Inflation, Oil Prices, Bond Yields and the One Brick Holding Markets Together
Written by the Editor-in-Chief – October 2026
Before we start, we want to go over a few key indicators we are currently seeing within markets and the economy, as of the close on October 6, 2026:
Those three numbers would not make sense together if the S&P 500 were not coming off one of the strongest earnings seasons in years, with earnings growth the strongest since 2021 and reported margins at a record, though flattered by one-off investment gains at a few mega-caps, with another strong quarter expected, in no minor part driven by AI, both through spending on AI infrastructure and through early cost savings.
So, what exactly does this mean and where do we go from here?
In the past, either of the first two indicators on its own has been enough to send markets into correction territory, at least over the short term – and earlier this year it did, with the S&P 500 down roughly 15% from its peak in March as oil spiked. So, the only way for us to interpret this is that markets believe AI benefits are here to stay, and so far, are strong enough to counterweigh headwinds in the broader economy.
Markets are, and always will be, a balance of opposing forces, and the two most familiar ones, even if you have never worked in finance, are supply and demand. However, how and why would supply and demand be affected by this?
Well, higher yields mean that low-risk investments, in this case government bonds, become more attractive because they pay investors a higher return. This theoretically attracts additional buyers and brings money out of the stock market and into bonds. Additionally, high oil prices hit consumers like a tax. With oil-derived petrochemicals going into roughly 95% of manufactured goods worldwide, companies will try to pass the higher costs to consumers rather than have their margins contract, a.k.a. inflation.
Inflation tends to come from one of two sources:
We are clearly in the latter, and supply-driven inflation is harder to rein in, as these macro forces (in this case the war in Iran) cannot be fixed as easily as raising rates to cool down consumer demand and bring back inflation to a manageable level.
Even more worrisome are the factors pushing yields higher, as the drivers below are interconnected and could create a reinforcing spiral that can make things go very bad very quickly until yields reach a level attractive enough to bring buyers back into Treasuries.
Fiscal deficits create the supply of bonds, inflation shapes monetary policy expectations, the unwinding yen carry removes a previously reliable source of demand for US Treasuries and AI borrowing competes for the same pool of capital (offering a more attractive return as they are riskier than Treasuries, supposedly the safest investment on planet Earth). We understand some of these may be unfamiliar or sound complex, which is why we will try our best to explain them simply below.
Fiscal deficit: US government debt recently surpassed $40 trillion for the first time ever, because the government spends more than it collects in taxes, and when there's a shortfall, it funds the gap by… issuing Treasuries. Simply put, this is the supply of Treasuries, and as we know, the more supply of a good, the lower we can expect the price to be (in this case, the more interest the government has to offer to finance its deficit).
Fiscal spending and government deficit are the largest single source of bond supply. The US federal deficit is running at approximately $2.1 trillion for fiscal 2026, or about 6% of GDP. The Treasury needs to borrow roughly $175 billion every month just to fund ongoing operations.
This is not a US-only phenomenon: France, Britain, Italy and Brazil, among others, all run fiscal deficits outpacing their GDP growth. When a government runs a fiscal deficit that consistently outpaces GDP growth, the debt-to-GDP ratio keeps rising.
In simple terms, governments are increasing their credit card debt every year rather than paying it down. If you did this and suddenly got a substantial pay raise, you could earn your way out of the problem. A country's version of a pay raise is faster growth, either by i) putting more people to work or ii) getting more output from each worker. The issue is that none of these countries is growing anywhere near fast enough for that to be the case, and with unemployment already low, there are few idle workers left to put to work. As anyone who has ever carried a credit card balance can tell you, this is far from prudent. Whether it's a country, a company or a household, too much debt leaves little room to respond when a shock hits.
AI project finance and "yield competition": this is the newest and most structurally novel driver. The AI infrastructure buildout – data centers, cooling systems – requires capital on a scale that exceeds what even the largest technology companies can fund on their own. As a result, these companies ("hyperscalers", think the Amazons, Googles, and Metas of the world) have turned to the bond market in force, and the numbers are staggering. Goldman Sachs estimated in July that $489 billion of AI-related debt had already been issued in 2026, surpassing its $322 billion estimate for all of 2025.
This is crowding out, but in reverse, an idea explored by Lombard Odier's Florian Ielpo. Traditionally, economists worried that government borrowing would crowd out private investment. What is happening today is that corporate borrowing for AI is, in part, crowding out the government.
