Market Daily Brief: The AI bill lands, and the oil windfall keeps flowing
By Arthur Ledger
Friday brought a sharp lesson in what happens when a market pays up for a story. Alphabet lost about $500bn of market value in a single day, Intel fell more than 20% and Amazon slid before the open. Meanwhile BP, Exxon and Chevron reported profits that would have looked fanciful two years ago. The patient owner asks the same question of both groups: what will this business earn in an ordinary year, and what am I paying for that today?
What the FT covered
Alphabet fell 13% on Thursday after a quarter of negative cash flow, capital spending of $51bn and plans for a large share sale. Intel guided fourth-quarter revenue to $15-16bn against $17bn expected. Amazon fell about 6% pre-market on softer sales and cloud growth, and chip shares fell across Asia. US third-quarter GDP grew 3.4% annualised.
In energy, BP's adjusted profit reached $8.5bn on refining margins near $47 a barrel. Exxon earned $15.4bn and Chevron $9.9bn, and both warned that diesel will stay short through 2027. Tesco is selling its Czech and Slovak arms to Ahold Delhaize for £1.5bn, with most of the money going to buybacks. NatWest's quarterly pre-tax profit was flat at £2.2bn, the weakest growth among the big UK banks. Elsewhere, Virgin Media O2 is weighing up BT's broadband network as its debt trades well below par, Rio Tinto proposed a large rise in its chief executive's pay, and the opinion page weighed Anthropic's planned listing.
What our checks found
Tesco: 14/20, the best of the day. A simple, well-run business with strong returns and a steady cash yield of roughly 4-6% at about 15 times earnings. The disposal looks broadly neutral for earnings per share. Around 430p it would edge to 15.
NatWest: 11/20. A return on equity above 21% and about 7.7 times earnings, but the shares fell 7% today, and we cost the bank surcharge rise at about £0.25bn a year, by our rough estimate. Fair, not a bargain.
Exxon 10/20, Chevron 9/20, BP 9/20. All three look cheap on this quarter, but this quarter is the peak of a war-driven refining cycle. On our mid-cycle estimates Exxon trades at about 24 times and Chevron about 25 times. BP is cheaper, but still carries more debt than its peers, has no buyback, and took $4bn of impairments that we treated as one-offs. We stripped inventory gains and asset-sale gains throughout.
Amazon: 10/20. A great franchise, but free cash flow over the past year is about minus $42bn and long-term debt has more than doubled to fund capital spending. We stripped $68bn of investment gains, mostly on Anthropic. At about 29 times operating earnings, there is no margin of safety.
Intel: 4/20. Even after the fall it is about 69 times annualised guidance, with the foundry losing close to $2bn a quarter. Too hard.
Mention only: Rio Tinto (pay), Liberty Global and Telefónica through Virgin Media O2 (a debt story), and Anthropic and Nscale (unlisted) were not scored. Binance was left out because it is a crypto company.
Also on our desk (since Thursday's article)
- Fortinet 13/20 (X sweep): a founder-run, net-cash business with high returns, but even after a 20% fall it is about 38 times earnings. Nearer $90 would interest us.
- Estée Lauder 9/20 (X sweep): great brands, but margins are a fraction of their old level and the price already assumes the recovery.
- Accenture, Markel and Fairfax came up on X with insider buying, record buybacks and a closed sale. Nothing changes our earlier scores.
Nothing new cleared 15 today. When the market marks a great company down 13% in a day, it is tempting to call it a bargain. A bargain needs a price below what the business will reliably earn, and that is not yet on offer.
Research only, not investment advice.
Chronicle