What moves hyperscaler stocks?

This article is an on-site version of our Unhedged newsletter. Premium subscribers can sign up to get the newsletter delivered every weekday. Standard subscribers can upgrade to Premium here, or explore all FT newsletters

Good morning. The first estimate for real US GDP growth in the third quarter came in at a ripsnorting 3.4 per cent. The big difference was not AI data centre spending — all those semiconductors get netted out because they are imported. Instead, there was sustained growth in household consumption and a big rebuild in inventories. There will be revisions, but Unhedged’s view that US economic growth remains broad-based is holding up. Email us your interpretation of the numbers: unhedged@ft.com.

Tech’s tough week

What moves hyperscaler stocks? This earnings season, the answer seems simple enough: cash flow, or the absence of it. Meta, Alphabet and Amazon all committed the sin of surprising the market with higher than expected investment spending — in slightly different ways. Meta threatened to spend more in the future, paid for with new equity issuance, while it was the actual spending at Alphabet and Amazon that gave fright. All three stocks have been hit hard.

Microsoft, on the other hand, surprised the market with parsimony, and generated $19bn in free cash in the quarter, and its shares have held up.

Here’s a look at free cash generation across the group:

Microsoft is not completely without sin, however: the morning after its reassuring earnings report it announced that it would issue $25bn in bonds to fund its AI build-out, and its shares lost the gains it had enjoyed in late trading the evening before.

Oracle, meanwhile, did the opposite. Its cash burn is among the fastest of the hyperscalers, and its corporate credit rating just a hair above junk. But on Wednesday, it said it expected to return to positive free cash flow sooner than expected, and pledged to limit additional funding to $20bn until the 2030 fiscal year. Its shares and its bonds rallied.

Cash doesn’t quite tell the whole story. Part of the reason Alphabet lost a cool half-trillion in value yesterday was that it compounded the sin of heavy spending with results that suggested the AI revolution might hurt its core business. Search advertising revenue was a little lower than hoped, and management intimated that “AI mode” inquiries were harder to sell ads against than old-style searches. The response was horror, despite the fact that the ad business continues to perform quite well. I’ve added Meta’s ad growth to this chart of revenue growth for context:

Eleven per cent growth in ad revenues is about the average growth rate for the past several years, and Alphabet’s cyclical growth pattern mirrors Meta’s. But when no one knows anything about what the future holds, we all jump at shadows.

It didn’t matter to investors that Google Cloud, where Alphabet’s own AI business sits, grew more than 100 per cent in the quarter, crushing its rivals at Microsoft and Amazon. This makes some sense: Alphabet’s ad revenue is running along at about $330bn annually; Cloud at just $125bn.

So, at the present moment, the pattern is something like this: cash flow is investors’ top priority; hints of bad news from legacy businesses are secondary but important; growth of AI initiatives comes last. Fear before greed, for now.

ECB conclave

Christine Lagarde has announced that she is leaving the European Central Bank before the end of her eight-year term. This didn’t move the market — ECB watchers had suspected she would peel off before her time was up, but the announcement livened up an otherwise straightforward rates decision (a hold) and press conference.

The timing obviously has something to do with the political ructions in France. She will step down after the monetary policy decision in February, two months before the first round of the French presidential election. At a minimum, it is about allowing President Emmanuel Macron a role in deciding the next slate of top European jobs. (The two main candidates to succeed him, the far-right and far-left leaders Marine Le Pen and Jean-Luc Mélenchon, are Eurosceptics.) But there is also the possibility that Lagarde wants to play a role in French politics herself, as she seemed to hint yesterday: “I don’t think I can remain neutral,” she said. She wasn’t talking about R*.

Unhedged has avoided speculation about her successor because it’s not clear that it means all that much for markets yet. ECB top jobs go to members of a small pool of familiar names, most of whom have been institutionalised by decades of Eurozone policymaking.

That said, we do enjoy it when sell-side analysts put their quant-y skills to speculative ends. Earlier this year, BNP Paribas’s Josha van Spronsen, Duncan Forsyth and João Vieira Pinto repurposed a University of Bocconi model, which in 2025 correctly predicted Robert Prevost as the next Pope, to model Lagarde’s successor. “The allocation of top European posts,” they write, “shares features with the papal succession, functioning as a set of bargaining games where candidates must show sufficient prestige, credibility and consensus-building ability to progress.”

The BNP Paribas study compared candidates’ professional backgrounds with those of previous ECB presidents and factored in the political horse-trading involved in European appointments. (For senior EU jobs, national politics is at least as important as individual merit.) The authors said that by layering these dimensions, they hoped to answer the question of who is “papabile”, or most suitable, for the requirements of the ECB presidency and other top positions.

Who comes out on top? Klaas Knot, former president of the Dutch central bank, and Olli Rehn, governor of the Bank of Finland. For a study completed in March, it has proven prescient. Knot is clearly in the running. But he’s facing close competition from Pablo Hernández de Cos, head of the Bank for International Settlements, who the BNP Paribas model suggested more closely matches the profile for the vice-president role.

The authors also backtested the analysis on past ECB leaders. It correctly positioned Mario Draghi, Jean-Claude Trichet and Wim Duisenberg in the ‘president’ sphere and got several former vice-presidents, chief economists and executive board members right. But the model unfortunately cannot account for the risk of a truly left-field candidate snagging the job. Lagarde was so far off the scale last time around that she didn’t fit in any of the model’s spheres.

One good read

What killed Weil Gotshal?

FT Unhedged podcast

Can’t get enough of Unhedged? Keep following Rob Armstrong and Katie Martin’s incisive and unfiltered analysis on the Unhedged podcast, a twice-weekly, 15-minute lively dive into the latest markets news and financial headlines. New episodes every Tuesday and Thursday.

Recommended newsletters for you

Due Diligence — Top stories from the world of corporate finance. Sign up

The AI Shift — John Burn-Murdoch and Sarah O’Connor dive into how AI is transforming the world of work. Sign up