As Fed Chair Kevin Warsh said at the press conference after the September 2026 meeting: "The so-called hyperscalers are out in the market raising funding. And so the competition for capital is real. And I think it partly explains the increase in yields."
Savings available to fund bonds are not unlimited. When two massive borrowers, the Treasury and hyperscalers, show up in the market simultaneously, both must offer higher yields to attract investors. The result is upward pressure on yields across the curve, including for Treasuries.
The Japan carry trade unwind: you have probably heard about the Japanese carry trade over the past few weeks. For years, the yen carry trade (borrowing in a low-interest country like Japan and investing it at a higher yield like the US) was a cornerstone of global market liquidity. Institutional investors borrowed yen at rock-bottom rates and used the proceeds to buy higher-yielding assets worldwide, particularly US Treasuries. This steady stream of artificial demand allowed governments to run large deficits without triggering a spike in borrowing costs.
That era is ending. The Bank of Japan has been gradually raising its policy rate, most recently to 1.25% in September, its highest in 31 years, and Japan's 10-year yield has moved above 3% for the first time since 1996. As Japanese yields rise, the economics of the carry trade weaken. Borrowing in yen becomes more expensive and investors come under pressure to unwind their positions before losses build. Unwinding involves selling foreign assets, including US Treasuries, to repay yen-denominated loans, further increasing the supply of US bonds in the market.
Although economists debate how much the carry trade contributes to the overall rise in yields, the unwind removes a buyer from the market precisely when you would most want one. Perhaps it is not the main story, but it amplifies the impact of the other drivers by reducing the market's capacity to absorb new supply.
Inflation and monetary policy expectations: this is the driver that ties the other three together. A bond's yield reflects two things: where investors expect short-term interest rates to average over the life of the bond, and an extra cushion for the risk of locking money up for that long (the "term premium," which we come back to later). With oil at ~$100 and inflation driven by supply rather than demand, the Fed has turned to raising rates, and investors want more compensation for the risk that inflation stays higher for longer.
This puts the Fed in a bind. Raising rates does little to fix a supply shock (higher rates will not produce more oil), but holding back while prices are still rising risks letting inflation become entrenched. As long as markets believe the Fed will hold rates high, or worse, that it may stop too early and lose control of inflation, long-term yields have little reason to fall. In simple terms, if you lend someone money for ten years and expect inflation to eat away at what you get back, you will ask for a higher interest rate to make it worth your while.
In short: the bond market is telling a story that the equity market, at least at an index level, is not yet pricing in. The cost of capital is rising, the supply of savings is finite, and the competition for that capital is intensifying, yet the only equity index showing any meaningful drop from its all-time high is the Russell 2000, down 7.8% from its August high. The S&P 500 and the Nasdaq, the latter usually highly sensitive to rates, both closed at new all-time highs on October 6.
The path forward hinges on a single, uncomfortable truth: the stock market has been pricing a future where capital remains abundant and cheap, while the bond market keeps sending the message that this will not be the case.
Today, a handful of mega-cap stocks are carrying the index while the typical stock lags behind (what analysts call weak "breadth"). That gap has to close one of two ways: either the index falls to meet the rest of the market, or the rest of the market rises to meet the index. Morgan Stanley's Mike Wilson puts it plainly: the share of S&P 500 stocks trading above their 200-day average has fallen from roughly 75% to below 50% while the index has held up, a divergence that "cannot persist forever." Which way it breaks will most likely be decided by volatility in the bond market.
If the MOVE Index (a measure of Treasury volatility) stays elevated while the VIX (its stock-market equivalent) stays calm at around 15, Wilson's view is that bond volatility could spill over into stocks and pull the S&P 500 down by about 5% to 10%. He does not see that as a bad outcome: a final drop in the index is often how a multi-month correction beneath the surface ends.
If bond volatility eases, the opposite could happen: the rest of the market catches up and the rally rests on a broader foundation. One possible catalyst would be the Fed stepping in with liquidity if financial conditions tighten too far.
This is where the damage could become structural rather than sentiment-driven. Lower-rated US companies, many of them small caps, face a rising wall of debt: high-yield bond maturities are projected to climb from about $68.5 billion in 2027 to $314.1 billion in 2029, according to a Reuters analysis of LSEG data. Companies that locked in 4%–5% rates in 2021 would refinance today at more than 8%, a coupon shock of 300–400 basis points that directly siphons free cash flow.
The risk is a slow-motion solvency crisis for the weakest quartile of the market. About 40% of Russell 2000 companies have no earnings or negative earnings, according to Apollo chief economist Torsten Slok, up from 14% two decades ago.
