending, to be fair, is not smoke and mirrors. Amazon, Microsoft, Google, and Meta are set to spend around $725 billion in capital this year, 77% more than in 2025, with consensus already talking about surpassing one trillion in 2027 and Goldman adding up to $5.3 trillion between now and 2030. Nvidia collects the toll on that highway, and trades as if that toll will never stop being paid. What has changed in recent months is that the market has stopped applauding spending for its own sake and has started demanding accountability. Meta fell 9% on the day it raised its capex forecast. Alphabet was punished for the same reason. Amazon’s free cash flow is flirting with turning negative. And Nvidia, which seemed untouchable, saw $600 billion in market value wiped out in a single session in January, the largest one-day loss in history for a listed company. The underlying question is the same as always, even if it comes wrapped in shiny paper: does all this money earn its cost of capital, or does it merely look like it does?

Because a large part of the demand is suspiciously circular. More than $800 billion in deals have been counted where the supplier finances its own customer. Nvidia puts $100 billion into OpenAI, OpenAI signs with Oracle, Oracle buys Nvidia chips, and the same dollar ends up being counted as revenue across three different balance sheets. With the uncomfortable detail that OpenAI is on track to lose around $14 billion this year, almost triple last year’s figure, while credit default swaps on Oracle and Microsoft have almost doubled since September, and that MIT study reminds us that 95% of corporate AI investments still have not delivered a measurable euro in return.

I am not saying AI is fake. I am saying the price already assumes it is not, with no room for doubt or error. In the meantime, I continue buying barrels, copper, and businesses that throw off cash at six times earnings, while others pay forty times for the fish biting its own tail. The party is going wonderfully, it must be said. And in this same context, there are a number of companies that have been labeled as “irreversibly disrupted” and have fallen more than 80% in just over a year, after being compounders and examples of quality for decades, while their businesses do not yet seem to be suffering under the weight of that disruption. Some paradigmatic examples are or , which continue to improve their businesses and trade at less than 10 times earnings, although the doubts around their terminal value are by no means resolved. Opportunity or value trap?

Apple has begun passing the shock of higher memory and storage costs on to consumers, raising prices across several iPad, MacBook, HomePod and Apple TV models. For now, the iPhone has been left out of the adjustment, but that is probably only a matter of timing. Apple may prefer to clear the noise before the autumn launch and avoid having the narrative around the new iPhone dominated by headlines about price inflation.

Even Apple, with one of the most powerful supply chains in the world, can no longer absorb the rise in costs. The investment cycle in artificial intelligence is draining memory capacity toward Nvidia and the large data center builders, leaving consumer electronics in a much weaker position. Micron and other manufacturers are prioritizing longer-term, higher-margin contracts with AI customers, which makes complete economic sense, but it shifts pressure onto PCs, smartphones and consumer devices. This adds a new layer of tension to the AI narrative. Until now, the market has wanted to view data center capex as an almost exclusively positive story: more chips, more servers, more growth. But the same investment that is feeding the profits of Nvidia or Micron is also making basic components more expensive for the rest of the technology economy. In other words, AI does not only consume electricity and capital; it is also beginning to compete for scarce physical inputs.

The problem for Apple is twofold. On the one hand, price increases may partially protect margins, but at the cost of putting pressure on demand in categories that are hardly essential. On the other hand, if the iPhone eventually incorporates these increases, the risk moves to the heart of the business. Consumers have so far tolerated high prices for premium products, but doing so in an environment of upgrade fatigue, pressure on disposable income and an expected decline in PC and smartphone sales is a different story. It serves as another reminder that the AI boom does not happen in a vacuum. It has consequences for the real economy, supply chains and the margins of companies that, until now, seemed immune.

The Artificial Analysis chart summarizing the evolution of frontier model intelligence leaves a conclusion that is hard to ignore: the gap between U.S. and Chinese labs has compressed from years to months. At the top of the index, Anthropic, OpenAI, Google and xAI still dominate, but just below them an increasingly tight pack has emerged (DeepSeek, Alibaba, Kimi, MiniMax and Z AI) no longer advancing with the comfortable lag they had a year and a half ago, but right on their heels. Artificial Analysis itself attributes this to a combination of algorithmic improvements, increasingly capable open-weight models and fierce competition among local labs. All three factors point in the same direction: the frontier has become much easier to reach than the market had assumed.

What matters is not so much the national pride of who is in first place, but what that convergence means for the economics of the sector. If leading models end up delivering “sufficiently similar” output, differentiation evaporates, and with it pricing power. This is exactly the dynamic we discussed in relation to hyperscaler capex, where researchers rotate between companies, architectures are copied quickly and the end user can switch providers with uncomfortable ease.

When a Chinese open-weight competitor also appears, performing at a comparable level and deployable almost for free, the pricing floor for the entire industry sinks a little further.

These companies are being valued as if they were going to capture monopoly rents, when the competitive structure increasingly resembles a commodity market. Billions of dollars in compute, energy and talent are being deployed under the assumption that technical leadership will be defensible and long-lasting. But a chart showing half a dozen labs converging within a matter of quarters is the exact opposite of a moat. In fact, UBS says that 60% of companies are already controlling their AI budgets and moving toward cheaper options (some of them Chinese) after paying bills more than 200% above their allocated quotas.

China is once again reminding us that the energy transition is neither linear nor remotely free. After the first decline in coal-fired power generation in a decade, the country now appears headed for a rebound in consumption, driven by a combination of rising electricity demand, lower hydropower availability, weaker wind and solar generation, and more expensive gas, which is being pushed back into a marginal role for peak-demand periods.

The data is relevant because it once again dismantles the simplistic narrative of “more renewables = less coal.” China has installed renewable capacity at an unprecedented pace, to the point of reaching in 2024 its 2030 target of 1,200 GW of solar and wind. But capacity is not useful generation, and useful generation is not firm availability. When wind weakens, solar utilization falls in western provinces, hydropower is hit by El Niño and electricity demand keeps rising because of air conditioning, manufacturing, electric vehicles and data centers, the system falls back on the only source capable of dispatching massive and stable power: coal.

On top of this comes the geopolitical effect. If the war in Iran pushes LNG prices higher and China reduces imports to avoid the cost shock, gas stops being a real alternative and is reserved for moments of maximum stress on the grid. In practice, coal acts as the energy insurance policy of the Chinese system.

The deeper problem for the aggressive energy-transition narrative is that China is still building new coal plants, and once they are in operation, they do not magically disappear from the energy mix. Many operate under medium- and long-term contracts that guarantee minimum generation volumes, making it harder for renewables to displace them even when clean capacity is available. In other words, the system keeps coal not only out of technical necessity, but also by contractual and institutional design.

Thermal coal remains an asset despised by the consensus, yet indispensable for sustaining the real electrification of the economy. Electricity demand is growing faster than the renewable system’s ability to replace firm generation, and every climate or geopolitical shock once again exposes the fragility of the narrative.

The changeover in Downing Street opens a new period of political uncertainty in the United Kingdom, although for now markets appear to have received the news with some relief. Keir Starmer has announced his resignation after acknowledging that he no longer has sufficient support within the Labour Party to lead it into the 2029 election.

In practice, his departure does not seem likely to trigger a long internal battle, but rather a controlled transition toward Andy Burnham, who has just returned to Parliament after winning in Makerfield and is emerging as the clear favourite. Burnham represents Labour’s desperate attempt to reconnect with an electorate increasingly tired of broken promises. The United Kingdom has now endured a decade of instability since Brexit, and the possible arrival of a seventh prime minister in ten years confirms that the British political system remains trapped in a deeper crisis of legitimacy than many are willing to admit. The names change, but the underlying problems remain: weak growth, deteriorating public services, tax pressure, high debt and a population increasingly sceptical of the political class.

In a country with the highest financing costs in the G7, limited fiscal space and growing investment needs in defence and infrastructure, any sign of institutional stability is rewarded. For now, Burnham has built more of a narrative than a programme. He talks about fundamental change and reducing the cost of living, but has yet to lay out a clear roadmap on the economy, foreign policy or defence. In fact, he has already made his first misstep by talking about revaluing pensions — in other words, more of the same. He will face the same wall: a state with little capacity for additional spending, bond markets unwilling to tolerate much more borrowing, and a society demanding visible improvements without being particularly willing to accept the cost of financing them.

The pressure from Reform UK adds another layer of risk. Farage has already called for a general election, presenting Labour’s leadership change as yet another internal manoeuvre by an exhausted system. It is a politically effective criticism because it touches a real nerve: British voters have watched prime ministers come and go without the country appearing to function any better. In that context, Burnham may benefit from an initial renewal effect, but he will have little time to prove that he represents something more than just a change of packaging.

The consensus has decided that the oil problem is already solved, and it has done so by looking at exactly the wrong variable. The bearish thesis has, on paper, a solid argument: flows through Hormuz have recovered strongly. Persian Gulf exports are back to 75–80% of their pre-conflict level, with 13 million barrels leaving the region in just three days, and the crude trapped in the Gulf — nearly 200 million barrels stranded on vessels — is being drained at full speed. With that snapshot, banks have rushed almost in unison to cut their forecasts. JPMorgan now sees Brent at $80 in the fourth quarter and $64 next year, Capital Economics and Wood Mackenzie have turned more cautious, and only Goldman is holding on to its bullish stance, with an $85 forecast for this year. Once again, the narrative is being built around the headline of normalization.

The problem is that this recovery in flows is not sustainable — loadings and entries into the Gulf remain depressed — and, above all, it does not change the arithmetic of the physical balance at all. While the market celebrates tankers moving again, global onshore inventories continue to draw down at the fastest pace in history: Kpler reported a 49-million-barrel drop just last week, and Barclays’ aggregate indicator, seasonally adjusted, is already at its lowest level in recent years, breaking below the five-year range. Buffers are being depleted, not rebuilt. The release from the U.S. SPR falls to zero from August onward, and U.S. commercial crude inventories will break below the psychological 400-million-barrel level — the threshold the administration fears most — as early as the second week of July. As so often happens in commodities, the damage has already been done before the price has even noticed.

Here lies the disconnect the consensus does not want to see: consumers do not use crude; they use gasoline and diesel. And that is precisely the signal revealing that there is no demand destruction. Refining margins remain elevated, U.S. refineries are running at 95% utilization and those in PADD 2 above 100%, while Chinese teapots are languishing at 50% and China’s crude imports fell by 4.7 million barrels per day year-on-year in June. If demand for final products were truly collapsing, cracks would not be holding up where they are. And if, even with that demand, you maintain a production shut-in close to 7–8 million barrels per day excluding Iran and Qatar, barrels are lost all the same: the math simply does not add up anywhere, unless one is content to look at the WTI screen struggling to hold $70 and declare the case closed.

And it is precisely at that point that positioning turns a reasonable thesis into an asymmetric opportunity. Hedge funds’ short positions in Brent have reached around $18 billion, the highest level in at least a decade, after tripling in three months. This is already the seventh consecutive week of selling, with almost $25 billion accumulated. Put differently, with inventories falling at the fastest pace ever recorded, the market is structurally shorter than ever. Complacency is not a risk to the thesis; it is the fuel.

As always, the geopolitical leg remains — and that has not been resolved either, no matter what the headline suggests. Iran once again showed this weekend that it can close the Strait when it suits its interests, this time under the pretext of the conflict in Lebanon and the use of the Omani route in traffic, and it is doing so even after the United States announced 60 days of sanctions relief on its crude.

SpaceX’s stock market debut is starting to lose some of its initial euphoric edge. After becoming the largest IPO in history and briefly surpassing a $2 trillion valuation, the stock has corrected by more than 15% over the past two sessions, although it is still trading more than 30% above its listing price. Nothing especially unusual after such a vertical move, but it is an early sign that the market is starting to distinguish between narrative and fundamentals, and that the gasoline of hype is running out.

The company went public amid a full-blown frenzy for anything combining space, artificial intelligence and Elon Musk. In the first few days, retail flows pushed hard, with more than $300 million in net purchases in barely three sessions. But that momentum has cooled abruptly: by midday Thursday, net purchases had barely exceeded $9 million.

The move has also dragged down other space-sector companies such as Rocket Lab, Planet Labs, AST SpaceMobile and Intuitive Machines, suggesting that part of the correction is not specific to SpaceX alone, but rather a broader expectations reset across the whole segment. The market had extrapolated too quickly that the success of the IPO automatically validated higher valuations for the entire space ecosystem. The most delicate point is precisely the valuation. SpaceX is not simply a rocket company with Starlink as its growth engine; the market is beginning to value it as a hybrid platform spanning space infrastructure, telecommunications, defence and artificial intelligence. The acquisition of Anysphere, the startup behind Cursor, for $60 billion in stock does show, however, that Musk masters capital allocation and knows how to use this elevated valuation. The whole manoeuvre has been perfectly executed, straight out of the altcoin playbook, using a low float and a high FDV — in this case, market capitalisation — to stage the show.

Kevin Warsh’s first meeting at the helm of the Fed has marked a far deeper regime shift than the decision to leave rates unchanged would suggest. The relevant move is not in the current level of the fed funds rate, but in the way the Fed communicates, how it reads the cycle and the institutional redesign Warsh wants to impose from day one. The new Fed is abandoning much of the expansive language and forward guidance that dominated the post-GFC era, returning to a much more sober, almost Greenspanian style. Warsh was clear: he does not want to pre-commit to the next move. There was also a wide dispersion in the dot plot.

In practice, this forces investors to operate with less of a communicative safety net, something that is theoretically healthy, but potentially uncomfortable for markets that have grown used to the Fed holding their hand. Although inflation remains clearly above target, the statement introduces a more favourable reading on productivity, capital investment and supply shocks, especially in energy. In other words, Warsh leaves the door open to the idea that part of the current inflation may be transitory in nature, but without using the complacent language of the previous era. The Fed is in no hurry to cut rates, and could even tighten further if the data fails to cooperate, as several of its peers around the world are already doing.

This leaves Trump in an uncomfortable position. His bet was to replace Powell with someone more willing to cut rates aggressively, but Warsh does not seem prepared to deliver that script so quickly. Over time, his Fed may well become more supportive of looser monetary policy, especially if productivity improves and energy inflation moderates, but the timing will not be the one the White House wanted.

The United Kingdom, Canada and Australia have converged within a matter of months on exactly the same measure: banning access to social media for children under the age of 16, with New Zealand already recommending the same path. Australia has had it in force since December, affecting Facebook, Instagram, TikTok, X, YouTube, Reddit, Snapchat, Threads and Twitch; Canada introduced its legislation on June 10, using the opportunity to also regulate AI chatbots; and the British government is selling it as a historic move to “give children their childhood back.” When three democracies from the same bloc adopt the same policy almost simultaneously, it is worth asking whether this responds to a spontaneous concern or to a form of coordination that prefers not to present itself as such.

The argument is, emotionally, almost impossible to refute: protecting minors. And that is precisely where the trap lies, because the problem is not the stated objective, but the mechanism required to achieve it. To prevent a child under 16 from entering a platform, it is not enough to simply ask their age; it has to be verified. And to verify that a user is over 16, the age of absolutely everyone has to be verified. The “age assurance” mentioned in these texts with antiseptic technical neutrality means, in practice, digital identity, document upload or biometric recognition for anyone who wants to open an app. In other words, the real cost of the measure is not paid by children, but by the anonymity of everyone else.

And this is where the logic of precedent matters more than the specific content of each law. Once the infrastructure for universal verification has been built, the marginal cost of expanding its scope is practically zero. Today it applies to one age group and a defined set of platforms; tomorrow that same scaffolding can be used to filter other content, other groups or other “risk categories” defined along the way. It is no coincidence that New Zealand is pairing its recommendation with the creation of a new regulator, nor that Canada is placing AI models in the same package. The perimeter of what can be regulated never narrows; it only widens, always under a justification that is uncomfortable to reject in public. The rhetoric follows the usual pattern: “most parents support it,” “Canadians are asking for it,” “this is what society demands.” Formulas that shield the measure from criticism, because opposing it becomes equivalent to being against protecting children. It is the perfect packaging for a silent transfer of control.

Everyone can judge for themselves whether childhood is made safer. What seems difficult to deny is that, layer by layer, the state continues to colonise spaces where it was not previously present, and that access to information — which is indeed a good for all ages — is gradually becoming a privilege subject to permission. It starts with one age group and a handful of apps. It never ends there.

The interim agreement between the United States and Iran represents one of the most significant geopolitical shifts in the Middle East in recent years. Not so much because of what it resolves — which, for now, is very little — but because of what it implicitly acknowledges: Washington has accepted that it cannot overthrow, isolate or force the capitulation of the Iranian regime through military means. After three months of war, thousands of deaths, global energy tension and the risk of regional spillover, Trump has chosen to freeze the conflict and open direct negotiations with Tehran. For Iran, this is already a strategic victory, especially considering the terms of total capitulation — symbolically signed in Versailles.

The agreement extends the ceasefire for 60 days, includes Lebanon within the de-escalation framework and opens negotiations on the nuclear programme, sanctions, frozen assets, the reopening of the Strait of Hormuz and a potential reconstruction programme worth at least $300 billion. Iran commits not to manufacture or acquire nuclear weapons, but keeps its current nuclear programme in place during the talks. This point is crucial: there is no immediate dismantling — and I do not think there will be — no technical surrender and no structural renunciation of nuclear leverage.

From Iran’s perspective, the balance is highly favourable. It secures the end of the war, avoids new sanctions during the negotiation period, obtains immediate waivers for oil and petrochemical exports, progressive access to frozen funds and the promise of massive reconstruction. It also gets the United States to lift the naval blockade and de facto recognise an Iranian role in the future administration of maritime traffic through the Strait of Hormuz, even if alongside Oman. In other words, Iran not only survives the conflict, but emerges with greater regional and international legitimacy than before.

For Israel, the agreement is much harder to digest, and this will probably be what blows up the peace. The war was framed as an opportunity to irreversibly weaken the Iranian regime, cut off its nuclear capabilities and reduce its proxy network. For now, the result appears to be the opposite. Washington signs directly with Tehran, imposes no clear restrictions on missiles or Hezbollah, and limits Israel’s operational freedom in Lebanon through the ceasefire. Netanyahu is trapped between a narrative of victory that is impossible to sustain and a reality in which his main ally has decided to end the war for its own reasons. Trump has even justified the Persians’ possession of ballistic capabilities, in a disgraceful speech.

