Huang's exit reshapes the US-China summit; inflation bites from Paris to Mumbai

Nvidia's board left itself exposed to a textbook case of key person risk. Huang held no supervoting shares, just 3.6% of the company, so the board could have removed him at any point. What it could not replace easily was the web of relationships and strategic instincts he had built over three decades. When a single person is that central to a company's direction, their sudden exit creates genuine uncertainty about whether the strategy continues, and markets hate uncertainty more than they hate bad news.

The US-China angle is where this stops being a corporate governance story and becomes a foreign-policy one. Export controls on advanced chips are already the sharpest economic weapon in the US-China rivalry. The allegation that Huang had a relationship with someone connected to Chinese intelligence gives Congress a concrete reason to demand stricter enforcement, or new restrictions entirely, right before a summit where chip access was already on the agenda. If controls tighten, Nvidia loses revenue; if they loosen as part of a deal, the scandal makes that politically toxic. The summit's outcome just got harder to predict.

The RBI's inflation target works as an anchor, a public promise about where prices will be kept that shapes what businesses and households expect prices to do next. When inflation stays inside the 2-to-6% band, people plan accordingly. When it breaks through the ceiling, those expectations can drift upward, and once people expect higher prices, they ask for higher wages, which pushes costs up further. That feedback loop is exactly what the RBI is trying to prevent by signalling a rate rise now, before expectations become unmoored.

Food prices are doing most of the damage, and that is where the evaluation gets complicated. A rate rise works by making credit more expensive, which slows spending and investment. But food prices are rising because of a weak monsoon and high energy costs, not because people are spending too freely. Raising rates does not make it rain. The RBI is being asked to use a demand-side tool against a supply-side problem, and the risk is that it slows growth without fixing the underlying cause. One month of data above the ceiling is not proof that inflation is entrenched, but the RBI cannot afford to wait and see if it turns out to be.

A bond's yield, the annual return a buyer earns, moves in the opposite direction to its price. When investors sell gilts, prices fall and yields rise automatically. The question is why they are selling. The most likely answer is that investors across the world are repricing how long interest rates will stay high, partly because of the US inflation data earlier this week and partly because UK inflation has proved stickier than the Bank of England hoped. When the world's biggest bond market, the US Treasury market, sells off, it pulls other sovereign bond markets with it, because global investors compare returns across countries.

The precedent that fits here is the autumn of 2022, when the Truss government's unfunded tax cuts triggered a gilt crisis that forced the Bank of England to intervene as an emergency buyer. The parallel is instructive but does not hold all the way. This sell-off is driven by global rate expectations, not by a single government's fiscal shock, so the scale is smaller and the Bank is not being asked to rescue the market. The honest question is whether the Budget's spending plans remain credible if yields stay elevated for another quarter, because the government's debt interest payments rise with every basis point.

The shuttle between Moscow and Kyiv is a classic example of what diplomats call mediation, where a third party carries proposals between two sides who will not sit in the same room. The US is not a neutral mediator here: it supplies weapons to Ukraine and has sanctioned Russia. That gives Washington real power over both parties, since it can threaten to reduce arms flows to Ukraine or to ease sanctions on Russia depending on which side it wants to pressure. The actor's dilemma for the Trump administration is that any deal it brokers will be judged by whether it looks like a Ukrainian surrender.

The Abu Dhabi venue is worth noting. The UAE has positioned itself throughout this conflict as a country that talks to everyone, hosting Russian oligarchs, Ukrainian officials and Western diplomats without choosing sides. That makes it a useful neutral ground, but it also means the talks carry no enforcement mechanism. The gap between signing a ceasefire and actually stopping the fighting has been the graveyard of every previous attempt, from Minsk in 2015 onward. What is different this time is that the US is directly in the room, which raises the cost of walking away, but does not guarantee that either side has changed what it fundamentally wants.

France's inflation number is a problem for the ECB because the bank cannot set different rates for different countries. This is the core tension in a currency union, where a group of countries share one currency and one central bank but have separate governments, separate labour markets and separate inflation rates. When Germany's inflation is falling and France's is rising, the ECB has to make a single call that is wrong for at least one of them. Cutting rates to help Germany risks letting French inflation run; holding rates to contain France risks tipping Germany into a deeper slowdown.

The France cost-of-living thread has been running since November, and what I find telling is that the October jump comes despite the ECB having already raised rates significantly over the past two years. That suggests the inflation in France has a supply-side component, energy costs and food prices, that rate rises cannot easily fix. The ECB's rate tool works by reducing demand, making it more expensive to borrow and spend. If prices are rising because of costs rather than excess spending, the tool is blunt. The risk is that the ECB tightens enough to slow growth without actually bringing inflation back to its 2% target.