Priced as bondholders
The S&P 500 closed Monday within a fraction of its record, +2.0% since 11 September
The big US banks report into a correction, and the order of their falls points at the bond market, not at borrowers. Interest income will not settle it; the paper losses might.
The big US banks are being priced as owners of bonds, not as lenders. A steeper curve is supposed to be their best season, and the lenders whose profits come mainly from lending have held up best. The heaviest falls are where value depends on bond prices or on bond buyers: the large securities books and the capital-markets houses. So today's results will probably show healthy interest income and settle nothing. What matters is the paper loss on the securities the banks already hold, and whether deals that need bond buyers are still clearing; tomorrow's inflation print decides which way both move.
Step back and this is the second time in four years that a rise in long yields has turned bank shares into a bond trade. Lending margins improve slowly, loan by loan, as old assets roll into new ones at higher rates; the loss on fixed-rate securities bought when yields were lower is marked at once. In 2023 that timing mismatch, not bad loans, decided which banks fell, and the market has not forgotten it. What is different now is that the rise comes from the long end rather than from the Fed, which steepens the curve and helps margins at the same time as it hurts the bond portfolios, so the two effects compete inside the same income statement. Add a capital-markets business that needs bond buyers to take the AI and buyout financing it underwrites, and the gap between the houses and the lenders is the gap between a mark and a margin. **Falsification.** This read is wrong if the banks reporting today show paper losses on their securities shrinking since midyear and their shares still fall, which would point to loan quality after all; it is also wrong if the bond-heavy names outperform the lenders on results while long yields hold above 5.2%. Conviction: high that the ranking of the falls tracks bond exposure; medium that results will not settle it; low on how far it runs before the inflation print.
The Executive Note
**Priced as bondholders.** JPMorgan, Goldman Sachs, Citi and Wells Fargo report this morning into a sector that has fallen while the S&P 500 sits near its record. The order of the falls is the story. Since 11 September the lenders, Wells Fargo, JPMorgan and Citi, have lost 5% to 8%. Goldman Sachs, Bank of America, Morgan Stanley and Schwab have lost 9% to 13%.
That ranking does not fit a credit scare, which would hit the lenders first, and it does not fit a squeeze on lending margins, because the curve has steepened. It fits the bond market. A ten-year yield above 5.2% lowers the value of the fixed-rate securities banks bought when yields were lower, and it tests the appetite of the buyers who take the AI and buyout debt the capital-markets houses arrange.
The engine's analyst panel has not moved. Across the seven names it recorded no new Sell rating in three months, and its favourite is among the worst performers. Forecasters are ranking the banks by earnings power; the market is ranking them by what they own and whom they need.
So the result to read today is not interest income, which will probably be strong. It is the paper loss on the bond portfolio, and what Goldman says about its financing pipeline. Bank of America and Morgan Stanley follow on Wednesday, Schwab on Thursday.
Wednesday also brings September consumer prices, consensus 3.7% with core at 2.5%. Meanwhile the Gulf war reached Saudi Arabia's capital again, and Brent closed at $106.65.
*Methodology.* Equity and commodity prices are Monday 12 October settlements; month changes run from the Friday 11 September close. The US Treasury cash market was closed on Monday, so yields are Friday 9 October marks. The analyst panel counts Buy, Hold and Sell ratings per company from our data provider; it does not track price targets or estimates, and its totals may differ from public trackers.
What mattered
The lenders fell least; the bond holders fell most
Since the close on 11 September: Goldman Sachs -13.1%, Bank of America -12.9%, Morgan Stanley -11.4% and Schwab -9.5%; Wells Fargo -7.7%, JPMorgan -6.2% and Citi -5.0%. The regional bank fund is down 11.6% from its August high and Bloomberg reports the big banks near correction territory going into results. Over the same month the gap between three-month bills and the ten-year widened from 107 to 119 basis points. CNBC notes that Goldman and Morgan Stanley carry earnings momentum into this week, and Citi is pushing to clear its regulatory burdens.
Two readings can be ruled out. A credit scare would hit the lenders and the loan funds first; instead the lenders fell least, leveraged loans sit 1.0% below their high and the high-yield bond fund 3.3%. A margin squeeze would not fit a curve that steepened. What the four worst share is exposure to bond prices, through large fixed-rate securities books, or to bond buyers, through underwriting and financing. Yesterday we wrote that the AI financing question would be answered through the lenders; their own shares say the question is wider than AI.
The read —Interest income will likely beat. The line to read is the paper loss on securities, and whether Goldman's financing revenue says bond buyers are still taking new deals.
The war reached Saudi Arabia's front door
The Houthis struck Riyadh's airport again, killing one person and injuring 11; the FT reports oil company chiefs staying away from the kingdom's energy congress, and Delta has postponed its Saudi service to March. Brent rose to $106.65, up 6.4% in three sessions, and the energy sector fund closed at a three-month high. Asian refiners have been buying spot November Saudi crude. In Ukraine, Russia kept hitting energy targets after Washington's ceasefire.