The risk for Israel is twofold. On the one hand, it loses the ability to shape American strategy. On the other, Iran gains room to rebuild its regional architecture. Hezbollah, far from being isolated, is indirectly incorporated into the negotiating framework. Beirut protests that Iran cannot negotiate on behalf of Lebanon, but in practice the Lebanese conflict becomes part of a higher-level table between Washington and Tehran. The Gulf countries also come out of this uncomfortable. For decades, they have lived under the assumption that American protection offset their strategic vulnerability to Iran. The war has called that premise into question. First, because they could not avoid being spectators to an escalation that directly affected their energy security. Second, because they now see Washington accepting Iran as an unavoidable interlocutor. The message is clear: if neither the United States nor Israel has managed to subdue Tehran militarily, the rational strategy for Saudi Arabia, the Emirates or Qatar will be to accelerate regional accommodation, not confrontation.

Oil reacted violently to the downside, with clear signs of capitulation that I find hard to understand. The gradual reopening of the Strait of Hormuz and the absence of fees during the first 60 days should ease part of the geopolitical premium in crude, although Iran retains friction capacity by requiring permits 48 hours in advance. It is not a closure, but neither is it a full normalisation. The strait is operating again, but under a new reality.

Trump himself acknowledged that he had been pressured into signing peace at almost any price to avoid an economic depression, given that inventories are four weeks away from hitting critical levels. What the market is overlooking is that, even with a reopening of the strait now — which I consider unlikely — logistics make any solution within four weeks impossible, meaning we would still reach the level so feared by the administration. The damage has already been done.

Nationalization is back in fashion, and this time the target is not failed banks, electric utilities or strategic mines, but the major artificial intelligence companies. It is no longer just that Bernie Sanders is proposing that the state take 50% stakes in OpenAI, Anthropic and similar companies; Trump also appears to be exploring a similar idea. When two opposite political poles converge on the same point, it is usually a sign that the underlying consensus is changing.

The logic is that if AI is going to capture a huge share of future economic surplus, displace jobs and become critical infrastructure for national security, the state does not want to limit itself to regulating from the outside. It wants to participate in the capture of value. Sanders presents it as a way to socialize part of the benefits of a technology built on collectively generated data, knowledge and content. Trump frames it more as a partnership between companies and the American public, in line with his growing use of the state as a strategic shareholder. The precedent already exists. The administration took a 10% stake in Intel, as well as positions in rare earths and quantum computing companies. The difference is that applying this model to AI would have far greater implications, because these companies are not simple industrial suppliers, but potential central platforms of the future economy.

On the one hand, state ownership could reinforce the idea that these companies are too strategic to fail, reducing political risk and facilitating access to contracts, financing and infrastructure. On the other hand, it introduces a huge layer of uncertainty, with the politicization of governance, pressure on margins, possible limits on returns to private capital and the risk that the state increases its control over time. What begins as a minority stake can end up becoming a lever for intervention. We are entering a new phase of economic nationalism, where governments no longer trust the market to allocate capital neutrally in strategic sectors. Energy, semiconductors, uranium, rare earths, defense, quantum computing and now AI are all part of the same board. The underlying question is troubling: if artificial intelligence becomes infrastructure as essential as electricity, telecommunications or the financial system, will governments continue to treat it as a conventional private sector? Probably not. AI was born as a story of venture capital, technical talent and almost infinite scalability. It may end up looking much more like a strategic utility, with private shareholders, yes, but under ever greater state oversight and ownership. What worries me is the direction the United States has been taking lately… Who is John Galt?

The May inflation data confirms that the normalization scenario many had expected is once again becoming more complicated. U.S. CPI rose 4.2% year-on-year, its fastest pace since April 2023, marking three consecutive months of strong price pressures. The main catalyst has been the energy shock caused by the conflict in the Middle East, with energy prices up 23.5% year-on-year and gasoline rising by more than 40%. In other words, this is not classic demand-driven inflation, but rather a renewed cost shock that puts the Fed back in an uncomfortable position. As a result, rate cuts are moving further away. The Fed will probably keep rates unchanged, but the important point is the shift in bias. Until recently, the debate centered on when monetary policy would begin to ease. Now, with oil under pressure, employment still resilient and headline inflation accelerating, the market is even starting to consider the possibility of further hikes, although the bar remains high. In practice, Powell does not need to raise rates to tighten financial conditions; he only needs to remove any expectation of cuts.

The key nuance is that core inflation remains much more contained than headline inflation. Core CPI rose 2.9% year-on-year and only 0.2% month-on-month, suggesting that, for now, the energy shock has not broadly filtered through to the rest of the basket. It also seems that the impact of tariffs is beginning to fade in goods, with declines in furniture, new vehicles and core goods. This reduces the urgency of an immediate hike, but it does not eliminate the problem. If energy remains elevated for longer, as we believe it will, the risk of second-round effects increases: transportation, airfares, fertilizers, food, services and inflation expectations.

For households, the data is more worrying than it is for macro models. Real wages fell for the second consecutive month, which explains why economic discontent is not merely a matter of perception. Inflation is once again eating into nominal income growth, especially for middle- and lower-income families. If consumption remains resilient, it is partly because households are still drawing on savings, a dynamic that cannot be sustained indefinitely. Politically, the data is dynamite. Trump won by promising to bring inflation down, but this renewed spike hits him just ahead of the November midterm elections. His response, saying that he “loves inflation” and that it will fall like a rock once the war with Iran ends, captures the problem well: the White House wants to frame the rebound as temporary and external, but voters do not pay for gasoline or groceries in terms adjusted for geopolitical narratives. They pay nominal prices.

SpaceX is coming to market in the middle of a technological storm, but it is doing so with demand bordering on the absurd, with more than $350 billion in indicative orders for a deal seeking to raise $75 billion. A 4-4.5x oversubscription which, beyond the headline number, confirms that appetite for the great private winners of the past decade remains intact, even in an environment of extreme volatility, sharp Nasdaq declines and bitcoin trading 37% below its highs.

SpaceX is probably the most coveted private asset in the world, with a unique combination of leadership in space launches, Starlink as a global connectivity business and a narrative of almost unlimited future optionality. The company is selling the market something far more ambitious than rockets or satellite internet: the possibility of building critical infrastructure beyond Earth, including data centers in orbit to solve the energy and computing bottlenecks of artificial intelligence. It is an incredibly powerful story, almost inevitably designed to capture capital in a cycle obsessed with AI.

But that is precisely where the risk lies. The presentation talks about a $23 trillion opportunity in artificial intelligence, supported by the idea that SpaceX can escape the physical limitations of terrestrial infrastructure. It is a brilliant narrative, but also a very demanding one. The market is not only valuing the current business, but a promise of dominance across several industries that have yet to fully materialize: launches, connectivity, defense, space computing and AI. In this way, with little tangible yet but enormous potential TAMs, it becomes much easier to sell both the idea and the valuation.

Massive institutional demand does not eliminate valuation risk. In fact, it often amplifies it. Buying pressure can become so strong that it distorts the market in the short term, and some analysts have even suggested that part of the recent weakness in risk assets could be related to selling in order to fund allocations to SpaceX. If that is true, this would be a deal capable of absorbing liquidity from across the entire technology ecosystem. The company probably deserves a premium over almost any traditional listed company. That is not the question. The question is how much future can be priced in before even an exceptional company has to prove it in the numbers.

Positioning in oil remains extraordinarily attractive precisely because the consensus still fails to understand the basic arithmetic of the physical market. The survey shared this week by Goldman is almost a gift for anyone with a bullish view on crude, not because it provides new information, but because it confirms just how complacent the market still is. If onshore inventories are being drained at the fastest pace in history and there are still investors convinced that there is no supply problem, then the bearish narrative has not yet been purged. In fact, positioning is the most bearish ever recorded.

The confusion comes from looking only at the price of WTI. Because it is trading around $84, already far from the $100 seen just two weeks ago, many conclude that there cannot be a shortage. It is the classic mistake of extrapolating from the financial screen to physical reality. The price still does not fully reflect the tension because part of the adjustment has shifted from crude to refined products, and because governments have temporarily softened the shock through releases from strategic reserves. But that does not change the nature of the problem. What was once an obvious crude shortage has turned into a product shortage. China has restricted exports of refined products, independent refiners have stopped flooding Asia with cheap supply, and plants without sufficient access to crude have reduced throughput to preserve inventories. At the same time, SPR releases have artificially softened the crude market, shifting part of the pressure downstream.

The balance, in reality, remains very tight. Millions of barrels per day of production have been lost, and although part of that impact is offset by lower refining activity, reserve releases and inventory drawdowns, the lost barrel does not disappear from the math. If product inventories continue to fall globally and there is no COVID-style demand destruction, the deficit is still there. Whether it shows up in crude or refined products is more of an accounting distinction than an economic one. The important signal is that the market has managed to convince itself that there is no shortage. Even many oil traders seem to have accepted the idea that if the price has not exploded, the problem does not exist. But in commodities, the opposite is often true: the best setups appear when the consensus looks at the price, ignores inventories and underestimates the physics of the system. Here, complacency is not a risk to the bullish thesis; it is a central part of the opportunity.

Nvidia is once again making a move to bring artificial intelligence out of the data center and directly into the PC. The company has unveiled RTX Spark, a new chip developed alongside MediaTek and Microsoft, with which it aims to turn the personal computer into a machine capable of running AI agents locally, without always depending on the cloud. The idea is that if the next computing interface is not an app, but an autonomous agent working in the background, today’s hardware falls short. After the phase of massive training in the cloud, the next battle appears to be shifting toward the edge — that is, toward devices capable of running models and agents locally. Qualcomm is also framing it this way, describing 2026 as “the year of agents.” The logic makes sense: if these systems are going to operate continuously, with low latency, access to private data and a degree of autonomy, not everything can go through remote servers. There will be growing demand for distributed computing.

For Nvidia, this opens up a new growth avenue beyond data centers, where it already dominates overwhelmingly. With RTX Spark and the Vera CPU, the company is moving directly into the traditional territory of Intel, AMD, Qualcomm and Apple. It is not simply trying to sell a more powerful GPU for gaming or design, but to redefine the PC as an AI-native platform. The pre-market declines in AMD, Intel and Qualcomm reflect fears that Nvidia could do in PCs what it has already done in servers: impose its standard before the rest can react. Still, it is important not to confuse narrative with immediate adoption. The PC with local agents still needs clear use cases, reasonable costs and a real improvement over today’s cloud-based experience. The iPhone reorganized the mobile ecosystem; ChatGPT reorganized the software layer; Nvidia is now trying to do the same with PC hardware. Whether RTX Spark will truly become the foundational moment the market is promising remains to be seen, but the strategic direction is clear: AI is no longer just a tool in the cloud; it is beginning to become a permanent layer of the personal computer.

Lisa Cook reinforces an idea that the market still does not seem willing to fully accept: the Fed is not yet in a position to declare victory over inflation. Quite the opposite. Although she supports keeping rates unchanged for now, her message is clearly more hawkish than one might expect in a context where Trump has placed Kevin Warsh at the head of the Fed with the expectation of accelerating rate cuts.

The problem is that the macro picture has become much more uncomfortable. Tariffs are still feeding through into prices, oil has become more expensive because of the war with Iran, and the investment boom in artificial intelligence is creating additional pressure in chips, software and data center construction. In other words, inflation is no longer coming from a one-off shock, but from several simultaneous forces pointing in the same direction. Cook still expects some of these pressures to moderate in the coming months, but she acknowledges that the risk is clearly tilted toward more persistent inflation.

What matters is not so much whether the Fed is going to raise rates tomorrow, but that the bar for cutting them appears higher than part of the market is discounting. After five years of inflation above 2%, the concern is no longer just the monthly data point, but the possibility that companies and workers start embedding structurally higher inflation into their pricing and wage decisions. The other side of the mandate, employment, remains relatively stable. Cook acknowledges that the accelerated adoption of AI could lead to job destruction before generating new productivity gains, but for now she does not see enough deterioration to justify preventive cuts. Even Trump, who has been so vocal about rate cuts, now recognizes that the macro reality has changed and is no longer constantly calling for cuts, giving poor Warsh a bit more room to maneuver.

Anthropic has confidentially filed paperwork to go public in the United States, moving ahead of OpenAI in a race that is starting to look less technological and more financial. The company has not yet disclosed the size or terms of the transaction, but its latest funding round already valued it at close to $965 billion, after more than doubling its valuation since February. If confirmed, this would be one of the largest IPOs in recent history, not only because of its size, but because of what it implies for the narrative underpinning a large part of the US market.

It is worth remembering that SpaceX is also moving toward a mega-IPO at a $1.75 trillion valuation, while OpenAI has been preparing its own process for weeks. In other words, some of the most important private companies in the world are trying to access the public market almost at the same time, competing for an amount of capital that, while enormous, is not infinite. The market window is open, but precisely for that reason everyone wants to get through it before it closes.

Anthropic’s listing will carry a dual interpretation. On the one hand, it will likely be the first major foundation model company to show its audited numbers to the market, setting the framework through which investors will assess revenue, margins, cash burn, computational capex and dependence on cloud providers. On the other hand, it is taking the risk of being the first to face real scrutiny. Until now, the private market has valued these companies based on growth, promise and fear of missing out. The public market will demand something more uncomfortable: unit economics, visibility and returns on an extremely capital-intensive infrastructure. If Anthropic manages to convince investors that its business model can scale profitably, it will validate a large part of the ecosystem. If, instead, its accounts reveal a structure that is too dependent on compute spending, cross-subsidies or constant external financing, it could open a crack in the narrative surrounding the entire sector.

For its part, Alphabet has announced a capital increase of up to $80 billion to finance the expansion of its artificial intelligence infrastructure, in a move that confirms just how much the race for compute capacity has become a balance sheet war. The transaction includes a direct $10 billion investment from Berkshire Hathaway, split between Class A and Class C shares, at prices slightly below the market close. Alphabet had already raised its annual capex guidance to $180-190 billion, after acknowledging that demand for AI and cloud solutions is exceeding its available capacity. The problem is that this demand requires an increasingly aggressive investment scale. The company has raised more than $85 billion in debt over the past year, taking its total balance above $100 billion, and is now also turning to equity. In other words, even one of the best businesses in the world is beginning to need massive external financing to sustain the pace of the race.

Berkshire’s entry provides an important reputational endorsement. For the market, the fact that Buffett — or perhaps already Greg Abel — is strengthening Berkshire’s position in Alphabet lends credibility to the idea that AI spending is not simply a capex bubble with no return. But if a company with Alphabet’s cash generation needs to issue shares, raise debt and launch at-the-market programs to fund its rollout, then the capital intensity of the new technology cycle is far greater than many investors had been discounting. The market is valuing many companies as if we were facing a new Google Search, but the competitive structure looks much less like a natural monopoly and much more like an infrastructure race with uncertain returns.

The key point is that “AI” is not a single market. In search, Google built a flywheel that was extremely difficult to replicate: more users generated more data, better results, more advertisers, more revenue, more investment, and the cycle started again. In addition, the product was cheap to serve relative to its monetization power. In generative AI, at least for now, the dynamic is different. Leading models are converging, they are trained on similar corpora, researchers move between companies, architectures are copied quickly, and end users can switch providers with relative ease. If the output of Claude, Gemini, ChatGPT or Llama ends up being “similar enough,” differentiation compresses. If there is no clear moat, there is no pricing power. And if there is no pricing power, massive capex becomes a dangerous bet. Everyone is investing as if they were going to capture monopoly rents, but perhaps the market will end up paying commodity-like returns.

The oil market is behaving as if it still had room to maneuver, but physical reality is starting to tell a different story. In theory, prices should be pricing in demand destruction to prevent global inventories from reaching minimum operational levels. However, the market seems more focused on headlines about memorandums of understanding, ceasefire rumors, and false diplomatic signals than on fundamentals.

The problem is that this is no longer a distant risk. According to this view, we could be just a few weeks away from reaching critical inventory levels. What appears to be taking place is one of the fastest declines in onshore oil inventories in history. Even if the Strait of Hormuz were to reopen immediately, a large part of the inventory drawdowns would already be inevitable because of the lag between supply, consumption, and replenishment. With strategic reserve releases close to 9-10 million barrels per week, preliminary estimates point to another 13-14 million barrel draw next week. In addition, with U.S. refineries processing around 17 million barrels per day, U.S. commercial crude inventories would begin to reflect the full pressure of the physical deficit.

The situation is starting to resemble living paycheck to paycheck. Immediate shortages may be avoided, but any additional disruption — a refinery outage, a pipeline issue, or a hurricane — could trigger real pressure in gasoline and distillates. The market continues to treat the closure or control of the Strait of Hormuz as a temporary event, when it could actually represent a structural change as long as the IRGC retains effective control over the passage. Governments are using strategic reserves as if they were a bridge for just a few weeks, but if the problem is not resolved quickly, those reserves merely buy time while commercial inventories continue to drain.

This is the real geopolitical fork in the road: either the IRGC controls the Strait of Hormuz, or it does not. There is no comfortable middle ground. Analysts keep looking for a path toward diplomatic normalization, but unless there are massive concessions, I see that as impossible. Investment banks are already starting to move toward this conclusion, and JPMorgan is laying out a base case with Brent around $100 per barrel through the end of the year, assuming the Strait of Hormuz reopens this very month. But each additional month of closure would add significant pressure: around $5 per barrel in the third quarter and up to $15 per barrel in the fourth quarter.

Samsung has managed to avoid a massive strike in its semiconductor business, but the price paid may end up being higher than it initially appears. The agreement reached with the unions, approved by 74% of the workers who voted, allocates 10.5% of the operating profit of the chip division to special bonuses for its employees. On paper, the news removes an immediate risk: a strike by 48,000 workers lasting 18 days, which would have hit not only Samsung, but also the South Korean economy and an already extremely sensitive global chip supply chain.

The problem is that the agreement breaks with a deeply rooted corporate norm in South Korea. Traditionally, bonuses were calculated after taxes and within a more discretionary structure. Here, however, Samsung has committed in writing to distributing a fixed percentage of the operating profit of a specific division. This creates a very uncomfortable precedent, because other unions could use it as a reference point to harden their own demands. It is no coincidence that South Korea’s president, business groups, academics, and even minority shareholders have expressed concern. The line between properly compensating key employees and structurally transferring part of the profit pool from shareholders to workers becomes increasingly blurred.

