Market Comment - May

Equity markets went completely haywire, which can happen, but once again, we did not expect it to happen in 2026. Both the presidential election cycle, as well as the decennial cycle did not have such kind of behavior in the cards. Markets enjoyed two tailwinds: First, the US government is selling crude oil from its Strategic Petroleum Reserves like never before. The US spent 40+ years building the SPR to 700 million barrels as a buffer against exactly this kind of supply shock. As of today, more than half of this reserve is gone. Currently, that is not a problem. But, to suppress the price of oil, the government has to continue selling reserves and we will see where we end up. To put things into perspective, every month more than eight times paper oil trades in the market compared to the physical oil that exists. As a conclusion, we witness massive market manipulation by the U.S. government and it would not come as a surprise, if Scott Bessent were leading it. The second tailwind is the return of the A.I. insanity. Both the Mag7 and the semiconductor industry are getting bought with full hands. While the buying in semiconductor stocks may make sense, the Hyperscalers, who are expanding their capex like never before, should trade lower, not higher. It is impossible for them to monetize on the irrational capex spending in the short run, even ever. But currently, nobody seems to care and the flow into the technology sector continues to be strong. The question for investors is not whether this market behavior will end badly, but when. As previously mentioned, both the election cycle and the decennial cycle have never seen such kind of behavior. Both seasonal trends are currently completely out of order. What is striking in the current A.I. narrative is the similarity it has with the infamous Enron story of 2000. By December 2001, Enron had filed for what was then the largest corporate bankruptcy in American history. A year earlier, the company was the seventh largest corporation in the U.S. While Enron applied bogus income calculations for its earnings reports, A.I. companies like Anthropic and OpenAI work with the premise to build increasingly larger models, which become better, while becoming bigger, or simply because they increase in size. The premise is that this will improve future earnings. It is a bogus calculation as well. It could be that those earnings expectations may never happen. Because there is a trend emerging, within the industry, called “open models” which achieve 90% of the performance of closed models when they are released, and the price of running inference is 87% less. The cash flow models currently presented by closed, proprietary AI inference models, namely those from OpenAI, Anthropic and Google, could turn out to be completely wrong. Both OpenAI and Anthropic are planning IPO’s in 2026. Better sooner than later if the mentioned trend to OpenRouters persists. But still, currently we see the same run for semiconductors like we saw in 1999 for servers. A Cisco server in 2000 was sold for USD 250`000, in 2003 for USD 25`000. Market breadth in the U.S. during the month of May was weak. Bullish Percent Indices, which have already started to fall in April, continue to fall in May, while the market moved up. This condition has led to a first cluster of Hindenburg Omen, which can be a precursor for a market correction. While new 52-week highs are expanding, so do 52-week new lows. Another surprising event is the flattening of the US 2-Year yield curve. It dropped almost 30% or 12 bps. If it continues to do so, it will become a tailwind for the US Dollar. The month of June will provide the first FED meeting with its new chairman Kevin Warsh. May he bring some clarity to the markets? In conclusion, we are looking at an unhealthy market, but still, there is no sign of a top. Avoiding any election and decennial cycle studies, the S&P500 should hold up well until September, while semiconductors should see weakness from mid-July, and Europe should trade weaker than the US in the coming months. Currently, markets are living by the crystal ball. Let us see for how long.