While strong corporate profits drove prices of the largest companies up, the price of most stocks went down. The market cap-weighted S&P 500 rose just over 2%, while the equal weight S&P 500 fell just under 2%. Mid-cap stock indices dropped more than 2% and the Russell 2000 small cap index plummeted more than 7%.
| Index | Q3 2026 | Year to Date |
|---|---|---|
| S&P 500 (Large Cap) – Cap Weight | +2.30% | +12.75% |
| S&P 500 (Large Cap) – Equal Weight | -1.87% | +10.03% |
| Russell 2000 (Small Cap) | -7.23% | +13.71% |
In the bond market, as predicted in our previous few updates, long-term interest rates continued to rise. At 5.27% as of this writing, the 10-year Treasury now yields about as much as the S&P 500 earns relative to its price. Further increases in rates are the main risk we see to this bull market.
Economic Overview
Investors began the year expecting the Fed’s incoming chairman, Kevin Warsh, to cut rates, as President Trump wanted. Instead, in September the Fed unanimously voted to raise them for the first time since 2023. Consumer prices (CPI) rose 3.4% in the year through August, or 2.4% excluding food and energy, both slower than in the spring. The Fed pays more attention to the Personal Consumption Expenditures (PCE) price index, which weights health care more heavily and housing less. PCE inflation usually runs below CPI, but excluding food and energy it was 3.0% in August, well above the Fed’s 2% target. Nearly all Fed officials projected another increase this year.
The Fed can do more about inflation driven by demand than about inflation driven by supply shortages; no rate increase will send a tanker through the Strait of Hormuz. The strait remains unsafe, and U.S. crude ended the quarter about 30% higher than it began. Heating oil cost about 50% more in August than a year earlier, and airline tickets almost 25% more. Several of the world’s largest fertilizer exporters ship through the strait, and the price of the most common nitrogen fertilizer nearly doubled between December and May, ahead of spring planting. Farmers planted less corn and wheat this year, and the smaller harvests will push up food prices next year regardless of what the Fed does.
Markets
This year’s market has been led by a small, volatile group. Technology stocks rose almost 39% through September, led by the chip makers, while consumer discretionary stocks were the worst of the eleven S&P 500 sectors, down about 8%. Higher rates have weighed on businesses that depend on borrowing, such as housing and autos, and consumer spending, about two-thirds of the economy, has not powered this bull market the way it usually does. Smaller companies gave back nearly all of their lead from earlier in the year.
Stocks moved in a sawtooth pattern during the quarter, and the leaders moved together. Chip stocks fell sharply, though temporarily, in midsummer; the Philadelphia Semiconductor Index dropped 28% between its June 22 record and late July. A drop of that size is tolerable in one stock, but not so great when it hits a large share of a portfolio at once. We try to ride stocks like these while they are rising, but we trim the sails a bit to limit losses.
Energy stocks have been the year’s other leader, up about 40% and priced for wartime oil. We have kept them below their index weight. We doubt that this “skirmish” with Iran lasts forever; even a number of Senate Republicans are calling for an end to it. When a ceasefire came into view this spring, crude fell nearly all the way back to its pre-war price, and another ceasefire could do the same. It seems clear enough that Iran will try to hurt Trump by keeping oil prices high through the November 3 midterms. A “peace is at hand” announcement before the vote, like the Nixon administration’s in 1972, is still possible, and that could produce a sharp rally in stocks, though not in oil stocks. After the election, the Administration has less political reason to hurry a settlement, and it’s anybody’s guess.
So far, higher rates have not dented corporate profits. S&P 500 companies are on track to report third-quarter earnings nearly 30% above a year ago, on revenue up 12%, with sales growing in every sector. Profit margins have widened in about half the sectors, most of all for chip makers and oil companies selling into shortages. They usually shrink once a shortage ends.
Analysts normally cut their estimates as a quarter goes on. This year they have been raising them, and they now expect S&P 500 index earnings per share to be about $400 over the next twelve months. Divide the S&P 500’s quarter-end level of about 7,650 by that $400 and you get a price-to-earnings ratio of 19, down from about 23 late last year and close to its 10-year average. Turn the ratio over and companies are expected to earn about 5.2 cents for every dollar of share price, while a 10-year Treasury paid 5.3% at quarter-end. By this measure the equity risk premium, the extra return investors normally demand for owning stocks instead of Treasuries, is gone.