Artificial intelligence has driven a sharp increase in profits in the memory business, and workers in that division want to capture a proportional share of the cycle. Some employees could receive bonuses close to $416,000 this year, while other areas of Samsung, such as foundry or consumer electronics, will receive much lower amounts. This introduces a significant internal fracture. The same company, the same logo, but a completely different labor reality depending on the division in which one works. In such a large conglomerate, that disparity can erode morale and create management problems deeper than the wage negotiation itself.

Ultimately, Samsung has bought labor peace, but not necessarily stability. The agreement solves the immediate problem, avoids a potentially highly damaging strike, and allows the company to continue competing in the memory cycle. But it also opens a Pandora’s box: more union pressure, greater tensions between divisions, potential shareholder litigation, and a precedent that could spread across the rest of South Korea’s corporate landscape. As often happens in major investment cycles, when an industry starts generating extraordinary profits, the real dispute does not take long to appear: who gets to keep them.

China is executing a much colder and more calculated strategy than the market appears to be pricing in. Amid the tensions in Hormuz, Beijing is not resorting to the usual playbook of releasing strategic reserves to stabilize the market. Instead, it is forcing a much simpler adjustment: reducing imports, cutting refinery utilization, prioritizing gasoline and diesel, and sacrificing the petrochemical chain. The signal is powerful, because a country that has imported close to 11Mb/d in recent years saw its purchases fall to 9.3Mb/d in April, while maritime arrivals in May and June point to extraordinarily low levels, around 6.5Mb/d.

The key point is that China is not willing, at least for now, to open the tap on its strategic reserves. Unlike in other periods of tension, such as the 2021 energy crisis, when the mandate was to secure supply almost at any price, Beijing is now blocking the use of strategic inventories and allowing only adjustments in commercial inventories. But even there, refineries are reluctant to draw down their stocks, while state traders are taking advantage of the situation to resell expensive West African cargoes and accumulate discounted Russian barrels. The equilibrium point lies in refinery utilization. If China tried to maintain the average processing rate of recent years, close to 14.1Mb/d, a drop in maritime arrivals below 9.2Mb/d would quickly force it to choose between drawing on strategic reserves or returning to the international market and paying elevated premiums. That is why the most realistic scenario is a cut in refinery runs of around 5%, to roughly 13.4Mb/d. With domestic production and pipeline flows operating normally, that level would allow China to hold out for several months with maritime imports in the 7.9-8.5Mb/d range without triggering an immediate shock in essential fuels.

The problem is that the early figures for May and June are even below that threshold. If arrivals consolidate near 6.5Mb/d, the 5% adjustment is no longer enough and the system starts to come under physical strain. China could take the cut to 10%, but that would mean entering much more delicate territory: protecting gasoline and diesel at the expense of crushing the availability of naphtha, LPG and other petrochemical feedstocks.

The domestic alternative is coal. China is pushing its Coal-to-Chemicals complex as a strategic shield under the next five-year plan, using cheap and abundant coal to produce olefins and methanol. But this substitute has clear limits. The infrastructure is concentrated in inland regions such as Inner Mongolia, far from the manufacturing heartland of the southeastern coast, which introduces meaningful logistics costs. In addition, coal gasification-based routes cannot replicate the full range of aromatics or the specialized chemical chains linked to LPG. It works as a patch, not as a full substitute for the barrel.

Once again, the risks around the real profitability of artificial intelligence are moving back to the forefront. The market narrative continues to operate under an almost religious premise: AI will be transformational, therefore any investment made today will eventually be justified. But those two ideas are not the same. A technology can be revolutionary and, at the same time, destroy capital for years for those financing the infrastructure required to build it.

The first data points are starting to become uncomfortable. Microsoft has reportedly cancelled a large part of its Claude Code licenses due to cost, Uber has allegedly burned through its entire 2026 AI budget in just four months, and several executives are beginning to publicly acknowledge that costs are harder to justify than expected. There is even talk of companies spending hundreds of millions on tokens because of a lack of usage limits. These are anecdotes, yes, but they all point in the same direction: enthusiasm around adoption is still not translating into clear economic returns.

The problem is that the market appears to be ignoring that second derivative. Stocks remain at all-time highs, while the implied return numbers on AI capex are beginning to deteriorate. According to FT/Panmure Liberum estimates, even under extremely favorable assumptions — revenue against capex, with no operating costs — most hyperscalers would fail to cover their investment between 2025 and 2030. Microsoft, Alphabet, Meta and Oracle would generate negative returns; only Amazon would appear to be positive. And that is before factoring in accelerated GPU depreciation, electricity consumption, salaries, inference costs and technological obsolescence.

That is where the real risk lies. Not in AI “not working”, but in AI working without generating enough profitability to justify the amount of capital being deployed. Hyperscalers are spending historic sums on data centers, chips and energy under the assumption that future demand will catch up with present supply. But that is not a certainty; it is a leveraged bet on the adoption and monetization curve. The comparison with 2000 is not perfect, but it is useful. The internet was real. It changed the world. But that did not prevent a huge amount of the capital invested during the bubble from being destroyed. The infrastructure survived, the users arrived, the winners were extraordinary… but many investors who bought too early, at impossible multiples and extrapolating infinite growth, took decades to recover their money.

AI will probably follow the same pattern. The technology is real, the productivity shift will be real, and there will be enormous winners. But it does not follow from that that every hyperscaler will smoothly earn back its capex, or that any price is justified. For now, the market continues to reward the promise while looking the other way when the bills show up. The relevant question is no longer whether AI will matter. It will. The question is who captures the economic return, when, and how much capital will have been burned along the way.

This is the common picture for oil inventories across almost every region of the world: they are now dangerously close to the minimum operational level required for the system to function. And there is no real possibility of reversing the trend, because, due to the logistics of tanker flows, even if Hormuz were reopened today — which is unlikely — the trend would not change for at least another 2-3 months, by which point the peak summer demand season would already be over.

In fact, market participants seem to believe that everything is fine, based on the fact that crude prices are falling and equity indices are making new highs. Recency bias leads them to assume that, if everything is fine today, it will also be fine tomorrow. The only reason we have not yet seen a rebound and spike in crude prices is the release of strategic reserves, which are acting as a buffer. But that buffer has an expiration date, and it is rapidly approaching. Worse still, it leaves the system even more fragile than it was at the start of the conflict.

Peace headlines keep coming one after another, although none of them has materialized yet, and even Trump seems to be taking the resolution of the conflict for granted. I do think we will see the signing of an MOU soon, but one in which the United States will have conceded on the release of funds and on control of Hormuz, and, of course, where the nuclear issue will not be addressed at all. A total defeat that they are trying to hide at all costs, but where the economic pressure of strangling 20% of global oil flows has carried more weight than ballistics. The opportunity is the most asymmetric one in years.

The era of subsidies in artificial intelligence is starting to run out in real time. Over the past two years, much of enterprise adoption has been built on an implicit premise: that inference costs would fall quickly enough to justify mass integrations, aggressive flat-rate pricing, and growing budgets. That premise is starting to break down.

The Microsoft case is paradigmatic. The same company that has invested $13 billion in OpenAI, and that also controls one of the world’s most important cloud infrastructures, has decided to cancel internal Claude Code licenses because token-based billing made the cost difficult to justify. We are not talking about a startup without financial muscle, but about a company with almost unlimited compute capacity. If even Microsoft looks at the bill and decides it is not worth it, the message for the rest of the market is fairly clear.

The problem does not seem to be the lack of usefulness of these tools, but rather the economic structure behind them. Usage-based billing is forcing enterprise customers to confront the real cost of running advanced models at scale. And that cost, once flat-rate plans or initial promotions disappear, is much higher than many companies had assumed in their internal models. Uber’s example points in the same direction: according to its CTO, the company reportedly consumed its entire AI budget planned for 2026 in just four months. It is a sign that adoption is moving faster than budgets can absorb.

At the same time, AI software prices in the United States have risen by between 20% and 37%, while GitHub — owned by Microsoft — is beginning to move away from flat-rate plans toward consumption-based models. AI companies, especially in a year of IPOs, can no longer keep giving away capacity indefinitely. They need to monetize actual usage, because the cost of serving these models remains material, especially in intensive workloads such as coding, agents, or automated workflows. This connects directly with the debate around Gemini Flash and the supposed structural collapse in inference prices. Yes, more efficient models exist and will continue to improve, but the commercial reality is more complex. Anthropic, OpenAI, and Google have raised effective prices in recent months, not lowered them. Competitive pressure is not eliminating the cost; it is merely redistributing it across labs, customers, and cloud providers. From here, the sector faces an uncomfortable dilemma. Either companies reduce usage to fit within their budgets, which would slow the revenue growth labs need to justify enormous valuations before going public; or labs cut prices again and absorb losses, further deteriorating unit economics that are already difficult to defend. In both scenarios, the narrative of unlimited growth starts to collide with a much more prosaic reality: someone has to pay the bill.

Korea is starting to show one of those classic symptoms of a late-stage euphoric phase: retail investors are not only buying the dominant narrative, they are buying it with borrowed money. The viral case of the civil servant who invested 2.3 billion won in SK Hynix, financing 1.7 billion through margin loans, is not relevant because of the individual himself, but because of what it reveals about the market mood. When these kinds of stories stop being seen as reckless and start being celebrated as examples of conviction, the cycle has already entered a different phase.

The Korean rally has been spectacular, driven by Samsung and SK Hynix itself. The KOSPI has gone from around 4,000 points to above 8,000 in less than half a year, fueled by enthusiasm around semiconductors, memory and AI-related demand. Consensus is beginning to project increasingly aggressive targets — J.P. Morgan is already talking about 9,000 points in its base case and 10,000 in its bull case — supported by a reasonable thesis: the memory cycle can remain “higher for longer” if AI demand continues to absorb capacity.

But the problem is not that the thesis is necessarily false. The problem is the price at which it is being bought and, above all, the risk structure being built around it. Leveraged positions in the Korean stock market have reached a record 36.47 trillion won, while the major brokerage houses generated 600 billion won in margin lending interest income in the first quarter alone, up 56% year over year. In other words, the brokerage business is directly benefiting from the transformation of retail optimism into leverage. The dynamic is familiar. First comes the rally, then FOMO, then financing that appears cheap, and finally the conviction that the cycle still has several years ahead of it. As long as prices keep rising, everything looks rational. Leverage accelerates gains, validates the narrative and attracts more participants. But margin is not permanent capital; it is unstable fuel. With rates at 7–9%, any sharp correction can turn a long-term thesis into a forced short-term liquidation.

This does not invalidate the structural thesis around SK Hynix, memory or AI. In fact, Korea is probably one of the markets most directly benefiting from the scarcity of advanced semiconductors and the global need for memory capacity. But it does suggest that the market is entering a much more fragile phase, where the fundamental direction may remain positive while the short-term asymmetry deteriorates. In fact, when looking at the outlook for corporate earnings, it is increasingly dependent on just a handful of names — 25% coming from only five companies — which is never a sign of market health.

Inflation is once again hitting new highs, leaving the Fed with no room to maneuver. April CPI rose to 3.8% year over year, the highest reading since May 2023, with a 0.6% monthly increase, while core inflation accelerated to 2.8%, clearly above the Fed’s target. Energy accounts for more than 40% of the increase, with gasoline up 28.4% year over year and fuel oil up 54.3%. But the truly worrying point is that core inflation is also accelerating — and this only reflects one month of higher crude prices. Services inflation remains alive, with shelter at 3.3% and food at 2.3%, meaning this can no longer be attributed solely to crude. The pass-through from the energy shock is beginning, not ending.

PCE, the metric that truly matters for the Fed, confirms the same story. Headline PCE is already at 3.5% year over year, and the Philadelphia Fed’s Survey of Professional Forecasters projects PCE at 4.5% and core PCE at 3.4% for the second quarter, compared with previous estimates of 2.7% just three months ago. In a very short period of time, consensus has gone from projecting a normalization toward target to accepting figures not seen since 2022. And this still does not incorporate the upward shift in the crude forward curve already visible in the market, with firms such as Piper Sandler projecting Brent above $100 for both this year and next. If this materializes, Q2 and Q3 will be substantially worse than the current data.

The transition from Powell to Warsh could not come at a worse time. Warsh was nominated with the explicit expectation of cutting rates, but the minutes from the latest meeting show that most of the committee already anticipates hikes if inflation continues to apply pressure. The market has gone from pricing in four cuts for 2026 to just one, and now assigns a 37% probability to a rate hike before year-end. Any attempt to steer the committee toward cuts with this data would be interpreted as pure politicization, destroying anti-inflation credibility. Trump’s proposal to suspend the federal gasoline tax follows the same flawed logic as cutting rates in this environment: stimulating demand in the face of a supply shock is the textbook example of what not to do. Higher for longer, including rates. If Warsh allows it.

I see the possibility of an agreement between the United States and Iran as highly unlikely, since there are two issues, both the nuclear question and the Strait, where the positions appear irreconcilable. Iran is once again using the Strait of Hormuz as a geopolitical lever, this time not through a direct threat to close it, but through a more subtle formula: the administrative redefinition of the area under supervision and the requirement for prior coordination for the use of frequencies within the maritime corridor. On paper, the measure is presented as a technical mechanism for managing maritime traffic, delimiting the area between Kuh Mobarak and southern Fujairah in the east, and between Qeshm and Umm al-Qaiwain in the west. In practice, however, it represents an attempt to strengthen Iran’s ability to control one of the world’s most important energy chokepoints.

The key issue is not only whether Iran carries out a physical disruption, but that it is showing a willingness to turn energy traffic into a tool of pressure. Marco Rubio has already confirmed that a toll in the Strait is unacceptable to them, and also to the Gulf countries, since it would give Iran de facto control over OPEC+. Therefore, in order to reach an agreement, they would have to concede either on the nuclear issue or on the Strait, or otherwise attempt to open it by force.

One of the most relevant points now starting to appear in reports from major banks is that the physical oil market no longer has much room to keep absorbing a closure of the Strait of Hormuz. JPMorgan puts it quite directly: Hormuz will have to reopen one way or another, because if it does not do so before June, global inventories will reach critical levels. The key issue is not only the volume of barrels that are no longer leaving the Middle East, but the speed at which visible inventories are being drained. In just two weeks, observable stocks — satellite data, crude in transit and EIA data — would have fallen by around 121 million barrels. Goldman has also validated this dynamic, estimating visible withdrawals of close to 8.7 million barrels per day during May. These are enormous figures, difficult to reconcile with the apparent calm in prices, which remain at the mercy of increasingly dubious and desperate headlines.

The problem is that even if Hormuz reopens, the market does not automatically return to normal. There is a logistical constraint that financial markets tend to underestimate. Shut-in production does not come back online overnight, cargoes do not instantly reappear, and export chains take weeks — probably months — to rebuild. According to JPMorgan, even if the strait reopened now, Middle Eastern flows would not return to normal until August at the earliest. In other words, the market would continue to lose inventories even after the reopening. This explains why many traders are waiting precisely for that moment to position long with greater conviction. A reopening of Hormuz could trigger an initial bearish reaction on the headline, but the physical fundamentals would remain bullish: shut-in production, logistical delays, falling inventories and a growing need to release strategic reserves. In fact, at the current pace, the market is heading toward a cumulative loss of up to 1.59 billion barrels, a figure that vastly eclipses the roughly 400 million barrels released from the SPR in previous episodes.

The paradox is that attempts to contain prices through headlines, reserve releases and political measures may be delaying the natural demand adjustment. Japan, for example, is carrying out one of the largest reserve releases in its history and trying to keep prices at pre-conflict levels, which in practice stimulates consumption rather than destroying it. In the United States, demand would even have increased by around 600,000 barrels per day, while the net decline across Japan, China, Europe and the U.S. would barely amount to 500,000 barrels per day across almost half of global demand. There is no clear historical analogue for a supply shock of this magnitude combined with low inventories, logistical delays and governments trying to cushion the impact on prices. The key message is that a reopening of Hormuz would not immediately solve the deficit; it would simply mark the beginning of a slow normalization process. Until then, oil will depend less on daily headlines and more on a much more uncomfortable physical reality: barrels are missing, and inventories are falling too fast.

The truce between the U.S. and Iran is once again on the verge of breaking down, although this comes as no surprise to us. The American proposal sought first to freeze the fighting and leave the more complex issues — the nuclear program, sanctions, security guarantees — for later. But Tehran’s response goes much further: an end to the war on all fronts, including Lebanon; compensation for damages; the lifting of the naval blockade; non-aggression guarantees; the reopening of oil sales; and recognition of its sovereignty over the Strait of Hormuz. In other words, unacceptable without an acknowledged humiliation on the part of the United States.

It is a maximalist position, but not an irrational one from Iran’s perspective. Tehran knows that its main leverage is not on the conventional battlefield, but in the energy market. Keeping Hormuz practically closed amounts to applying direct pressure on crude prices, American inflation, and Trump’s electoral calendar. With Brent above $109 and only a trickle of tankers passing through the strait, Iran is turning its military weakness into economic bargaining power. The problem for Trump is that the war is beginning to carry a visible domestic cost. Two out of three Americans believe he has not clearly explained the objectives of the intervention, and the rise in gasoline prices is arriving at the worst possible political moment, less than six months before key legislative elections. The administration can tighten sanctions, pursue Iranian crude shipments to China, or threaten to break the truce, but every day that Hormuz remains blocked increases the pressure on the American consumer and on the political credibility of the White House.

Moreover, Washington is still failing to build a solid international coalition. NATO allies do not want to send ships to reopen the strait without an international mandate and a broader peace agreement. The regional dimension is also becoming more dangerous, and according to The Wall Street Journal, the UAE has reportedly carried out attacks against Iranian targets, including a refinery on Lavan Island. Until now, several Gulf allies had tried to move cautiously, aware that they are vulnerable to Iranian retaliation. A more direct UAE involvement would open a new phase of the conflict, with a greater risk of horizontal escalation across energy, infrastructure, and maritime routes.

Trump’s visit to Beijing comes at a particularly delicate moment for U.S.-China relations, and it can be described as a total failure relative to the expectations that had been created. After last year’s clash, it was an attempt to stabilize a relationship that has become the central axis of the new geopolitical order: trade, critical minerals, Taiwan, AI, nuclear weapons, Russia, and Iran. Everything now runs through Washington and Beijing.