Saudi Arabia has been the supplier that fills gaps left elsewhere in the Gulf. A war that targets its capital changes what the oil price is insuring: not one lane of shipping, but the country the market relies on to replace lost barrels. The airport attacks also go after the kingdom's plan to become a finance and events hub, which raises the cost of the Saudi investment story beyond oil.
The read —Watch whether attacks move from airports to export terminals. Until they do, crude is pricing the possibility, not a loss of supply.
Tomorrow's inflation print is mostly fuel
Consensus for September consumer prices is 3.7% against 3.4%, with core at 2.5%, a gap of 1.2 points that is largely energy. MarketWatch frames the question as whether the data let the Fed delay a rate hike; BlackRock's research arm has turned positive on short-term Treasuries and cut equities, while Goldman has turned more positive on stocks. University of Michigan sentiment fell to 46.3 on Friday against 47.6 expected.
On 8 October we wrote that a weaker sentiment reading with fuel costs rising would add to the pressure for relief at the pump; it fell, and the administration's diesel deal is that relief. For the banks the print is the swing factor: a fuel-driven headline with core in line leaves long yields where they are, while a core surprise deepens the paper losses and brings a hike into view.
The read —Read core, not headline. A core figure at or below consensus would be the first relief for bond holders in a month.
What we see that the tape doesn't
The engine's analyst-consensus panel on the seven largest US bank and brokerage names: JPMorgan, Goldman Sachs, Morgan Stanley, Bank of America, Citi, Wells Fargo and Schwab, 360 ratings in all. Between 13 July and 12 October it recorded no change at all for 3 of the seven, and the changes it did record leaned positive: Sell ratings went from 16 to 15 and Buy ratings rose by 2.
Set that against the prices. Over the past month every one of the seven fell while the S&P 500 rose, and the panel's only change in that month was one more Buy, at Wells Fargo. Its favourite, Bank of America, with 35 Buys of 54 ratings, the highest share of the group, is among the worst performers, 15.8% below its August high. The objection is that ratings are sticky and a quiet panel proves little. But the forecasts and the prices are not merely out of step; they rank the banks differently. Analysts rank them by earnings power, which favours the bank with the most deposits to reprice. The market ranks them by exposure to bond prices and bond buyers, which puts that same bank near the bottom. A panel turning cautious would show it first in a downgrade at the margin; in three months no rating moved down. The disagreement says prices are carrying a variable that earnings forecasts do not: the mark on bonds already held, and the appetite of the buyers who take the deals the houses arrange. Neither appears in a rating, both appear in a share price. Limitations: the panel counts ratings, not price targets or estimates, which can move without a rating changing; our counts come from one data provider and may differ from public trackers. The next test is whether any of the seven is cut after results.
What to watch
- The paper loss on bond holdings that JPMorgan, Wells Fargo and Citi report today, against the midyear figure.
- Debt-underwriting revenue at the capital-markets houses: is AI and buyout financing clearing or waiting?
- Fed speakers Waller, Barkin and Collins today, for any change in how they describe a hike.
- Saudi spot crude sales to Asian refiners as the attacks on Riyadh continue.
Risks on the radar
Houthi strikes move from Saudi airports to export terminals
medium · severeRiyadh's airport was hit again with one dead, several countries now warn against travel to the kingdom, and Turkey, Pakistan and Egypt have joined Saudi Arabia in condemning the attacks. Iran's foreign minister has urged diplomacy in a call with Riyadh. The scenario is a strike on Saudi export infrastructure, which would remove the supplier the market relies on to replace other Gulf barrels. Carried 12 October at medium probability, severe impact and rising; still rising.
The Fed turns from holding to hiking
medium · highThe policy debate has moved to whether the Fed can delay a hike, and a former governor says this week's Fed commentary matters most. Three-month bills yield 4.05%. The scenario is a hike signalled before year-end, which would lift short rates, flatten the curve and take away the margin benefit banks are counting on. Last carried 2 October at low probability, medium impact and falling; raised to medium probability and high impact.
The energy truce collapses formally and refinery strikes resume
high · highRussia kept striking Ukraine's power plants and gas infrastructure after Washington declared its energy ceasefire; Zelenskyy accuses Moscow of breaching it, the EU says it does not believe in it, and Washington is again threatening Ukraine's arms supply. The scenario is a formal collapse with Ukraine resuming refinery strikes. Carried 12 October at medium probability, high impact and rising; raised to high probability as the breaches continued.
Listed private-credit funds slide toward redemption limits
medium · highThe listed business-development-company fund is 10.6% below its August high while leveraged loans and high-yield bonds are near theirs. A failed AI data-centre listing in Australia has unsettled Asia's new-issue market. The scenario is a run of redemption requests at non-listed private-credit vehicles as listed peers trade below asset value, forcing limits on withdrawals. Carried 1 October at medium probability, high impact and rising; still rising.
India's currency defence moves the strain into its bond market
low · mediumThe rupee firmed to about 95.3 per dollar, but Indian debt funds are set to extend a $4.5 billion bond sell-off after the central bank's policy shift, banks are asking for relief from the new defence measures, and blackouts are hitting factories as coal stocks run down. The scenario is the cost of the defence surfacing as tighter domestic credit. Carried 12 October at low probability, medium impact and falling; now stable.
— Antevo Executive Brief