The latter triggered a stiff selloff across AI enablers at the start of the week on fear that the slower the development, the slower the investment, the slower the data-centre buildout, the slower the chip and other raw-material sales to builders, and the slower the revenue growth.

Well, Nvidia’s Jensen Huang spent the week trying to convince investors – and the world – that that’s not true. From taking a phone call from the US President on stage to announcing that Nvidia’s sales will DOUBLE next year, he did what he could to prop up battered appetite. And he managed to pull his company’s shares more than 2.50% higher yesterday.

On the sales outlook front, to give some context, the company CFO had said last month that revenue would increase by 70% – and that number would be 100% if there were no supply constraints. Either supply constraints evaporated, or Nvidia – which has been systematically beating its own forecasts since the beginning of 2023 – decided to give itself a smaller cushion to save the day.

One or the other, the week has seen appetite recover for semiconductors. The news that SK Hynix is in talks with Intel about potentially producing memory chips in the US probably helped ease some of the AI-pocalypse worries. Philadelphia’s Semiconductor Index jumped more than 3%, while the Korean Kospi is up around 2.50% at the time of talking.

Broadly, chipmakers and AI enablers have been giving back gains from their summer peaks. If we leave the apocalyptic scenarios aside for a minute, the need for massive capital investment, evaporating free cash flow, the increasing need for financing through stock and bond sales at a time when yields are rising, and interest rates from major central banks are projected to climb higher remain major headaches that could limit upside potential.

So far, the strength in earnings growth has been sufficient to counterweigh the ugly macroeconomic and geopolitical backdrop. The question is how long growth companies can continue to outperform the rest. Since the beginning of the Iran war, we have seen the technology sector attract capital. If high energy prices (hence borrowing costs) remain the primary concern, we could see capital flow into energy and broader value names.

Easing oil prices give relief

The good news is that oil prices may have topped for the moment. US crude is pulling back for the third session as concerns over Middle Eastern supply constraints ease, with WTI trading around $101 per barrel.

I am afraid, however, that if we don’t get a sustainable decline in oil prices, refined-product prices will remain elevated and continue to keep inflation expectations at uncomfortable levels. Gasoil and diesel exports from both the Middle East and Russia have been squeezed by the wars. With no end in sight, supply constraints will become a growing headache, especially as global oil inventories and strategic reserves have been falling. In other words, the clock is ticking louder.

The good news is that JP Morgan updated its outlook for global oil inventories. Back in April, its “no-resolution” scenario – which assumed that flows through the Strait of Hormuz would remain close to zero – projected that global inventories would fall to the operational stress level of around 7.6bn barrels by May, before approaching the operational floor of around 6.8bn barrels shortly afterwards, near September.

In reality, inventories have been falling much more slowly than feared and remain around 7.9bn barrels, comfortably above the operational stress level. That buys the world some precious time to find a solution, but probably not enough to reroute and restore energy supplies to pre-war levels if the wars drag on. So risks to oil prices remain tilted to the upside. It is almost impossible to make precise forecasts here – ultimately, it is about geopolitics. Oil companies therefore remain, in my view, a better hedge against actual energy-supply risks than volatile spot energy prices.

Keeping up with the Central Banks

The Bank of England (BoE) kept its policy rate unchanged in yesterday’s decision, pointing to inflation risks. It also echoed that inflation depends on energy prices which, in turn, depend on the geopolitical situation – impossible to predict. But “the longer this volatility persists, the bigger the impact it will have on inflation, and the more likely it is we will need to raise the Bank Rate”, Governor Bailey said.

But that wasn’t the highlight of yesterday’s meeting. What really made the difference was the BoE’s overhaul of its QT programme. The Bank will retain £120bn of its longest-dated gilts to back banknotes, allow £222bn of shorter-dated gilts to mature naturally and sell the remaining £146bn at an annual pace of £20bn, potentially through sales to the government. APF gilt auctions have been paused in the meantime. Bond investors celebrated the reduced pressure on the long end by buying longer-maturity gilts, pulling 10- and 30-year gilt yields sharply lower.

Better yet, global yields eased along with oil prices yesterday, and I believe that the Federal Reserve’s (Fed) decision to hike rates – defying the White House – to defend its inflation objective also helped pour some cold water on the heated sovereign bond markets.

And finally, today, the Bank of Japan (BoJ) also delivered a highly expected – and wanted – 25bp hike, taking its policy rate from 1% to 1.25%. This was the second rate hike in three months – well faster than the previous pace of roughly one hike every six months. But the knee-jerk reaction was a swift rise in USDJPY, as two officials dissented.

The market’s answer was clear: one hike is not enough; more is needed to bring the BoJ rate somewhere around neutral. So what BoJ Governor Ueda says matters more than the rate hike itself. How Ueda sees inflation risks evolving and how he sees fiscal concerns fitting into this context will be important for the USDJPY’s next direction.

As I wrote yesterday, Japan may need to raise rates towards the 1.5–2% range to move deeper into what could be considered a neutral zone. The BoJ’s own estimates put the nominal neutral rate in a much wider 1.1–2.5% range. That means at least one and up to three additional 25bp rate hikes to reach the 1.5–2% range in the coming quarters. If traders believe that the BoJ can’t deliver that, it will be hard to keep the USDJPY from bouncing back towards, and potentially above, the 160 level.