On the most tangible side, both countries seemed willing to announce mechanisms to facilitate trade and investment, as well as Chinese purchases of Boeing aircraft, agricultural products, and U.S. energy. This is the kind of agreement Trump understands well: visible headlines, American exports, industrial jobs, and a narrative of negotiating victory. The extension of the rare earths truce was also discussed, probably the most important point from a strategic perspective. China still controls a critical part of the global supply chain for essential minerals, and Washington needs to buy time while it tries to rebuild domestic capacity and diversify supply. The dependency remains, even if it is politically dressed up as an “agreement.”

But the truly relevant part lies in the difficult issues. Iran has become a central piece, since China is one of the main buyers of Iranian crude and therefore a key source of financing for Tehran. Trump wanted Xi to use that leverage to pressure Iran toward an agreement that would end the conflict that began after the U.S. and Israeli attacks in February, but the Chinese president offered no concession in that regard.

Taiwan will once again be the most structural point of friction. Beijing considers the island part of its territory and has intensified its military pressure in recent years, while Washington remains its main international backer and arms supplier. The Trump administration insists that its policy will not change, but the risk of miscalculation rises when both powers mix military rivalry, technological pressure, and domestic politics. It is the kind of tension that is not resolved by a summit; it can only be managed.

Meanwhile, after all the theatrics around last year’s tariffs, the trade deficit remains enormous, and it does not look likely to shrink much further.

In that sense, the trip has been an absolute failure for the United States from a strategic point of view, as it leaves without any major agreement, headline, or help on the Iranian front, where much of its hopes had been focused.

Inflation is once again complicating the macro picture in the United States. April CPI rose 0.6% month-on-month and accelerated to 3.8% year-on-year, the largest increase in three years, with a composition that is particularly uncomfortable for both the Fed and the White House: energy, food, and services all pushing higher at the same time.

The immediate cause lies in the energy shock triggered by the U.S.-Israel war with Iran. Gasoline rose another 5.4% after March’s record jump, diesel and other fuels advanced 17%, and energy accounted for more than 40% of the index’s monthly increase. The problem is that this no longer appears to be limited to the gas pump. Disruptions in the Strait of Hormuz, higher transport costs, and pressure on fertilizers threaten to spill over into the rest of the basket in the coming months. Moreover, for the first time in three years, prices are once again rising faster than wages, eroding real income and helping explain the deterioration in consumer sentiment.

Core inflation does not offer much relief either. Excluding energy and food, CPI rose 0.4%, the largest increase since January 2025. Part of the rebound is explained by a technical adjustment in rents after data collection problems during the government shutdown, but even so, the signal is clear: services inflation is still alive. With the labor market still solid and inflation once again moving away from target, the room for rate cuts disappears. In fact, the market is starting to assume that rates could remain in the current 3.50%-3.75% range until 2027.

Politically, the blow to Trump is obvious. His 2024 victory rested on the promise of reducing the cost of living, but the geopolitical conflict and rising energy prices have brought inflation back to the center of the debate just before the midterms — and the worst part is that this is a self-inflicted problem. The proposal to temporarily suspend the federal gasoline tax may ease the immediate pain somewhat, but it does not change the root of the problem, and stimulating demand in the face of a supply shock is a textbook example of what should never be done. Against such a complicated backdrop, Warsh takes over from Powell at the head of the Fed, facing a historic challenge with no clear solution in sight.

The oil market is entering a much more delicate phase than many analyses focused solely on “counting barrels” would suggest. The main idea is simple: the buffer has already been exhausted, and the damage has been done. First, the excess oil on water and floating storage disappeared; from now on, any further deterioration will have to show up as a visible decline in crude inventories, refined product inventories, or, ultimately, demand destruction. And that last part does not yet appear to be happening.

When refineries start running out of available crude in storage, their first reaction is not to shut down abruptly, but to try to protect throughput by buying more barrels in the physical market. That urgency translates into more aggressive backwardation in the front-month contracts. In other words, the market rewards the barrel available today far more than the barrel available in the future. That is why timespreads should remain strongly backwardated as long as the shortage is concentrated in physical crude rather than in a simple financial dislocation.

The natural consequence is that refining margins remain elevated. If the constraint is in crude, the regions with less access to sufficient inventories will have to pay more for the raw material, but they will also be able to pass that pressure on to refined products. Asia is the most vulnerable point: if its refineries cannot access the crude they need, they will begin importing more refined products. That additional demand from the East will pull margins higher in the West, first in Europe and then in the United States, incentivizing Western refineries to maintain or even increase throughput.

This process can feed on itself for some time. As Asia imports products, inventories of gasoline, diesel, and other refined fuels in Europe and the United States will fall quickly, perhaps within a matter of weeks. Western refineries, seeing better domestic prices, will have incentives to keep products in their own markets. Asia would then not only have a crude problem, but also a refined products problem. At that point, Asian margins could take another leg higher, allowing refineries to bid much more aggressively for any available physical barrel.

It seems that even the IEA, always so negative on crude, is now forecasting a deficit that will last at least until Q4, and probably beyond. That has very relevant implications not only for prices this year, but also for future years, given the pressure it places on inventories.

A few analysts, still not many, are already daring to put numbers on the physical reality we are observing and the consequences that follow from it. One example is Piper Sandler, which is now forecasting Brent above $100 for all of this year and next. Higher for longer.

GameStop’s bid for eBay looks more like financial spectacle than a conventional, serious corporate transaction. Anyone who watched the CNN interview where he was questioned about the financing will know exactly what I mean. Ryan Cohen is trying to buy a company almost four times larger, combining cash, stock, and a $20 billion financing promise in a transaction that the market, for now, does not believe. The evidence is in eBay’s share price, which rose barely 6% and remains well below the $125 offer price, reflecting a fairly low implied probability of the deal closing.

It is understandable that GME would want to acquire eBay, apply the same cost-cutting playbook it used at GameStop, and combine the online marketplace with a physical network of around 1.600 stores to compete more effectively against Amazon. On paper, it sounds ambitious, but the industrial fit is limited. eBay is an asset-light business built on charging commissions for connecting buyers and sellers; GameStop remains a traditional retailer, with inventory, physical stores, and a very different customer base. There is some overlap in collectibles, trading cards, and second-hand goods, but not enough to justify a transaction of this size.

The main problem is financing. GameStop has around $9 billion in cash, but it also has debt, and it is trying to buy a company of much greater scale. Issuing shares would massively dilute existing shareholders, as it would require more than doubling the current share base, while a leveraged buyout would turn the deal into one of the largest LBOs in history. In other words, for the deal to work, Cohen needs the market to accept a combination of faith, leverage, and GameStop paper. It is no surprise that analysts are asking for more details.

Even so, the offer could have consequences even if it fails. eBay had been performing well in the market, with investors recognizing its repositioning toward higher-value niches such as sneakers, luxury fashion, antiques, and collectibles. GameStop’s attempt could put it on the radar of other buyers, especially if the market starts to perceive that the business has more strategic value than its current valuation reflects.

The meme component is also back. Retail investors have reappeared as buyers of both GameStop and eBay, and some are already speculating about a kind of “GameStop Hathaway,” blending the Warren Buffett narrative with Reddit culture. But there is an important difference between building a rational conglomerate and financing a gigantic acquisition with a stock whose valuation depends largely on retail enthusiasm. Michael Burry has been especially critical, selling his entire GameStop position and arguing that the strategy is not transformational at all. In his view, the transaction is not really about competing with Amazon, but about dominating the collectibles and second-hand niche, at the cost of more debt and more dilution.

Half cash, half stock.

Coinbase has announced a workforce reduction of up to 14%, around 700 employees, as part of a broader process to turn the company into a leaner, faster, and more “AI-native” organization. Brian Armstrong’s official explanation combines the weakness of the crypto cycle and the inherent volatility of the business with the accelerating impact of artificial intelligence on internal productivity.

The cyclical part is easy to understand. Coinbase remains highly exposed to trading volumes and speculative appetite in the crypto market. Although the ecosystem is maturing — stablecoins, prediction markets, tokenization — revenues continue to swing sharply from one quarter to the next. In a bear market, cutting costs before the deterioration becomes deeper is a rational decision, especially for a company that has already lived through several cycles of euphoria and purge.

But Armstrong does not present the layoffs as a simple defensive measure, but rather as a structural reconfiguration of the company around AI. Fewer management layers, managers who are also individual contributors, small teams, “pods” capable of managing AI agents, and even the possibility of one-person teams combining product, engineering, and design. In other words, the company does not simply want to use AI as a tool, but to redesign its organizational structure around the premise that many white-collar tasks can be automated or compressed. For years, AI has been discussed as a productivity boost, but now we are beginning to see how that productivity translates into explicit headcount reductions. Not because the company is on the brink — Coinbase remains well capitalized — but because the standard of efficiency is changing.

If an engineer with AI can do in days what previously required weeks of work from a team, the traditional corporate structure starts to look bloated. Middle layers, pure managers, and large teams go from being a sign of scale to becoming a coordination cost.

Trump has signed several executive orders to revive Keystone XL and Dakota Access, two pipelines that had become symbols of Obama’s climate agenda and of environmental resistance in the United States. Keystone XL is the more symbolic case. Obama blocked it in 2015 because he believed that approving a major pipeline to transport Canadian crude to Gulf Coast refineries would undermine the United States’ credibility in international climate negotiations. Trump, by contrast, frames it as a matter of energy sovereignty and job creation, inviting TransCanada to resubmit its application and adding a politically useful condition: that the steel and pipes be manufactured in the United States. It is the same framework we will see repeated across many areas of his economic policy: more domestic energy, more infrastructure, more local content, and fewer environmental restrictions.

Dakota Access has a somewhat different dynamic, because the project was practically finished and only the section under the Missouri River remained, near the area where protests by the Standing Rock Sioux tribe and their allies were concentrated. For opponents, the issue was the risk to drinking water and culturally sensitive sites, but for the company, Energy Transfer Partners, the blockage was a political decision by the Obama administration. Trump explicitly aligns himself with this second interpretation and orders the pending permits to be accelerated.

What matters is that energy is once again at the center of U.S. industrial policy. The White House, especially after seeing the risks exposed by Iran, wants to unblock projects, reduce regulatory friction, and favor the oil industry. In practical terms, Dakota Access was far more likely to be completed quickly, since it was almost finished and did not require the same international process as Keystone, which had to cross the Canadian border and needed a presidential permit. But even before a single additional barrel starts flowing, the message has already been sent.

The outlook for copper supply continues to disappoint to the downside, as we have come to expect. After the revision and delays at Ivanhoe, we can now see that Freeport has also lowered its expectations over the past three months. A 10% cut to the production profile in a single quarter is highly significant for one of the world’s largest producers.

In fact, despite the economic slowdown that is already starting to filter through as a result of the Iran conflict, demand is holding up while supply keeps surprising to the downside, leaving the net balance more positive than expected. At the end of last year, when the Grasberg mine, one of the most important in the world, collapsed, we highlighted how relevant that event was for the copper price, which has risen more than 50% since then. Now, after issuing very optimistic guidance on the recovery and ramp-up, they have publicly acknowledged that it will take at least another year to restore activity. With prices comfortably above $6/lb, many explorers and producers are in a sweet spot to generate extraordinary value. If Iran allows it.

Another week in which we remain stuck in Groundhog Day, with constant peace headlines, almost always coming from the same unreliable source, which are then denied. In reality, there is no new information, just recycled material, but it achieves the desired effect of keeping crude prices depressed. The physical reality is that the Strait remains closed, and another 70 million barrels of inventories have been lost this week. In fact, the Iranian government has published the following form, which must be filled out by all vessels wishing to transit the Strait through a newly created body designed to manage the route, which seems to be an attempt to cement a new paradigm of control.

Additionally, tensions in the Gulf are escalating again just as the market was, wrongly, beginning to price in a diplomatic resolution to the conflict. The exchange of fire between U.S. and Iranian forces in the Strait of Hormuz, along with renewed attacks on the UAE, puts at risk a ceasefire that had barely been in force for a month. Trump has tried to play down the episode, stating that three U.S. destroyers crossed the strait under fire without suffering damage, while Iran accuses the U.S. of violating the ceasefire by attacking a tanker and civilian areas in Qeshm. As usually happens in these episodes, both sides claim victory and deny any meaningful damage, but the key point is not so much the military report as the strategic signal: the conflict is still very much alive.

Despite the constant headlines and manipulation, which mainly affect the front month, meaning the crude for most immediate delivery, the entire curve has shifted higher over the past month and is now above $75/bbl for all of 2028. That would make all the valuation models currently considered consensus in the industry obsolete. My view is that we will see another upward shift during May, making the numbers even more interesting. The oil market continues to show a level of tightness that consensus seems unable to recognize. Historically, a 1–2% imbalance in global balances has already been enough to move prices violently, because oil is not priced based on the average barrel, but on the last available barrel.

At present, even including the SPR, the accumulated shortage is around 8.5 million barrels per day. The deficit exists, the inventory drawdown is real, and the headline of the day does not change much. In commodities, noise dominates in the short term, but balances eventually prevail. And right now, those balances still point to a market far tighter than the dominant narrative suggests.

The Fed kept rates unchanged at 3.5%–3.75%, but the key point was not the decision itself, which was fully priced in, but the internal fracture within the committee. The vote was the most divided since 1992, with eight members supporting the decision to hold rates, three rejecting the dovish bias in the statement, and one, Stephen Miran, voting directly for a 25-basis-point cut. In other words, Powell is leaving behind a Fed split into three camps: those who want to wait, those who believe the next move should still be a cut, and those who no longer even want to hint at easing.

The change in language also offered quite a few clues. Inflation is no longer described as “somewhat elevated” but simply as “elevated.” It may seem like a minor nuance, but it is not. The Fed is acknowledging that the energy shock stemming from the war with Iran, with oil back above $100, makes the normalization narrative much more complicated. Until now, the market was still clinging to the idea that cuts were only a matter of time. After the meeting, futures no longer priced in any cuts in 2026 and even briefly assigned a non-negligible probability to a rate hike over the next year. The market is beginning to accept that the Fed may not have as much room to cut as previously thought, despite Warsh’s arrival, especially if the economy continues to grow at a solid pace and energy inflation threatens to filter through into expectations, wages, or core inflation.

Kevin Warsh will inherit a much less compliant Fed than he probably expected. If his intention was to arrive with a clear mandate to cut rates, the committee has just sent him a preventive message: there will be no blank check. The three hawkish dissenters did not vote against holding rates, but against maintaining the easing bias. If the oil shock fades over the coming months, the Fed will be able to refocus on the weaker side of its mandate: employment, growth, and the risk of a slowdown. But if crude remains high, the narrative changes completely. Central banking history is full of mistakes made by treating energy shocks as transitory when they ultimately contaminate the rest of the economy.

SpaceX has approved a new compensation package for Elon Musk that, even by Musk’s own usual standards, borders on science fiction. The company’s board has reportedly designed a pay plan tied to extraordinarily ambitious milestones: reaching a $7.5T valuation, establishing a permanent colony on Mars with at least one million people, and operating data centers in space with 100 terawatts of computing capacity.

The figure is difficult to put into context because there is simply no comparable precedent. We are not talking about improving margins, growing revenue, or expanding EBITDA, but about objectives that belong more to SpaceX’s founding mission than to the traditional corporate governance playbook. The package includes up to 200 million restricted super-voting shares if the milestones linked to Mars and valuation are achieved, plus another 60.4 million shares tied to additional valuation targets and the construction of orbital computing infrastructure. All of this would be in Class B shares, carrying 10 votes for every Class A share, further reinforcing Musk’s control over the company.

The obvious reading is that SpaceX wants to ensure Musk remains mentally anchored to the project ahead of a potential IPO. The company is reportedly targeting an IPO around June 28, Musk’s birthday, at a potential valuation close to $1.75T. If that figure already seems extraordinary, the plan approved by the board implies multiplying that value several times over and turning SpaceX not only into the dominant space company, but into critical infrastructure for a multiplanetary civilization and, potentially, for the future of computing.

The uncomfortable point is that SpaceX and Tesla are, in practice, competing for the attention of the same person. Tesla already justified its own massive compensation package by arguing that it needed to keep Musk focused on the company, especially at a time when the electric vehicle business faces greater competition, margin pressure, and a narrative increasingly dependent on autonomy, robotics, and artificial intelligence. Now SpaceX is using a similar logic, but with even more extreme objectives. The result is an incentive conflict that is difficult to ignore: two enormous companies, with different shareholder bases, trying to buy time and focus from the same founder.

If Musk does not achieve the targets, he does not receive the shares. But if he does, the reward could be enormous and come with even greater voting power. In theory, this aligns incentives with value creation. In practice, it also consolidates a structure in which the founder increasingly concentrates control, and where minority shareholders must accept that the investment thesis depends extremely heavily on a single individual. The risk, meanwhile, is just as obvious: extreme valuation, founder dependence, technological complexity, and objectives with no historical precedent. But so is the reason why the market may be willing to pay multiples that would look absurd in almost any other sector.

OpenAI is once again facing the same question hanging over the entire artificial intelligence industry: not whether demand exists, but whether it is arriving fast enough to justify the level of investment already committed. According to the WSJ, the company has reportedly missed several internal user and revenue targets in recent months, raising concerns within management itself about its ability to absorb the huge computing and data center contracts already signed or under negotiation.

The company continues to grow, and ChatGPT remains the benchmark consumer AI product, but the economics of generative AI require near-perfect scale. Every additional user consumes infrastructure, inference capacity, chips, energy, and data center resources. Unlike traditional software, where incremental margins tend to be extremely high, here the cost of serving demand remains material.

For months, the market has been funding the expansion of AI on the assumption that future demand will absorb any excess capacity. The other relevant angle is competition. OpenAI has reportedly lost ground to Anthropic in key markets such as coding and enterprise, precisely where monetization is most attractive and where customers can justify higher spending. This matters because the consumer market can provide scale and brand recognition, but the real profit pool will likely be in enterprises, productivity tools, workflow automation, and vertical software. If Anthropic begins to capture a growing share of that segment, pressure on OpenAI increases, especially given such an aggressive cost structure.