That risk premium was also gone through most of the 1980s and 1990s, and stocks still did very well as investors correctly bet on continued economic and earnings growth. But interest rates were falling then, and they are rising now. If they keep rising, more investors may decide Treasuries are the better deal, and that is the main risk we see to this bull market.
On the other hand, a 10-year Treasury pays the same 5.3% every year until it matures, while corporate earnings can grow. Since 1950, S&P 500 earnings have grown at least 4% a year in about 80% of 10-year periods. If earnings grew even a fraction as fast as they have recently, stocks would come out ahead. Still, with stocks no cheaper than Treasuries, we are tapping the brakes by modestly trimming the positions that have risen furthest.
Artificial Intelligence
Investment in AI and other technology accounted for roughly a third of U.S. economic growth in the year through June. Building AI has strained the supply of chips, copper and electric power. High prices eventually cure a shortage by paying for new factories, power plants and mines and by driving buyers away. Electrical equipment looks like the next bottleneck, with large transformers taking about two years to arrive. Our technology holdings include chip makers selling into these shortages and cloud companies buying from them.
About seven in ten Americans told Gallup this year they would oppose a new data center near them. The greatest shortage in AI may turn out to be public goodwill. In January we suggested that rising power bills could lead lawmakers to make data center operators bear more of the cost. New Jersey did so in July. In September the House voted 417 to 3 for a bill that would require every state to consider doing the same. We expect more rules of that kind, and stricter testing of the most advanced models, but no drastic federal action. Neither party wants to be the one that stopped so much of the economy’s growth.
AI stocks are reasonably valued by the usual measures, though if earnings stall they can fall hard. Two of the largest cloud companies now spend more than their businesses generate, mostly on AI, and we worry about whether all that spending will earn a return. For that reason, we monitor free cash flows just as we do earnings. The AI buildout may turn out like the internet boom of twenty-five years ago, when telecom companies poured hundreds of billions into fiber-optic networks. In 2002, only a fraction of that cable was in use, and several of the builders had gone bankrupt. Demand caught up eventually, and much of the cable was put to work by new owners who bought it at a steep discount. Unlike most of the fiber builders, the cloud companies we own have large, profitable businesses in search, advertising, software and online retail.
Debt and Interest Rates
In January we warned that heavy Treasury borrowing could push long-term interest rates up even as the Fed pushed short-term rates down. Long-term rates have since climbed further, and the 30-year Treasury yield ended September at 5.64%, its highest since 2002. Rates have risen across the developed world as well. Japan’s 10-year yield is the highest since 1996, Germany’s since 2009 and France’s since 2002, and in early October Britain’s 30-year yield topped 6% for the first time since 1998. Part of this is a return to normal. From 2011 to 2021, as central banks bought trillions of dollars of bonds and inflation remained low, the 10-year Treasury yield mostly stayed between 1.5% and 3%, and in 2020 it fell below 1%. Over the past sixty years it has averaged close to 6%.
The rise also reflects government finances that keep getting worse. The Congressional Budget Office expects a federal deficit of $1.9 trillion this fiscal year, 5.8% of GDP. Deficits that large have historically come with wars or recessions. Federal debt held by the public has reached 101% of GDP and is on course to pass its 1946 record of 106% within a few years. Interest on this debt now costs more than $1 trillion a year. Neither party has shown much inclination to address any of this, except when it is in the minority. France, Britain and Japan face versions of the same problem. The companies building AI are borrowing heavily too, selling more than $150 billion of bonds this year, about ten times their 2024 total.
Before 2020, we held longer bonds completely in the corporate space when inflation was under control and the yield curve was normal, providing higher rates for the added risk of time and company credit. Since Covid, when rates hit abnormally low levels, we’ve stayed short, but flexible.
Long-term yields are the highest in a generation. When we sense stability and improvements in fiscal and monetary affairs are in sight, we will again build longer portfolios to lock in excellent rates. Until then, we are keeping liquid and short with T-bills and garnering some extra yield with agency bonds and small allocations to short-duration closed-end funds bought at a discount.
In short, we do not know when the war will end, how long the shortages created by the AI boom will last, or where long-term rates will settle. Stocks no longer offer the cushion over bonds they did for most of this century, and the government’s finances will not improve on their own. So, we are trimming the stocks that have run furthest and keeping our bonds short and liquid, ready for rebalancing to stocks should they decline or for redeployment to longer bonds should the rate picture change. Prudent asset allocation remains the best bet of all.