US crude briefly traded below $90pb yesterday on the news, and Brent below $97. Considering that the price was flirting with the $115pb level just two days ago, seeing levels below the psychologically important $100pb mark probably got investors ahead of themselves. Yields dropped, pricing in the end of inflationary risks, the US dollar fell across the board — the dollar index is now back to levels seen before the Iran war started — and equity markets rallied.

Major tech-heavy indices hit fresh record highs, also bolstered by strong earnings from chip companies, cementing the AI growth story. Samsung joined the trillionaires’ club after announcing a nearly 50-fold profit rise for its chip unit alone — that basically makes up 94% of its total earnings.

AMD rose more than 18%, rewarded for its own record earnings and strong guidance. The latter didn’t prevent rival Nvidia from adding nearly 6% itself! The Nasdaq 100 gained more than 2%. Gold rallied nearly 3%, taking advantage of a sharp drop in the US dollar and global yields, which compress the opportunity cost of holding the non-interest-bearing metal.

Now, all this is great, but there is no certainty it will last. We had peace proposals and market rallies over the course of the past weeks, and all of them ended in disappointment.

Given the chaotic diplomacy, and given how the US lost control of the situation — and given the earlier unfounded announcements of progress — I would like to say: ‘I will clap when Iran confirms.’

Because the longer this drags on, the greater the risk of oil shortages and sharper spikes in oil prices. And a 180-degree turn in the situation is just one headline away. Just one headline.

The thing is, you can’t enter a winning trade on the news, you must enter a great trade in expectation of the news. And this is what’s leading to this maybe-premature rush and massive volatility.

This morning, oil prices are higher than yesterday’s dip levels; the risks remain two-sided.

Topix and Hang Seng are up as they catch up with the past days’ holiday lag, while Kospi posts small gains following a 48% rally since the beginning of April. Remember, most of it is thanks to rising memory chip demand, which created shortages and pushed memory chip prices notably higher — we’re sometimes talking about triple-digit price increases.

Memory chip makers are used to boom-and-bust cycles. During boom cycles, they increase capacity, produce more and end up with too many chips, which then pull prices lower and weigh on margins.

Today, many are betting that AI demand will break that cycle and keep the memory chip sector in an eternal boom cycle. Is it possible? On one hand, computing needs double every 6–7 months — and this could accelerate as adoption accelerates and models grow — meaning that we will always need more chips. On the other hand, such demand – and margins! - would attract other players and eat into margins.

In fact, today, the parabolic rise in Samsung’s PE ratio warns that the stock price is rising faster than earnings, despite the 50-fold rise in chip profits. Rationally, there should be a correction. But sometimes it takes time before investors come back to their senses.

What’s interesting is that the AI trade has changed hands. It’s no longer Nvidia alone, but a cocktail of AI model providers — like OpenAI and Anthropic — a bunch of different chip makers — GPUs, TPUs, CPUs, memory chips to keep the system together — and construction equipment makers like Caterpillar

On the flip side of this trade are software companies shaken by headlines, like the one that dropped this week about Anthropic’s new agents for performing financial services tasks. LSEG was down another 2.5% in London yesterday, which I think is an opportunity to buy the dip when you consider that AI may replace their platforms, but the data is still theirs to sell.

Anyway, with all these changes, the earnings season is surely going well. Besides the fact that more than 80% of companies that have already reported earnings beat expectations, Deutsche Bank analysts called it ‘one of the best earnings seasons in 20 years’. The Middle East war impact was nowhere to be found, and the trade war hit — not much either.

But the ADP report yesterday printed a lower-than-expected figure, suggesting that the US economy added 109K new private jobs last month versus 118K expected, while inflation metrics around the globe are heating up. Yet optimism is such that investors prefer to see the glass half full.

Yesterday, they cheered the fact that job additions were notably up from the 61K added a month earlier. And you know, a softer-than-expected figure also justified lower US yields. Or simply, no one cares about jobs data anymore. War headlines and chip earnings are far more exciting than the good old jobs figures.

Whatever it is, the market mood is probably too optimistic to last. The higher we go, the faster we’ll fall if anything goes wrong. So the best thing to do is remain diversified, and conscious that we will more likely than not face high volatility in the coming days.