There is also a signal of product maturity. ChatGPT reportedly fell short of its internal goal of reaching 1 billion weekly active users by year-end, while dealing with subscriber cancellations and weaker adoption momentum. This fits with an idea we have been discussing: the phase of explosive growth and novelty-driven adoption is starting to fade. AI is no longer just a viral curiosity; it is becoming a recurring, useful, but more normalized tool. AI can be transformational and, at the same time, overcapitalized in certain parts of the value chain.

As it has done every Friday for weeks, Axios published a new “story” that seemed to bring permanent peace in the conflict between the United States and Iran closer. The objective, clearly, was to try to push crude prices lower, but the effect is becoming smaller each time, both in magnitude and duration, as it increasingly seems obvious that any solution will be complicated. If the Strait of Hormuz remains closed over the coming weeks, the oil market would enter territory that is almost impossible to model. Sell-side analysts will try to assign a price target to crude, as always, but the reality is that there is no comparable historical precedent. The potential supply loss — between 11 and 13 million barrels per day — cannot be absorbed through a simple price increase or marginal demand adjustments. The only useful reference point would be a COVID-style demand destruction shock: mandatory lockdowns and government-directed economic paralysis.

Outside that scenario, the market has never destroyed demand on this scale through purely economic mechanisms. Not even during the Great Financial Crisis of 2008 did we see anything resembling a 12% decline in global demand.

The latest inventory data are already starting to show the scale of the imbalance. The EIA reported a massive 24.1 million-barrel draw in petroleum inventories, even with a 7.1 million-barrel release from the strategic reserve. Commercial crude inventories in the U.S. alone fell by 6.2 million barrels. And this is happening while some VLCCs that were already loaded and heading to the United States are still being discharged, which is keeping imports artificially elevated in the short term.

But this buffer is temporary. Those same cargoes that are supporting imports today will turn into lower available volumes in one or two weeks. In addition, there is a fleet of empty VLCCs heading toward the United States, suggesting that the drawdown in commercial inventories could accelerate sharply. On top of this, refined product inventories are also likely to keep falling, while U.S. exports of petroleum products remain near historical highs. In practice, we could see weekly EIA reports showing crude draws of 10–12 million barrels even alongside 10 million-barrel releases from the SPR. In other words, an effective cumulative reduction of 20–22 million barrels in a single week. If that happens, even generalist investors will realize that the problem has become impossible to ignore.

The United Arab Emirates’ exit from OPEC represents a significant blow to the traditional architecture of the oil market, although its immediate impact is partially cushioned by the blockade of the Strait of Hormuz. Under normal conditions, one of the group’s largest producers leaving both OPEC and OPEC+ would be a clearly bearish signal for crude, as it frees Abu Dhabi from production quotas and allows it to increase output once logistical restrictions disappear. But while Gulf crude remains trapped, the ability to produce more does not translate into the ability to export more.

The decision reflects an increasingly visible fracture between the UAE and Saudi Arabia. For years, Riyadh has acted as the market’s central stabilizer, absorbing much of the political and economic cost of production cuts. The UAE, by contrast, has been investing for some time to expand capacity and no longer wants to accept limits designed for a balance that favors Saudi Arabia more than its own ambitions. This weakens OPEC’s coordination power over the medium term and could effectively mark the end of the organization, given the enormous loss of relevance — they now control 40% of global supply. If the UAE, one of the few countries with real spare capacity alongside Saudi Arabia, decides to act outside the cartel, Riyadh’s ability to manage the market is reduced. More importantly, it raises an uncomfortable question: if market adjustment increasingly depends on a single producer, how sustainable is Saudi Arabia’s role as the central bank of oil?

For Trump, the UAE’s exit can be sold politically as a victory. It fits his long-standing rhetoric against OPEC, which he accused of manipulating prices while the United States bore the military cost of defending the Gulf. The war with Iran has exposed a weakness that the market used to treat as a tail risk, but which is now at the center of the system. Around one-fifth of the world’s crude oil and liquefied natural gas passes through the Strait. If that artery is compromised, OPEC+’s weight in global supply falls mechanically, not because producers decide to cut output, but because they cannot place their production on the market. In fact, OPEC+’s share would already have fallen from 48% to 44% in barely a month, and could continue declining if production shutdowns intensify.

The “AI trade” thesis is starting to face its first major collision with physical reality. Over the past two years, the market has priced in an almost unlimited expansion of artificial intelligence infrastructure, as if capital were the only bottleneck. But the problem is precisely the opposite: there is plenty of money; what is missing is power, permits, transformers, electrical equipment, labor and social acceptance. According to Sightline Climate, of the 16GW of data center capacity planned in the U.S. for 2026, between 30% and 50% could face delays or cancellations, and only 5GW is currently under construction. The situation worsens for 2027: 21.5GW has been announced, but only 6.3GW has broken ground. In other words, there is a huge gap between press releases and the real capacity that can actually be connected to the grid. The market continues to value hyperscalers as if all that capex were automatically convertible into compute, but supply chains do not work that way.

The most obvious bottleneck is the power grid. Today’s data centers are not simply warehouses filled with servers, but giant industrial consumers that require firm power, redundancy and connections to infrastructure that takes years to build. Lead times for high-power transformers, which used to be 24–30 months, have stretched in some cases to as much as five years — as we saw, for example, in GE Vernova’s call. And without transformers, there is no data center, no matter how willing Microsoft, Amazon, Google or Meta are to spend hundreds of billions. The digital economy is once again dependent on very analog assets.

That is why hybrid solutions, or projects entirely independent from the grid, are starting to appear, with campuses incorporating their own gas generation, converted coal assets or renewables. But if major projects have to finance their own power plants in order to operate, it means the grid is not ready to absorb the demand. The problem is not only technical. Local resistance is starting to grow, as shown by Maine’s moratorium on large data centers until 2027.

As consumers begin to see higher electricity bills, pressure on water resources, noise, land use and few permanent jobs in return, social support could deteriorate quickly. Artificial intelligence was sold as a clean, intangible and almost magical revolution, but its material base is one of the most energy-intensive of recent history. The software narrative is colliding with the politics of infrastructure.

Kevin Warsh’s appearance before the Senate left a fairly clear message: he is trying to sell institutional continuity in form, while showing a willingness to significantly change the Fed’s internal functioning. Politically, the most relevant issue was to dispel the suspicion that Trump had chosen him in exchange for an explicit commitment to cut rates, and on that point Warsh was categorical. He said Trump had not asked him to promise rate cuts, nor would he accept such a demand. It was an almost mandatory answer, because the market’s main question is not only whether he will be confirmed, but to what extent he would act as an independent Fed chair or as an extension of the White House.

That said, denying such a pact does not mean the risk disappears. Trump has already publicly suggested that he expects lower rates under Warsh’s leadership, and Warsh himself avoided saying whether he would cut rates today or not. That ambiguity allows him to defend formal independence without directly confronting the president before being confirmed.

Where he did provide more substance was in his reform agenda. Warsh believes the Fed made a serious mistake with post-COVID inflation. On the one hand, he wants to review how inflation is analyzed, incorporating more data tools and adapting the framework to an environment in which AI could alter productivity, employment and prices. On the other, he wants a Fed that is less obsessed with verbally guiding the market. He criticized the fact that too many officials constantly speak about where rates should go, and hinted that he would prefer less forward guidance, more genuine internal debate and less chorus-like communication. For years, central banks have tried to control every market expectation through speeches, interviews and advance signals. Warsh appears to think this has gone too far and has ended up making the process more rigid.

There were also uncomfortable signals. He avoided commenting on the Administration’s political pressure on Powell, on the attempt to remove Lisa Cook, and even dodged saying that Trump lost the 2020 election. At the end of the day, he is Trump’s man.

The thesis of a ramp-up in U.S. LNG exports is starting to take shape. The start of operations at Golden Pass marks an important moment for the global LNG market — not so much because of the initial volume, which is still limited to the first train, but because it signals the beginning of a trend. The project, a joint venture between QatarEnergy and Exxon, has been under construction for seven years and aims to reach 18 mtpa once its three units are operational. For now, only Train 1 is running, but the first cargo has already departed from Sabine Pass, strengthening U.S. export capacity at a time when the global LNG balance has tightened abruptly.

And it comes at the ideal moment. Before the Iranian conflict, the market expected a relatively comfortable environment for 2027 and 2028, with surpluses of 1.7 and 2.1 Bcf/d, respectively. After the conflict, and the force majeure declared by Qatar, that view changes completely: 2026 moves from a moderate deficit of 0.5 Bcf/d to a much more severe deficit of 3.5 Bcf/d, while 2027 and 2028 also move out of surplus and into deficit territory. In other words, the market not only loses the expected cushion, but enters a situation of persistent scarcity just as Europe and Asia continue competing for flexible molecules.

Golden Pass’s first cargo helps, but it does not yet change the global balance. Capacity arrives gradually, with delays, operational ramp-ups and execution risks.

The energy crisis stemming from the war with Iran once again puts Europe in an uncomfortable position. Normally, traders buy gas in summer, when demand falls and prices are lower, in order to sell it in winter at higher prices. But today, summer prices have risen to almost winter-like levels, driven both by supply constraints linked to the conflict in the Middle East and by regulatory pressure to refill inventories. European storage levels are barely at 30%, with Germany at 23.5% and the Netherlands at a worrying 7.4%. The European Commission appears willing to lower its ambition in practice, keeping the 80% target but opening the door to 75% in order to ease pressure on prices. It is a pragmatic solution, but it also reveals the underlying dilemma: if refilling is pushed too hard, summer gas prices rise; if targets are relaxed, the risk increases of having to buy on the spot market during winter, probably at higher prices.

Even so, the most likely scenario is not a physical supply crisis, but another round of high prices. Europe can pay more than almost any other region, and it will do so if necessary. The problem is that this purchasing power shifts the adjustment onto poorer countries, which are priced out of the market when Europe and Northeast Asia compete for the same LNG cargoes.

In other words, Europe will probably avoid blackouts, but at the cost of exporting scarcity to the rest of the world and continuing to burden its industry with a structural energy disadvantage.

The situation in the Strait of Hormuz is escalating again. Iran seized two vessels on Wednesday, tightening its control even further over the world’s most important energy chokepoint, just after Trump unilaterally announced that he was indefinitely suspending new attacks against Iran. The problem is that this supposed extension of the ceasefire does not really appear to have been agreed by both sides. Washington is presenting it as a pause to give room to an Iranian proposal, but Tehran has not publicly validated that version and, in fact, continues to denounce the ongoing U.S. naval blockade. For Iran, as long as that blockade remains in place, there is no real ceasefire.

That is the core point of the story. This is not an orderly de-escalation, nor are negotiations clearly on track. It is a very fragile, almost improvised pause, in which each side keeps its position of strength fully intact. Trump is once again pulling back at the last moment after threatening to strike critical Iranian infrastructure, but he is not offering a clear diplomatic off-ramp or a concrete timetable. The White House has also given no new deadline for concluding talks.

Meanwhile, Iran is responding by reminding everyone that the strait cannot reopen normally while the U.S. maintains its blockade on Iranian maritime trade. In other words, the truce exists more in the American narrative than in the operational reality on the ground.

The seizure of the two vessels matters precisely for that reason. According to Tehran, the ships committed maritime violations and altered their navigation systems. In practice, the formal justification hardly matters: what matters is that Iran is signaling that traffic through the strait will continue to depend on its will unless the framework of the conflict changes. The fact that another vessel was also attacked, even though it was able to continue, reinforces the idea that commercial navigation in the area remains under real threat.

Every few days, we wake up to another headline trying to offer hope that peace is close. But oil prices are responding less and less to those theatrics, because the negotiating positions remain, as of today, very far apart.

The physical oil market is starting to enter a phase in which spreadsheets become less useful and the numbers are now so large that they are self-evident. The thesis is simple, although its implications are enormous: if the closure or disruption of the Strait of Hormuz has removed between 11 and 13 million barrels per day from the market, that loss has to show up somewhere. It does not disappear by magic. It will either be reflected in falling crude inventories, falling refined product inventories, or demand destruction. Most likely, all three at once.

What matters is that this reading is no longer coming only from external analysts or investors with a bullish energy thesis. Vitol and Trafigura, two of the major players in global physical trading alongside Shell, appear to be validating an accumulated loss close to 1 billion barrels. Between the three of them, they move around 27–28 million barrels per day, roughly a quarter of the global market. In other words, these are participants who are not giving opinions from a screen, but from real flows of cargoes, refineries, storage and freight.

The breaking point for the market comes when the last barrels loaded before the disruption finish unloading at destination. While vessels are still in transit, the system appears to function with a degree of normality. But once those cargoes are consumed, the real drawdown of onshore inventories begins. A reduction in refinery throughput should not be confused with an equivalent fall in end demand. If Asia and Europe reduce processing because they do not have enough crude, that does not necessarily mean consumers are using less gasoline, diesel or jet fuel. It means the pressure shifts from crude to refined products. In that scenario, product inventories can fall faster and end prices can rise sharply, eventually generating the demand destruction the market needs to rebalance.

The logistical bottleneck is just as important. Even if Hormuz reopened tomorrow, the market would not return to normal instantly. It would take weeks, perhaps months, to reorder flows, mobilize vessels and drain the excess crude trapped in storage in the Middle East. Flow estimates still point to a huge deficit even after incorporating buffers: 12 million barrels per day of shut-in production, 2.5 million barrels per day of strategic releases and 2 million barrels per day of demand destruction would still leave a net deficit of around 7.5 million barrels per day. That is a figure which, if sustained, would consume inventories at a pace incompatible with the current complacency. If, in addition, U.S. commercial inventories fall below 400 million barrels and move toward the operational minimum of 370–380 million despite a massive SPR release, the market will have to violently reprice scarcity risk.

Financial markets, with major indices at all-time highs, still seem to be debating whether the disruption is “real” or temporary, while the physical market is already starting to operate as if the shock were inevitable. Even if Hormuz reopens, onshore inventory draws are probably already underway. And when those draws become visible in official data, the consensus may discover too late that this was not geopolitical noise, but one of the largest logistical and energy shocks of recent decades.

U.S. utilities are bracing for a historic wave of investment. They need to modernize an aging power grid and absorb another jump in demand, driven above all by data centers tied to the artificial intelligence boom. According to PowerLines, 51 publicly traded utilities expect to allocate around $1.4 trillion over the next five years, more than 20% above what they were planning just a year ago. The grid needs maintenance, hardening against storms and wildfires, and expansion to connect new industrial load, electric vehicles, and now data centers with electricity consumption comparable to entire cities and round-the-clock demand. The problem is that the regulated logic of the sector turns much of that CapEx into future rate hike requests. And this is happening at a politically sensitive moment, with electricity bills rising faster than overall inflation and affordability becoming a bipartisan issue.

Residential electricity prices have already been climbing sharply, and regulators are approving a very significant share of spending requests. As is often the case, residential customers appear to have less bargaining power than large industrial or commercial users, so the cost allocation is far from neutral.

The big unknown revolves around AI. Many utilities and tech companies argue that the arrival of large customers helps spread fixed costs and, in theory, should ease pressure on the rest of the system. But that only works if the projected demand actually materializes and remains in place. If some of these projects are oversized, delayed, or never built at all, the system risks investing today for future demand that may not exist at the promised scale. That is when the worst-case scenario appears: stranded assets and traditional consumers footing the bill for excessive optimism. From a market perspective, this will create another tailwind for copper demand, which already faces major supply problems in the main producing regions worldwide. The sulfur shortage resulting from the Iran crisis will only worsen this issue, which is already structural rather than temporary.

The urgent meeting between Bessent, Powell and the major American banks reveals just how far the risk associated with certain advanced AI models has moved beyond an abstract discussion and become a matter of systemic stability. The fact that the Treasury and the Fed have summoned the CEOs of the main institutions is not a response to the usual noise around technological disruption, but to something far more serious: the fear that tools like Mythos could sharply increase the offensive capabilities of malicious actors in the cyber domain. What is interesting is that the problem does not seem to be simply that AI can automate attacks, but that it can do so with a degree of autonomy, scale and sophistication far beyond anything seen so far. Based on what has been described, Mythos appears to have shown the ability to identify and exploit vulnerabilities in operating systems, browsers and potentially in environments where sensitive financial information circulates. For years, banking cybersecurity has been understood as an incremental race between attackers and defenders. But if certain models drastically lower the technical barrier to uncovering complex attack vectors, that balance could break much faster than regulatory systems and internal controls are prepared to absorb.

Anthropic’s reaction is also revealing. The fact that the company has restricted access to the model to a very small group of large technology and financial firms suggests that even the developer itself does not view this as a conventional launch. In essence, it is a form of preventive containment; when an AI company withholds a product not because of weak demand, but because of fear over its offensive implications, we are entering a different phase of the technological cycle. From a banking perspective, the concern is quite logical. Systemically important institutions do not just manage enormous balance sheets, they also rely on old, complex and highly interconnected technological architectures. In theory, large banks invest billions in cybersecurity, redundancy and operational controls. But precisely because of their size and complexity, they also present a massive attack surface. And if operational risk stemming from cyberattacks was already difficult to model, it may now become outright nonlinear. That is likely the implicit message of the meeting: this is not just another threat, but a possible acceleration of risk that renders many previous assumptions obsolete.

It is also striking that the move is not limited to the United States. Canada and the United Kingdom are holding similar discussions with their major institutions, reinforcing the idea that this is not seen as an isolated corporate issue, but as a cross-cutting vulnerability within the Western financial system. The most useful parallel may not be with a new fintech or another wave of automation, but with the emergence of a dual-use technology: valuable for defense, but potentially devastating in the wrong hands. It is becoming increasingly clear that AI will not only transform productivity, margins and business models; it may also reshape the map of systemic risk. Over the last two years, the market has tended to view artificial intelligence almost exclusively through the lens of growth, capex and corporate winners. But the other side of the phenomenon is now emerging forcefully: that of a technology which, in addition to creating value, can also amplify critical fragilities in infrastructures that were already vulnerable.

The IMF has cut its growth forecasts, but the most relevant point is not so much the baseline revision as the implicit recognition that the central scenario is already outdated from the outset. Formally, the institution still maintains a relatively benign case, assuming a short-lived conflict in the Middle East and a partial normalization in oil prices during the second half of 2026, with an annual average of $82 per barrel. The problem is that its own chief economist admitted just minutes later that reality already looks closer to the adverse scenario than to the reference case. In other words, the energy market remains under strain, the disruption in Hormuz continues, and each passing day pushes the world further toward an environment of lower growth, higher inflation, and greater financial fragility.

The intermediate scenario outlined by the IMF is already fairly uncomfortable. It assumes oil close to $100 this year and global growth of 2.5%, clearly below the 3.4% recorded in 2025. Even so, the real risk lies in the severe scenario, where crude would average $110 in 2026 and $125 in 2027, dragging global growth down to 2%. In practice, that would amount to skirting a global recession. This would not be just a classic supply shock, but a simultaneous hit to consumption, margins, financial conditions, and inflation expectations. The regional breakdown also sends a fairly clear message. The United States is holding up reasonably well thanks to fiscal support, the momentum from previous rate cuts, and AI-related investment, while Europe is once again showing its structural vulnerability to energy prices. China is suffering less than one might expect due to stimulus measures and lower tariff pressure, but it remains trapped by its underlying domestic problems. Where the blow is truly severe is in emerging markets and, of course, in the Middle East and Central Asia, which are facing a combination of infrastructure destruction, declining exports, and logistical dislocations. India stands out as a rare positive exception, benefiting from its cyclical momentum and a better relative position vis-à-vis the United States. At its core, the IMF’s message confirms something fairly obvious: the world was already sliding toward a more fragile equilibrium even before the conflict, and the energy shock is simply accelerating that drift. Without the war, they probably would have revised growth upward. With the war, the debate is no longer about how much the world will grow, but how much pain will be required to contain a new inflationary wave.

What we are seeing around the Trump crypto universe is not a disorderly collection of improvisations, excesses or isolated conflicts of interest. It is a business model. Or, more precisely, an extraction economy built around the presidential surname, where politics, media attention, ideological loyalty and financial speculation merge into a single circuit. The mistake would be to analyze each piece separately. The memecoin, World Liberty Financial, the stablecoin, private dinners, regulatory-rescued partners and pardons are not isolated episodes. They are different nodes of the same machine. The logic is deeply Trumpian. For years, the Trump name functioned as a commercial license. It was monetized through hotels, golf courses, towers, television shows and branding deals. But in this new phase, the underlying asset is no longer just fame, but political power itself. The presidency stops being merely an institutional platform and also becomes a platform for private monetization. Not as an accidental side effect, but as a central mechanism. Politics generates visibility. Visibility attracts capital. Capital is channeled into vehicles controlled directly or indirectly by the family. And that capital, in turn, strengthens the business, media and political network that sustains the brand. It is a perfect feedback loop. Crypto fits ideally into this scheme because it combines opacity, speed and narrative. It allows money to be raised with very little friction, dresses the operation in the language of innovation and disintermediation, and dilutes responsibility across entities, wallets, partners, foundations, white papers and corporate structures complex enough to muddy any accountability. In that sense, the technology is almost incidental. What truly matters is that it offers the perfect format to monetize community, access and political affinity.

The buyer is not just acquiring a speculative asset, but a symbolic form of proximity to the leader, an emotional stake in the brand and, in some cases, the fantasy of getting early access to the cap table of a new political-financial order. That is the essence of this grift economy. It does not sell intrinsic value; it sells belonging. And precisely for that reason, it can survive even when the fundamentals are poor, the allocation is obscene or the losses for retail investors are obvious. The economic structure that has emerged around World Liberty Financial fits this pattern perfectly. The fact that a disproportionate share of the revenue ends up in entities linked to the family, despite not bearing the main economic risk, is not a design anomaly. It is the design. The fact that insiders have better terms, more information or more favorable vesting schedules is not a governance flaw either. That is why reducing all this to “conflicts of interest” is so weak. The phrase is too small. A conflict of interest suggests an uncomfortable interference between public duty and private benefit. Here, we are beyond that. Here, private benefit appears to have been built from the start into the very functioning of the ecosystem. The boundary between political action, regulatory favor, business relationship and family monetization is not blurred by accident; it is exploited precisely because it can be blurred.

That is the great discovery of economic Trumpism. Extraction does not need to be completely hidden if it can be normalized as part of the spectacle. And that spectacle works because the target audience no longer demands much moral coherence from its leaders, only tribal alignment. If the product enriches the family but humiliates cultural enemies, many will consider it an acceptable price. And if it also offers the possibility of making quick money, even better.

The irony is that this model, despite presenting itself as anti-system, depends entirely on capturing the system’s rents. It does not replace it; it parasitizes it. It does not build an alternative to the old relationship-driven capitalism of Washington; it turns it into a cruder, more personalist and more performative version. In the end, perhaps the best way to understand all this is to stop thinking of the Trumps as political actors with side businesses, and start seeing them as a family conglomerate whose main competitive advantage is political power.

Trump continues with his Kabuki theater, announcing by tweet that everything is fine now and that peace is close. In his defense, markets seem to believe him: we have hit new all-time highs, with a rally that sits in the 100th historical percentile. Objectively, the situation is now much worse than it was a month ago, yet markets appear to be pricing in an immediate return to normality.

The reality is that nothing has returned to normal, despite Friday’s headlines, when Araghchi announced the opening of the Strait until the end of the ceasefire. The message was misinterpreted. What it actually reflected was that the agreed conditions were now being met, and that vessels would circulate under Iranian supervision, with no more than 12–15 ships per day. Since then, the Strait has closed again and traffic remains practically non-existent.

There are three friction points that continue to block any stable resolution: the U.S. refusal to accept a toll system in the Strait of Hormuz; the demand that Iran hand over or sell all of its near-weapons-grade enriched uranium; and Iran’s request to release $27 billion frozen abroad, in addition to compensation for damages. Of these, the real breaking point is Hormuz. Iran has already played its most important card. Once used, giving it up without securing a structural concession would be equivalent to admitting that it has lost its main tool of future deterrence. There may be a gradual reopening of the Strait, even a partial normalization that keeps part of production out of the market for some time, but a toll system is politically unacceptable for the GCC. Saudi Arabia and the UAE cannot accept Iran turning the world’s main energy chokepoint into a permanent source of rent extraction and coercion. Meanwhile, Iran still has some room to resist. It holds around 170 million barrels in floating storage or in transit, which at physical prices of $135 per barrel is worth almost $23 billion. This is not a structural solution, but it does give Iran some oxygen to withstand the initial impact of the blockade, provided that the blockade is truly effective.

The problem for the global market is that the numbers start scaling very quickly. If the U.S. blockade cuts Iranian exports, estimated at between 1.6 and 2 million barrels per day, the total loss of global supply would already reach the 12.6–13 million barrels per day range. And that is not the extreme scenario.

If the Houthis also disrupt Bab el-Mandeb, affecting Saudi exports from Yanbu, another 4.5–5 million barrels per day could be at risk. In that case, the total shock would reach 17–18 million barrels per day, a figure that cannot be offset with inventories, diplomatic rhetoric, or small production adjustments.

In short, if Hormuz does not resume normal operations, the market will only rebalance through the most painful channel: demand destruction via much higher prices. Even if the Strait reopened immediately, the production currently shut in — around 11 million barrels per day — would not return instantly and would be unlikely to normalize before the end of June, with the cumulative loss exceeding 800 million barrels.

In the short term, the path for energy investing is painful, but it is the only theme I see clearly from a value investing perspective this year.

U.S. economic growth in the fourth quarter was somewhat weaker than previously thought, but the underlying message is not as simple as the headline suggests. GDP was revised down to an annualized 0.5%, from the previous 0.7%, mainly due to a smaller contribution from business investment and inventories, while consumer spending also came in slightly below expectations. At first glance, this reinforces the slowdown narrative, especially compared with the 4.4% recorded in the prior quarter. However, it is worth avoiding an overreaction to this figure, because both the third and fourth quarters were distorted by one-off factors, including the government shutdown, and therefore do not cleanly reflect the true momentum of the economy.

In fact, if one looks at somewhat more useful measures, such as final private domestic demand, the picture still points to an economy that is cooling, yes, but not collapsing. This metric grew by 1.8%, down from 2.9% in the previous quarter, confirming some loss of internal momentum, though still far from a severe recession scenario. The most striking part of the report was elsewhere: corporate profits rebounded strongly, accelerating clearly from the previous quarter. In other words, aggregate activity is moderating, but corporate profitability is holding up much better than one might expect in a cooling macro environment, which also helps explain why equity markets have behaved the way they have. Overall, the data fit with an economy that is less exuberant, but still reasonably resilient. The problem is that the margin for error is narrowing. If this slowdown is combined with a more tense geopolitical environment and the possibility of new shocks to energy or confidence, the start of the year could become complicated very quickly. For now, rather than an imminent recession, what we have is an economy losing momentum while profits remain resilient, a combination that usually supports the market for a while, but can also become fragile if consumption and investment continue to deteriorate. On the other hand, inflation data again came in somewhat higher than expected, but once again they are using lagged data or even incorrect readings; using the same categories in real time, as measured by Truflation, the reality of PCE stands at 2.42%, a much more reasonable level and closer to the Fed’s long-term target.

The problem is that none of this is really valid anymore after the disruption in Hormuz, and the global recession that will likely come with it, along with the inflationary rebound driven by fertilizers, gas, and oil. The Fed is now in a very difficult position, one that makes further rate cuts in the short term practically unworkable.

Yesterday, peace talks between Iran and the United States began in Islamabad, Pakistan. They follow in the wake of Wednesday’s ceasefire, which the market has welcomed as though it materially reduced the risk, but it is probably more accurate to see it as a tactical pause rather than the beginning of any real resolution. The conditions that would make a durable agreement viable are, as things stand today, almost impossible to reconcile. Iran is not going to give up easily its ability to influence Hormuz, nor does it seem realistic to expect peaceful acceptance of its enrichment program, let alone a political framework in which meaningful sanctions are lifted, reparations are paid, and the entire regional front, including Hezbollah, is frozen. It is too much for too many actors with incompatible interests. That is why, even if crude softens for a few days on initial relief, it is hard to believe this will lead to a stable peace or a rapid normalization of the energy market. Even if some passage through Hormuz is allowed for two weeks, that does not mean flows return to normal. At best, what we are likely to see is a partial release of crude trapped in the Gulf, not a symmetrical restoration of traffic. The ships already inside will try to get out, but shipowners will have far less incentive to return while they still do not know what will happen once that temporary window closes. That asymmetry is crucial. The market may read the reopening as a return to order, when in reality it may be nothing more than a partial release of floating inventory. And if maritime trade recovers to half of its pre-conflict levels, that would already be a relatively optimistic outcome.

Nor does it seem reasonable to expect a rapid rebound in production. If a meaningful share of Gulf and Iranian output has been disrupted, bringing it back online requires more than a ceasefire on paper. First, the stranded tankers have to leave, then new vessels must come in to load, and on top of that, onshore storage, which is likely saturated by now, needs to be cleared. Even if the ceasefire holds, the recovery in shut-in volumes would not be immediate but gradual, and in the meantime global inventories would remain under pressure. In other words, the supply shock has already happened, and its effects do not disappear just because the tone of the headlines changes. If Iran emerges from this crisis with the effective ability to regulate or condition flows through the strait, the problem for Saudi Arabia and the UAE would not be the toll itself, but the implicit surrender of regional hegemony. Saudi power has never rested solely on its reserves or its wealth, but on its role as swing producer, on its ability to influence the marginal balance of the global oil market. If Tehran ends up with a hand on the roughly 20 mb/d that pass through Hormuz, that primacy is structurally eroded. And that is something the Gulf monarchies are unlikely to accept passively. Seen this way, the conflict is not only about missiles, sanctions, or nuclear diplomacy, but about control of the most important valve in the global energy system. So far, as a precondition for the talks, Iran has requested the release of sanctioned funds, and it already appears set to receive $6 billion, which makes defending the American handling of the conflict increasingly difficult. The situation is now much worse than it was when the attacks began.

Of the sell-offs on Wednesday, one of the least understandable was that of European natural gas, given that no LNG tanker passed through the Strait of Hormuz. And while Qatar says it is beginning to lay the groundwork for a partial restart of Ras Laffan, the crown jewel of global LNG, taking advantage of the temporary improvement in security conditions after the ceasefire between the U.S. and Iran, the reality is that these are repairs that will take months, if not years. For now, what exists is not a real normalization, but a very limited reopening focused on maintenance work and the mobilization of technical staff to bring some lines back online in the coming days. Even so, it is important not to get confused: restarting part of the facility is one thing, but restoring a meaningful flow of exports is something very different. For that, LNG carriers still need to be able to cross Hormuz with some degree of normality, and on that front the situation remains extremely fragile. The key point is that, even if Ras Laffan manages a partial restart, the structural damage has already been done. Last month’s attack knocked out roughly 17% of Qatar’s annual export capacity for what could be an extremely long period, up to five years, which completely alters the balance of the global gas market.

In other words, the bottleneck does not disappear just because some volumes come back. I understand the current weakness in the context of the end of the cold season, which softens the impact, but if the bottleneck is not resolved soon, once we enter the inventory refill phase during the summer, we will see new price spikes not seen since 2022. On the positive side, China has sharply reduced its LNG demand, and it will offset the gap through pipeline imports and greater use of coal and renewables. Beautiful, clean coal.

Although the paper oil market has fallen sharply this week, from around $115/b at Monday’s open to roughly $95/b by the end of the week, the picture is very different in the physical market, where actual barrels for real delivery are trading at $125/b to $140/b. In fact, perception is shifting as well, and Goldman now expects prices above $80/b throughout both 2026 and 2027.

What we are seeing is that the decline in exports, both crude and refined products, at a global level is already very evident, pushing flows well below the range normally expected at this time of year.

Not only that, but if we look at the market balance the EIA is now contemplating, the contrast between its February, March, and April forecasts is enormous. Even so, they continue to insist on a rapid return to a very large surplus from Q4 onward, which is already unthinkable if one considers the likely duration of the disruption in Hormuz. Beyond the logistical problem, a great deal of infrastructure has also been damaged. JPM counts more than 60 assets affected, representing a total of 2.4 mb/d of refining capacity, with the east-west pipeline in Saudi Arabia as the latest victim, which will translate into a production loss of around 700 kb/d.

My view, as you already know, is that the market still has not understood the magnitude of the problem. We have a similar precedent in the Yom Kippur War fifty years ago. The market took almost a year, and also needed an especially violent bear market, to fully digest the 1973 oil shock. The Arab embargo, triggered after the Yom Kippur War, removed only about 4% to 5% of global supply, yet that was enough to quadruple crude prices within a few months. What followed is well known: surging inflation, gasoline shortages, and a devastating equity market collapse between 1973 and 1974, with the S&P 500 and much of the major global indices falling by more than 40%. The difference now is that the potential shock would be significantly larger, not only in scale, but also because of how interconnected the system has become. We are not just talking about crude, but also natural gas, refined products, petrochemicals, fertilizers, and, of course, logistics. On top of that, today’s market does not function like the market of the 1970s. Information now moves in real time, quantitative models react within seconds, and investors are far more conditioned to read supply disruptions and bottlenecks. Precisely because of that, the adjustment could come faster.

The historical lesson is fairly clear. Oil shocks rarely remain contained within the energy market itself. They usually act as the catalyst for something broader, putting pressure on inflation, consumption, monetary policy, and valuations all at once. In 1973, the market took months to connect all the dots. This time, it will probably take less. And that very speed of recognition could make the reaction more violent, not less.

The war with Iran is entering a more delicate and, above all, more binary phase. Pete Hegseth made it clear that the coming days will be decisive, combining the usual offer of a “deal” with an explicit threat of escalation if Tehran does not yield. In other words, Washington is still trying to present the diplomatic path as open, but the real message is that it is negotiating from a position of strength and with very little interest in concealing it. On the other side, Iran responded by widening the perimeter of the conflict, not only on the military front but also on the economic one, by threatening major U.S. corporations with a presence in the region. That list, which includes companies ranging from Microsoft and Google to Boeing and Tesla, is aimed at raising the perceived cost of the war and reminding everyone that the conflict is no longer limited to bases, militias, or energy facilities.

The Strait of Hormuz remains the true center of gravity of the conflict, because it turns a regional war into a global energy shock, and the solutions now being discussed, which involve Iran establishing a permanent toll, would amount to a constant Sword of Damocles hanging over the energy market, which would therefore deserve a geopolitical premium. Trump’s speech on Wednesday did nothing to calm tensions, and it already seems quite clear that there is no real plan and that nothing will return to the way it was before.

That shock is already beginning to filter into the real economy. In the United States, gasoline has risen above $4 per gallon for the first time in more than three years, which creates a first-order political problem for Trump and the Republicans ahead of the midterms, which are already being written off here, with expectations of a blue wave both in those elections and in the 2028 presidential race. Some conflicts are fought on the military front, while others are fought in the voter’s wallet; this one threatens to do both at once. That is why the White House insists on projecting military resolve while still leaving the door open to a negotiated exit. It needs to project strength, but it also needs to avoid higher energy costs damaging consumption, confidence, and ultimately the economic cycle. Internationally, visible cracks are also beginning to emerge. France and Italy have expressed reservations about parts of the military operations, the Pope has openly called for a way out, and China and Pakistan are pressing for an immediate ceasefire. In other words, the longer the war drags on, the harder it will be to maintain a coherent political coalition around Washington and Tel Aviv. The problem with long wars is not just their material cost, but that they gradually erode the legitimacy of those who sustain them. And that now seems to be starting to happen. In that scenario, energy once again moves to the center of the board. And, as so often happens, when geopolitics comes in through the door, market complacency goes out the window.

Quality investing has come into fashion on the back of a fantastic decade, but it contains far more problems than might initially be assumed. Nike remains stuck in a restructuring that is moving more slowly than the market had expected. Although third-quarter figures came in somewhat better than feared — with revenue flat at $11.28 billion and EPS of $0.35, beating estimates — guidance for the fourth quarter was clearly disappointing, with sales expected to decline by 2% to 4%, versus the slight growth the consensus had been pricing in. The underlying problem has not changed. China continues to be the main drag, not only because of weak demand and competitive pressure from local brands such as Anta and Li Ning, but also because Nike is still digesting operational mistakes and excess inventory; in fact, the company itself is forecasting a 20% collapse in sales in the region next quarter. The recent improvement there, after falling “only” 10% compared with the previous -16%, is real, but still far from sufficient.

The other major bottleneck is inventory. Nike is still relying on promotions to clear older product, especially in Europe and in its digital channel, which inevitably erodes margins. It is therefore no surprise that gross margin has contracted again for the sixth consecutive quarter, falling 130 basis points to 40.2%, with the impact of tariffs adding further pressure. The problem is that the market is beginning to wonder how much longer this cleanup can last, especially when Nike has spent several quarters promising progress without fully normalizing the business. As noted at the beginning, this is a world-leading brand, with products at the technological frontier, yet whose stock market performance has been, at the very least, disappointing. Price is always the most important thing, and drifting away from that discipline has damaging consequences.

Pam Bondi’s dismissal is a fairly revealing signal of how this White House understands loyalty and the political usefulness of institutions. Bondi is not being removed because she was somehow outside the Trump orbit. Quite the opposite: she had been one of his most reliable allies, both during the 2020 impeachment and in defending the subsequent electoral narrative. Precisely for that reason, her departure carries greater political weight, because it shows that in this environment not even loyalty guarantees survival if it ceases to be functional. The immediate trigger appears to have been the handling of the Epstein case, which has become a growing source of noise within the Republican bloc itself. The delay in releasing documents, the errors in the redactions, and the accidental exposure of victims’ names have given ammunition both to the opposition and to sectors of Trump’s own base. The subpoena from the oversight committee led by Republicans was already a clear sign that Bondi was starting to become a liability.

But limiting the interpretation to the Epstein case would fall short. Bondi was also under pressure for not having carried out aggressively enough the other major function Trump expects from his justice apparatus: going after political enemies. That is the real substance of the matter. In Trump’s universe, as in any banana republic, the Department of Justice is not merely a state institution, but a tool of power. It is not enough to want to fight the battle; it has to be done effectively. The choice of Todd Blanche as interim replacement fits perfectly with that logic. A man with a direct personal link to Trump and a far more combative profile when dealing with judges, career prosecutors, and internal checks. It is a move that will likely intensify the politicization of the DoJ rather than correct it. The market or institutional reading here is clear: Trump’s second term continues to consolidate an increasingly personalist model of power, in which job security depends less on technical performance than on the ability to execute the president’s political agenda without creating excessive collateral costs.

The saga and hype around SpaceX’s potential IPO continues to build, and it would be far more than just another major tech listing. It could become the defining financial event of the cycle. If a valuation above $1.75 trillion and a capital raise of more than $50 billion are ultimately confirmed, we would be talking not only about the largest IPO in history, but also about the arrival in public markets of one of the most coveted private assets in the world. In any case, the key is not so much the rockets as Starlink — which is precisely why this matters so much for ASTS. The epic narrative of colonizing Mars, returning to the Moon, or putting data centers into orbit helps sustain the story, but the part that truly makes a valuation of this magnitude defensible is the recurring satellite connectivity business. That is where the market may begin to see a very unusual and difficult-to-replicate combination: critical infrastructure with growth, defense contracts, geopolitical exposure, and a proprietary network already deployed on a global scale. The “Musk dream” adds to the appeal, but the potential cash generation is explained above all by Starlink.

That said, the deal would also open a very interesting Pandora’s box. A publicly listed SpaceX would force the market to price what could be called the “Muskonomy” far more rigorously — that increasingly intertwined ecosystem made up of Tesla, xAI, X, SpaceX, and the rest of his assets. So far, much of that value has moved in private markets, with less transparency and far more room for narrative permissiveness. Going public would imply much greater scrutiny of corporate governance, capital allocation, control structures, and above all the extent to which a single individual can continue simultaneously running several enormous companies without real frictions emerging. That is probably the main perception risk. SpaceX may be an extraordinary asset, but it would also come to market carrying an enormous narrative premium. And when a company goes public wrapped in an almost mythical story, the margin for disappointment narrows. It would not be surprising to see a valuation heavily dependent on the market’s faith in Musk, more akin to Tesla than to a traditional aerospace company. That can work strongly in its favor at the debut, but it also implies considerable potential volatility once the market starts separating fantasy, optionality, and tangible profits.

We have passed the point of no return in the global oil market. After a month of war, with disruptions affecting roughly 20 Mb/d transiting through the Strait, the loss of inventories is already enormous and will have long-term consequences for the crude market.

The market, including what is often referred to as smart money, still seems oblivious to this structural shift and continues to operate with a buy-the-dip mentality and first-order thinking, going short oil while waiting for the longed-for TACO headline.

The Trump administration is reportedly considering a long-term solution based on an Iranian toll on cargoes transiting the Strait. On paper, it sounds workable, since it would only raise cargo costs by around 1%, but in practice it is unviable because it would alter the balance of power among the Gulf countries and leave the rest under a constant threat. More specifically, it is highly unlikely that either Saudi Arabia or the UAE would accept such an arrangement, as it would leave them hostage to Persian leverage. As a result, it is likely that they will begin using their substantial influence in Washington to block such a resolution. Past this point of no return, the only sector that still looks clearly attractive is energy, and the broader market, both the indices and the rest of the sectors, is likely to wake up to a harsh reality in the coming weeks.

The potential IPO of SpaceX is starting to look like something more than just another big tech listing. If it is confirmed that Musk wants to reserve as much as 30% of the offering for retail investors, it would represent a fairly clear break from the usual Wall Street ritual, where most of the allocation typically ends up in institutional hands and the retail tranche plays little more than a decorative role. At least on paper, the idea makes sense, especially given the valuation being sought and Musk’s strong appeal among retail investors. Broadening the shareholder base to include small investors and family offices could help build a more stable and less opportunistic ownership structure, avoiding part of the classic cycle of initial euphoria followed by rapid distribution from funds simply looking to monetize the first-day pop.

SpaceX is already trading at valuations that belong to a different market regime, following the integration of xAI in an all-stock transaction that valued the combined entity at around $1.25 trillion. If it does go public in 2026 at a valuation close to or above $1.5 trillion, it would not only enter straight into the ranks of the world’s largest companies, but also serve as a stress test for the market’s real appetite for grand narratives around growth, AI, and hard tech. After months in which valuations across the AI complex have started to raise more questions than excitement, a deal like this would amount to a kind of referendum on how much capital is still willing to pay extraordinary multiples for “iconic” assets. It also fits Musk’s style quite well.

That said, it is worth keeping an eye on the element of exuberance. When an IPO starts being sold as a historic event before the price has even been set, we are usually no longer talking only about fundamentals, but also about market psychology. SpaceX is an extraordinary asset, probably a unique one, but precisely for that reason it may become the best possible thermometer for measuring how willing the market still is to reward stories built on scale, ambition, and almost unlimited optionality. If the deal comes out strong, it will reinforce the idea that the technological narrative bubble is still alive. If it stumbles, it could mark a much more important turning point than it seems. Elon to the rescue in moments of geopolitical turbulence.

Cracks keep appearing in the private credit market. Ares has become the latest major alternative asset manager to hit the liquidity brakes in private credit, capping quarterly redemptions in its ASIF fund at 5% after receiving withdrawal requests equivalent to 11.6% of the fund’s capital. The signal confirms that the pressure is no longer anecdotal or confined to a single vehicle. Apollo and BlackRock have made similar moves, while others such as Blackstone have chosen to absorb more paper in order to avoid sending an overly aggressive message to the market. In practice, the sector is entering its first real stress test after years of almost frictionless growth and abundant liquidity-driven exuberance.

The underlying problem is fairly obvious. These vehicles promise periodic liquidity on assets that, by nature, do not have it. As long as money keeps coming in, the structure holds; when outflows accelerate, the contradiction becomes visible. This does not necessarily imply an immediate deterioration in the underlying credit — in fact, Ares insists it has no non-accrual loans and still maintains available liquidity — but it does reveal that the confidence of the marginal investor is beginning to erode. And in products like these, perception matters almost as much as portfolio quality. The comparison with the stress episodes in non-traded REITs a few years ago is not exaggerated: once the market starts to question valuations and liquidity structure, redemptions tend to feed on themselves. What is most interesting is that the sector is still trying to frame this episode as a simple operational normalization. Prorating redemptions to 5% is built into the product design and helps avoid forced sales or an excessive cash drain. From the manager’s point of view, it is the right decision. But from the investor’s point of view, it is also an uncomfortable reminder of what they actually own: not a money market substitute with enhanced yield, but exposure to illiquid private credit wrapped in a promise of limited liquidity. For years, many buyers seemed to forget that second part. Private credit had become one of Wall Street’s favorite corners just as doubts are beginning to pile up around underwriting standards, spreads, refinancing capacity, and macro sensitivity. If fundraising continues to cool — something BDC fundraising data is already beginning to suggest — the industry could shift from a regime of almost automatic expansion to a much more selective one, where the gap between strong platforms and more fragile vehicles becomes far more visible.

The market reacted with relief to yet another Trump pivot in the Middle East, but it is giving him less and less credibility and increasingly understands that the United States has backed itself into a dead end. The president said that the U.S. would postpone for five days an attack on Iranian energy infrastructure and also claimed that there were “very productive” talks underway with Tehran aimed at reaching a full resolution to the conflict, although Iran was quick to deny it. That headline alone was enough to trigger a violent unwind of the fear that had built up over the previous sessions. Crude collapsed by more than 10%, the dollar lost momentum, sovereign yields fell, and equities rebounded sharply from oversold levels. As a side note, and a sign of the level of grift in the current administration, there was unusual selling in crude futures and buying in the indices just before the announcement.

As soon as the possibility of diplomatic de-escalation appeared, however fragile, we saw the typical reaction of a market that had become excessively loaded with defensive hedges. Oil and the dollar fell, bonds eased, and flows returned to equities, especially in more cycle-sensitive segments such as small caps. The next day, however, it already seemed clear that there was no real solution in sight, and everything “returned to normal,” with equity indices falling again and crude moving higher, despite the administration’s efforts to make it seem as though everything was going smoothly. Demand for protection, through puts and hedges, is rising quickly and is already back at December 2021 levels, even though the dominant narrative still insists that everything is fine and will soon be resolved.

Some investment banks, such as Citi, are already arguing that this situation reminds them of COVID, which is something we have also been saying for weeks, when the market took two months to finally break after the Wuhan lockdowns. It is important to understand the past in order to position ourselves in the present.

One of the consequences of this conflict will be a renewed sense of urgency around energy autonomy and security, given how fragile the logistical dependence on fossil fuels really is. Nuclear, renewables, and coal are all likely to benefit. Japan and the United States continue to deepen their energy alliance, and this time the focus is on one of the major structural narratives of the moment: the need for new power capacity to sustain the AI boom. The White House announced that the joint venture between GE Vernova and Hitachi will invest $40 billion to develop small modular nuclear reactors in Tennessee and Alabama, as part of a broader $73 billion energy agreement between the two countries. In addition, Tokyo will also channel capital into gas plants in Pennsylvania and Texas. In this way, the SMR story is no longer just an aspirational concept and is beginning to move into the phase of large-scale industrial deployment. These types of reactors promise shorter construction times and more flexible implementation than traditional nuclear power, something especially valuable in a context where electricity demand is rising sharply and the grid needs stable, dispatchable, and less intermittent solutions than pure renewables.

At the same time, and somewhat paradoxically, plans are now being drawn up for the construction of a new coal-fired power plant. The news is striking because it symbolizes precisely the opposite of what the market had taken for granted for years. Terra Energy Center will allocate $1 billion to a new coal project in Alaska, which would mark the first significant investment in new coal-fired power generation in the United States in more than a decade. The Trump administration is trying to reopen space for coal on the argument that electricity demand is set to surge because of the rise of data centers and artificial intelligence. In economic terms, coal continues to lose ground to gas and renewables, and its share of the U.S. power mix has been declining steadily for years, although under Secretary Wright, 15 of the 16 plants that had been expected to close in 2026 have seen their operating lives extended. The key nuance is that Trump’s energy narrative is no longer limited to “drill more” or to defending oil and gas, but now extends to trying to slow the closure of coal plants and even promote new construction. That fits with a vision of energy abundance at any cost, where the priority is not to optimize the energy mix but to secure firm megawatts for growing demand. In every major commodity bull market, coal has led the way, and this time I do not expect it to be any different.

The key point here is that the market has moved in a very short period of time from a problem of excess supply to a potential scenario of physical tightness that is far more serious than many are pricing in. With production shut-ins already on track to approach 12 million barrels per day by the middle of next week, that excess could be absorbed in as little as 56 days, including the volume removed from the market since March 5. And the most relevant point is that, once that cushion is gone, the system’s ability to respond appears to be extremely limited. Global production cannot come close to replacing those 12 million barrels per day at anything like the same pace, and at the margin we would be talking about only an additional 0.1 to 0.2 million barrels per day over the next three to four months. On top of that, the situation is becoming even more complicated because non-OPEC supply appears to have already peaked, which further reduces the system’s elasticity. In the case of U.S. shale, which remains the main variable many still look to as a potential source of rapid adjustment, real-time data point to stagnation, and the expectation is that it will remain flat. So if demand holds up reasonably well and production shut-ins become entrenched, the market could move from talking about elevated inventories to facing a genuine supply constraint on a much shorter timeline than it seems.

In fact, and this is precisely why energy companies are interesting both in the current scenario and in a post-conflict one, the entire oil futures curve has shifted aggressively higher, and companies could now hedge their production above $75/b all the way out to 2030 if they wanted to, something that would have been unthinkable just a few months ago. Without a real and clear solution in sight, and even if a minimum agreement is reached to buy time, the only truth is reality, and when politics and physics collide, physics always wins.

It is an early but relevant signal of how an energy disruption that appears geographically distant can be transmitted very quickly into the real economy. In Western Australia, diesel shortages are already becoming a tangible operational constraint for fuel-intensive sectors, especially for small and mid-sized players that lack both the bargaining power and the supply priority enjoyed by larger companies. The case of Blue Cap Mining, a contractor dependent on independent distributors, with high daily fuel consumption and very little room to absorb disruptions without impacting production, is especially illustrative. The most important aspect of this episode is not so much the exact number of workers affected, but what it reveals about the structure of the sector. When a company starts parking equipment and sending workers home not because demand for its product has fallen, nor because the asset has ceased to be profitable, but because it cannot secure the fuel needed to operate normally, it shows how the effects can cascade quickly and irreversibly. In a remote mining region, diesel is not just another input. It is the minimum condition for everything else to function.

In addition, situations like this usually expose something that remains more hidden under normal conditions: the enormous fragility of smaller players in the face of any shock in the physical supply chain. Large operators are better able to cushion a supply crisis because they have stronger contracts, greater storage capacity, and implicit priority within the system. Smaller players, by contrast, are left at the back of the queue, much like first-world countries versus those in a weaker position, which are already being forced to adopt drastic rationing measures. It is also worth underlining that this affects a particularly vulnerable sector. Australian mining, like much of the natural resources complex, remains diesel-intensive in a way that many energy transition narratives tend to understate. That dependence does not disappear simply because the dominant narrative revolves around electrification, batteries, or decarbonization. In practice, a large share of raw material extraction still relies on heavy machinery, remote transport, and operating chains in which diesel remains irreplaceable in the short term. That is why a prolonged shortage does not just pressure margins; it can translate directly into lower activity, lower production, and greater stress throughout the supply chain.

When energy supply comes under strain, the economy does not stop all at once, but it begins to deteriorate through the margins, first in the weakest links. The market tends to focus on the big names, but very often the first cracks appear much earlier among smaller players. That is usually where the physical reality of the problem begins to show.

The Fed has chosen not to make a move, as was already widely expected, but the underlying message was far less reassuring than a simple “rates unchanged” would suggest. It kept rates at 3.50%-3.75%, raised its inflation forecasts (diesel has already surged beyond its 2022 peak), sees no further deterioration in employment, and continues to point to just one rate cut this year. On paper, it looks like continuity. Here we can see the evolution of the PPI in its latest readings.

In practice, though, it is an uncomfortable pause, with a more hawkish bias than the market had expected. The explicit acknowledgment that the Fed is navigating blind in the face of a new energy shock was surprising to me. Powell repeatedly insisted that nobody yet knows what the real economic impact of the war between the U.S., Israel, and Iran will be (it will be a recession). And when a central bank repeats that uncertainty is “unusually high,” what it is really saying is that it has no room to commit to a clear path of monetary easing. The priority stops being support for growth and becomes avoiding a loss of control over inflation expectations. The market had been anchored to the idea that disinflation was still on track and that any macro weakness would eventually force the Fed to cut. But the rebound in crude makes that narrative much harder to sustain, and although Powell tried to separate the energy shock from the underlying trend, he also admitted that tariffs are still making inflation stickier and that progress toward 2% remains frustratingly slow. In other words, even if oil does not end up generating a severe second round of inflationary effects (which it will), it is arriving at the worst possible time, when inflation has not really been defeated yet.

The new projections reflect this shift quite well, with year-end inflation now forecast at 2.7%, unemployment held at 4.4%, and growth barely improving. In my view, once it becomes clear that this conflict is dragging on and that the damage to employment and the economy is serious, that is when the market will react violently, recognizing that there will be no rate cuts (until the proximity of the midterms forces them), with the biggest impact falling on the technology companies most sensitive to this narrative.

Meta may be preparing a major workforce reduction, potentially exceeding 20% of the company, in what would be its largest round of layoffs since the “year of efficiency” in 2022–2023. On the one hand, the company needs to free up resources to sustain what will be a massive escalation in AI infrastructure CapEx over the next few years. On the other, Zuckerberg appears increasingly convinced that the productivity gains driven by AI itself will allow the company to operate with much smaller teams. Meta, like the rest of the hyperscalers, is betting aggressively on the AI race, offering outsized compensation packages to attract talent and planning massive investment in data centers, while at the same time trying to slim down those areas where it believes human labor will become progressively replaceable or, at the very least, more efficient. It is the same pattern that is beginning to repeat across much of the tech sector: fewer people in traditional structures and more spending on compute, elite talent, and infrastructure.

What matters here is that this decision is not coming from a position of absolute strength, but also from a degree of competitive pressure. Meta has stumbled on several fronts with its Llama models, has struggled to meet performance expectations, and needs to show the market that its enormous AI investments are not simply a way of kicking the can down the road. When a company accelerates spending in such a capital-intensive area while simultaneously proposing mass layoffs, the implicit message is that it is trying to preserve margins while buying time for that bet to start delivering tangible returns. The promise of efficiency sounds good on paper, but it is worth keeping sight of the other side of the story. Many of these companies are committing colossal sums to technological infrastructure with returns that are still uncertain. Cutting headcount may soften the short-term impact, but it does not answer the bigger underlying question: whether all this investment will ultimately generate returns commensurate with the capital employed. And if not, it is worth remembering their great metaverse adventure, which they have already decided to wind down. And rightly so.

The Iranian attack on Ras Laffan represents far more than a temporary disruption, and it is the worst possible event for Europe. We are talking about roughly 17% of Qatar’s LNG export capacity being taken offline, with 12.8 million tonnes per year out of service for between three and five years, in addition to the further impact on condensates, LPG, helium, naphtha, and sulphur. In practice, this forces QatarEnergy to declare force majeure on long-term contracts with Europe and Asia and brings back into focus a reality that the market often underestimates during periods of apparent calm: critical energy infrastructure remains one of the most sensitive bottlenecks in the global system. It does not take a massive destruction of global supply to tighten prices and flows; it is enough to damage highly concentrated assets that are difficult to replace and require long repair periods. Until now, Qatar had been perceived as one of the most stable and secure hubs within the global energy complex, and that safety premium has just been broken. If the conflict becomes normalized and oil and gas facilities start being treated as legitimate targets within the regional escalation, the market will have to reprice not only Qatari gas, but the entire security of Gulf supply. In addition, the temporary halt to the North Field expansion introduces another important dimension, because it is not just current supply that falls, but also visibility on future growth that deteriorates. In a market that was already relying on new volumes to supply Europe and Asia over the next decade — which makes Europe’s self-inflicted mistake of cutting off Russian gas look even more painful now — this tightens the balance even further. Energy security is not negotiable, and instead of learning that lesson and trying to come out stronger, Europe is once again betting on the same exhausted formulas that only make the problem worse: taxes and price controls. Good luck.

Twenty-one days after the start of the conflict, the crude oil market has moved beyond the initial cushioning phase and is entering a far more uncomfortable dynamic. The floating oil buffer has already been exhausted, which means that each additional day of disruption now implies a decline of around 11 million barrels per day in crude-on-water, pressure that is now being transmitted directly to the onshore storage system. And there too, the margin is extremely thin: excluding China, onshore inventories only have a buffer of about 45 million barrels, enough to last barely four more days. The release of strategic reserves will ease some of the strain, but it will not change the broader picture. If the SPR starts flowing by the end of next week, it will do so at a pace far below the current gap, around 2.5 million barrels per day versus an 11 million barrel imbalance. Under these conditions, the drawdown in onshore inventories could reach close to 290 million barrels in April, a figure large enough to turn a logistical problem into a physical supply problem.

On the supply side, the deterioration continues to deepen. Shut-in production is already approaching 10 million barrels per day, with Iraq practically out of the market and Kuwait moving in the same direction. The UAE and Saudi Arabia may still be able to sustain production for a few more days, but everything points to them having to cut soon as well. Our estimate is that by the middle of next week, disrupted production could be approaching 12 million barrels per day. The key point here is that the market is no longer trading on expectations, but on increasingly tight physical balances. While floating oil was still available, the system could absorb part of the shock without its fragility becoming obvious. That margin has now disappeared. What comes next is a direct transfer of stress from the sea to onshore inventories, and then, unless there is a rapid normalization, into production and consumption. In commodities, small imbalances can be managed; imbalances of this magnitude cannot. In practice, this translates into a high probability of recession unless the situation is resolved immediately, and one could even begin to talk about a depression if the conflict drags on beyond the month of April.

As the saying goes, better late than never. Von der Leyen’s speech marks a fairly significant shift in tone in Europe. Not so much because it reveals anything new about nuclear energy, but because it politically normalizes an idea that for years was almost uncomfortable across much of the continent: that giving up a stable, relatively cheap, and low-emission energy source was, in hindsight, a strategic mistake. In a context of energy vulnerability and geopolitical tension, the message is that Europe’s energy autonomy will not be built on renewables alone, but on a more pragmatic mix of nuclear, renewables, grids, storage, and flexibility. In fact, this has been the major driver behind the erosion of European industrial competitiveness and one of the reasons Germany has gone through three years of zero growth. The announcement of a €200 million guarantee for innovative nuclear technologies and the future European strategy for SMRs points precisely in that direction. The figure alone will not transform the sector, but it matters as a political and financial signal. Markets often do not need so much initial capital as they need credible institutional validation that lowers perceived risk and encourages more investors to step in. If Europe wants to play a real role in the next generation of nuclear power, the first step was to stop treating it as an anomaly and start viewing it as a strategic technology.

The most important part of the message is probably not the funding, but the attempt to simplify regulation, harmonize frameworks across countries, and speed up permitting. In nuclear, the bottleneck usually lies in execution, bureaucracy, and lack of scale. And that is exactly what SMRs need, what Von der Leyen is putting on the table: a coordinated European market capable of turning an industrial promise into a real value chain. At its core, what we are seeing is a shift from a more ideological vision of the energy transition toward a more industrial and geopolitical one. Europe is beginning to accept that decarbonization is not simply about installing more renewable capacity, but about building an energy system that is robust, competitive, and less dependent on the outside world. The question is no longer whether nuclear fits into that equation, but how much time Europe has lost pretending that it did not.

The crisis in the Middle East has suddenly complicated the room for maneuver of Asian central banks. Until recently, much of the debate revolved around when and how much to cut rates in order to support growth. Now the problem is that an oil-driven supply shock pushes energy prices higher, fuels inflation, and at the same time weakens economies that were already growing with difficulty. It is the worst kind of inflation because it does not come with economic strength, but with fragility. In emerging Asia, the dilemma is even more uncomfortable. It is not just that more expensive crude makes rate cuts harder; it is also that an overly dovish stance in a strong-dollar environment can punish currencies and accelerate capital outflows. In other words, the issue is no longer measured only in future inflation, but also in exchange-rate stability. India is a good example: the central bank would probably prefer to prioritize growth, but if the rupee comes under sustained pressure, currency intervention could become the real short-term monetary policy tool.

In developed Asia, the problem is just as serious. Japan and South Korea are especially vulnerable because they depend on imported energy, stable global supply chains, and a reasonably orderly external environment. In Japan’s case, the situation is particularly uncomfortable because the Bank of Japan can no longer ignore inflation as easily as it did in the past. It has less room to look the other way, but also very limited space to tighten aggressively without damaging an economy that is structurally weak. At the heart of the issue is the fact that the market is once again facing a scenario it thought had largely been left behind: the risk of stagflation. Not as an unavoidable base case, but as a sufficiently real threat to shape central banks, currencies, and expectations if the conflict drags on long enough. When oil rises, not all shocks are the same. Some cool demand and help central banks. This one does exactly the opposite: it forces them to choose between protecting growth or protecting credibility, knowing they will probably not be able to fully protect both at the same time.

Despite the violent swing in the oil market, the most likely scenario is that this is not the end of the move, but merely the beginning of its real consequences. Recent history offers an uncomfortable lesson: in 2020, not even the historic OPEC+ intervention managed to stop the initial collapse in oil prices. This time, the logic works in reverse. If tanker flows do not normalize soon, the physical, balance, and inventory problem will become impossible to solve. Every day that passes without traffic resuming in the area means more crude trapped on land, more inventories piling up in Gulf countries, and less operational room to sustain production, which has already begun to decline. Iraq and Kuwait would already be nearing their storage limits, the UAE would be next, and Saudi Arabia has announced a 2Mb/d reduction. In fact, these shut-ins would already have begun, and estimates point to a cut of several million barrels per day that could expand rapidly if the disruption lasts just a few more days.

In oil, the arithmetic is fairly clear. If the strait returns to normal operation almost immediately, the damage would be manageable and probably consistent with Brent trading in a still elevated, but not out-of-control, range of around $75/bbl. If the disruption lasts another one or two weeks, the market stops pricing in a temporary scare and starts pricing in a material loss of supply, with Brent moving into the high $80s or even $90 after the situation is resolved. Every day that passes without flows means inventories fall by around 11Mb, and production restrictions have already reached 4Mb/d, with expectations that they could rise to 6.5Mb/d–7Mb/d by the end of next week. That leads to the following calculations:

• If everything is resolved TODAY, 210Mb would be lost.

• If everything is resolved by March 17, 290Mb would be lost.

• If it is resolved on March 22, 370Mb would be lost, implying oil at around $90/bbl for the rest of 2026.

• If it is delayed until the end of April, Brent would be close to $100 for the rest of the year.

The situation is extremely serious, and it closely resembles February 2020, when most of the world still seemed unaware of what was coming. Eyes wide shut. That alone, combined with problems in fertilizers and other commodities, would be enough to cause a significant recession, and we are already starting to see price spikes not seen since 2022, which will make the Fed’s job even harder. Trump already knows he has lost the election and will probably face impeachment afterward... and a cornered animal is dangerous.

To try to counter the effects of this crude shortage, the G7 has agreed to release 400Mb from its strategic oil reserves. Here, the role of the SPR should be kept in proper perspective. Yes, it can relieve some of the pressure and, above all, buy time while the market tries to navigate a disruption like the one in the Strait of Hormuz. But it is one thing to soften the initial impact and something very different to replace a physical flow of this magnitude. These are two completely different orders of problem. On paper, the maximum release rates look high, but in practice effective capacity is far more limited. The experience of 2022 already showed that the actual pace of drawdowns falls short of the technical maximums. And that matters, because the current disruption is running at around 11 million barrels per day of crude, while production shut-ins are already estimated at between 4 and 7 million barrels per day, with pressure building as the days go by. In other words, even in the most favorable scenario, the SPR does not fill the gap; it simply cushions it for a very short period. Below, we can see the distribution of the G7’s strategic reserves (1,100Mb).

In addition, the United States enters this episode with far less room than the headlines suggest. After the announcement of a 172 million barrel release, reserves would fall to around 243 million barrels, the lowest level in decades and barely a third of total capacity. If we also exclude the volume that must be maintained for operational reasons, the truly usable buffer shrinks dramatically. In practical terms, the American strategic buffer is starting to look more like a temporary stabilization mechanism than a tool capable of supporting the market through a prolonged disruption. That is why the broader message does not change. The SPR can buy you a few weeks, probably until the end of March, but not much more. Beyond that, if flows do not normalize, the market is once again confronted with the same underlying problem, only now with fewer reserves available and with a physical structure that is becoming increasingly strained.

Anthropic is an uncomfortable reminder that in AI, the technological “moat” is worth less than the market thinks once you collide with the State. State, Treasury, and HHS have ordered staff to stop using Claude by direct directive from Trump, joining the Pentagon, and redirecting teams toward alternatives like ChatGPT and Gemini, with fewer ethical constraints. This isn’t a vendor switch driven by performance—it’s a de facto political sanction: a way of turning a domestic player into an administrative pariah for not accepting the usage terms Washington wanted to impose, especially around military guardrails, autonomous weapons, and potential domestic surveillance. The administration is making it clear who runs the deployment chain when a technology is considered critical to national security and the country is in a global race against other powers like China. You can have the best model, but if you don’t accept the government’s operating framework, they flip the switch off—starting with the most visible and symbolic demand: federal usage. This also creates a second-order effect: if you’re blacklisted in core agencies, many vendors, integrators, and internal teams become allergic to reputational and compliance risk, and the ban becomes self-reinforcing even if there isn’t a broad “legal prohibition.”

For OpenAI and Google, this is close to an ideal scenario. They don’t just capture share—they also cement the standard. In fact, the text notes that OpenAI has already announced an agreement to deploy technology on the DoD’s classified network, and that Altman rushed to add a written clarification that it will not be intentionally used for domestic surveillance of U.S. citizens or nationals. In AI, the risk isn’t only competing against another model—it’s competing inside a framework where the dominant buyer can rewrite conditions, define acceptable uses, and pick winners by decree. The market tends to underweight this kind of risk until it materializes. It’s a shame, because Anthropic’s superior model has helped it gain enterprise market share exponentially, riding on Claude Code and specific plugins—but that may not be enough for total success.

The conversation around AI is still being framed as a stock-picking game, and that is why the essential point is being missed. If we really are entering a world where production is automated first on the cognitive side and then on the physical side, the natural outcome is not just higher productivity, but a structural decline in marginal costs and, with it, strong deflationary pressure. And deflation does not just lower prices—it erodes revenue pools in real terms and turns “advantage” into a battle to retain some share of the corporate surplus. The fact that a company can produce something more cheaply does not mean it will capture that margin for years, because everyone else is also lowering costs, and competitive equilibrium forces prices down. The second-order effect is that if labor is displaced, the personal income tax base shrinks, and the State will need another source of funding to sustain the social contract, probably through some form of UBI. That pushes political pressure toward capital and, in practice, toward the few winners that concentrate cash flow. You could end up with a toxic combination of falling prices, compressed margins, and rising effective tax rates. The transition itself may be highly profitable because the market takes time to price in these dynamics, but the end state does not look like infinite profits—it looks like an environment where part of the surplus is competed away and part of it is taxed. It is a plausible scenario that would destroy and reshape the structure of the equity market as we know it.

The operation in Iran has become far more complicated, and unlike in Venezuela, it is not clear what the objective is or how it is supposed to be achieved. Operation Epic Fury began with several objectives:

• Destroy and eliminate the regime’s missile infrastructure.

• Destroy the Iranian navy.

• Bring about regime change and the end of the nuclear program.

But after a week since it began, the path to achieving those goals no longer seems clear at all. Iran, while it does appear to be something of a paper tiger — that is, its ability to inflict major damage (in terms of human casualties, infrastructure, security, and so on) on its enemies is limited — has chosen to pursue a strategy of chaos and broader Gulf involvement, which could drag the conflict out significantly (think back to the Red Sea and the Houthis...) and create strong economic pressure to bring it to an end. The first thing to blow up has been the price of oil, which is already flirting with $100/bbl, something that seemed unthinkable just a few weeks ago, especially if we take into account that physical market fundamentals clearly pointed to a surplus. Right now, crude is expensive relative to fair value, but cheap relative to the tail risk if logistics seriously break down. Two key factors are coming together:

• Iran is one of the world’s largest producers, and any disruption to its operations could alter the market balance in a very significant way.

• The world’s main maritime artery for hydrocarbon exports, the Strait of Hormuz, has its narrowest passage right next to the country’s border, meaning that the threat of attacks or a blockade of traffic could also cause major disruptions to trade flows.

Iran, despite sanctions, had been producing practically at full capacity (around 3.55 mb/d), so the optionality is only to the downside for prices. If one strips out the noise, WTI should be closer to $60/bbl in a base-case scenario. But we are already close to $90. That gap is, quite literally, the price of uncertainty. And here the usual problem in commodities appears: when the risk premium compresses or normalizes, it does so quickly and without warning. That is why the return/risk profile gets worse and worse as WTI approaches $100. Producers, for their part, are hedging aggressively. After more than a decade of a bear market in oil, now that prices have finally rebounded, nobody in the sector is willing to stay exposed to the market taking that margin away from them overnight. But the interesting thing is that nobody is increasing capex. According to the latest U.S. production estimates, if WTI stays around $70/barrel, production could rise to roughly 14 million barrels per day. If WTI averages $80/barrel, it could reach a peak near 14.5 million b/d, but that would be close to the ceiling. That implies relatively limited additional growth, between 350,000 and 850,000 b/d compared with December 2025 volumes. It is not an especially aggressive expansion.

As for maritime traffic, even if there were no attacks — and in fact, a physical closure of the strait is highly unlikely — security risk alone is enough to deter navigation. If insurers cancel “war risk” policies (or refuse to renew them) for vessels transiting the Gulf and the Strait of Hormuz, the strait does not actually need to be “closed” for flows to slow. It is enough for the cost and availability of coverage to make transport operationally unviable for a meaningful portion of the fleet. In practice, that means fewer ships willing to load, more waiting days, more delays, and a rapid rise in freight rates and insurance premiums. The oil may exist, but it may not get where it needs to go on time.

In the Iran disruption, everyone instinctively looks to oil, but the more interesting shock could end up coming through gas, and not through Europe and TTF, but through Henry Hub via LNG. It is important to understand that the Strait of Hormuz is a trade route of critical importance for a large number of commodities, and all of them are at risk of experiencing spikes like the one seen in oil.

So far, the consensus on U.S. gas had been clearly negative, with the heating season ending, inventories closing near the 5-year average, and the sense that 2026 is “well supplied.” On top of that, there is the whole set of arguments that has kept HH contained: fairly well-known structural LNG/Mexico demand with no major surprises expected until 2027, Lower 48 production still capable of growing, and a 2026 with more renewables that could once again reduce power burn if gas gets too expensive in the summer, in addition to the usual gas-to-coal switching when prices become too high. In parallel, the market had already been buying into the 2027–2030 narrative of a possible global LNG glut from new capacity, especially from Qatar, which in theory would limit growth in U.S. exports and put pressure on prices. Last weekend changed the variable that was not in the model: supply reliability. Qatar declaring force majeure on its production is a practical demonstration that LNG from the Middle East carries an interruption risk that buyers can tolerate when everything is normal, but not when there is sustained geopolitical tension. If the market moves into a regime where LNG shipping rates surge and the probability of disruption is no longer remote, the premium is not paid only in the spot price; it is paid in the kind of contract you want to sign and with whom. That is where the U.S. becomes the preferred supplier, not because it is the cheapest today, but because it is the most financeable and “insurable” in terms of country risk and continuity of flow.

On top of that, the U.S. starts with obvious advantages: it is already the largest exporter, shale producers still enjoy solid economics, and the Permian, because of its associated gas dynamics, can end up selling molecules at absurdly low prices or even negative prices during periods of congestion. That last point matters, because the market tends to think of HH as a standalone story, when in reality supply elasticity is constrained by infrastructure and by the relationship with oil. If the world is asking for more secure LNG, the bottleneck stops being just the resource itself and becomes capacity, contracts, logistics, and permits. The point is not that the U.S. will run short of gas tomorrow. The point is that the contract map may have shifted for years, and that is exactly the kind of change the market is usually slow to price in.