These companies will likely not fail overnight. Instead, they are more likely to quietly restructure, issue new shares that dilute existing holders, sell assets or, eventually, fail. The overall index can hold up even as its weakest 10% is wiped out.
The Fed's response is likely to be the make-or-break factor. If the Fed tolerates more pain to break inflation, the refinancing wall could become a genuine credit event. If it provides liquidity preemptively, the reckoning would be deferred but not resolved.
The long-term outcome depends on whether the forces driving yields higher are cyclical or structural. The evidence points strongly to structural.
The supply of bonds is likely to stay high because it is driven by three forces that will not easily change:
In our view, the resolution is not lower rates – it is a structurally higher cost of capital. The era of zero rates is dead. The 2010s were defined by negative net equity supply (buybacks exceeding issuance) and central banks keeping long-term borrowing costs artificially low (suppressing the "term premium", the extra yield investors demand for locking up money for longer). Both tailwinds are reversing, with Goldman Sachs estimating that US net equity supply will be nearly flat in 2026, ending more than two decades of shrinking share counts through buybacks.
In other words, winners and losers are likely to be decided by how well each business holds up when borrowing costs 5% or more, not by which industry it belongs to. A cash-generating company can thrive in this environment whether it sells software or groceries, while a company that depends on cheap debt to survive is exposed no matter what sector it is in.
Anthropic's upcoming IPO sits at the center of everything discussed so far. The scale is all but unprecedented: a reported valuation of $1.8–$2.0 trillion, roughly 2x the most recent funding round of $965 billion, would top even SpaceX as the largest IPO in history. JPMorgan expects more than $260 billion of equity issuance in 2026, with SpaceX's $86 billion raise and an Anthropic offering reported to be at least as large making up most of it.
The mechanism of impact is not the IPO itself, but the selling that precedes it. Ahead of the SpaceX IPO, hedge funds cut their positions in mega-cap tech to make room for allocations, according to JPMorgan data. The Magnificent Seven went on to lose more than $2.3 trillion in market value in June, the month SpaceX listed, though worries about AI spending also drove that decline.
Anthropic's IPO matters – as a catalyst, not a cause. We expect it to force rotation within the AI complex, with capital flowing from adjacent names (smaller AI companies, AI-adjacent public equities) into Anthropic. When SpaceX went public, its listed proxies (stocks investors had used as stand-ins for SpaceX exposure) such as EchoStar and Destiny Tech100 fell roughly 15% on the first day of trading as investors consolidated into the category leader. While Anthropic is not a full-ecosystem player like SpaceX, it would still pull capital from somewhere, even if the impact is less severe than that of SpaceX.
So, supply shock or symptom? In the short term it acts as a supply shock, but at its core, Anthropic's IPO is a symptom of the same structural force driving bond yields higher: the AI buildout requires capital on a scale that exceeds internal cash generation. The IPO is not creating the capital scarcity; it is revealing it.
So, back to the one brick holding markets together: AI-driven earnings. As long as AI keeps lifting margins faster than rising borrowing costs eat into them, the index can keep shrugging off oil at ~$100 and a 10-year yield above 5%. But a market resting on a single brick is only as stable as that brick, and every force we have walked through (deficits, AI borrowing, the carry trade unwind and sticky inflation) is pushing in the same direction: a higher cost of capital, for longer.
We are not calling for a crash. We are saying the margin for error has narrowed. Over the coming quarters, we will be watching three things: whether bond volatility stays elevated, how the small-cap refinancing wall plays out from 2027 onward, and whether AI earnings keep pace with the capital being raised to fund them. Our positioning follows from that: we favor businesses that generate their own cash, and are wary of any that need cheap money to survive.
Our thesis rests on a higher cost of capital for longer, and several things could prove it wrong.
First, the supply shock could fade faster than we expect: a durable resolution in Iran that sends oil well below current levels would cool inflation, let the Fed stop hiking, and pull long-term yields down with it.
Second, we may be underestimating the "pay raise." If AI delivers the productivity gains that margins are already hinting at, faster GDP growth could stabilize debt-to-GDP ratios and calm the bond market, turning today's competition for capital into a temporary bulge rather than a permanent regime.
Third, policymakers have more tools than we have given them credit for: the Treasury can shift issuance toward short-term bills, regulators can loosen bank capital rules to create new buyers of Treasuries, and the Fed has repeatedly shown it will step in when markets seize up. Small caps may also prove more resilient than feared, as private credit and extended maturities soften the refinancing wall.
Finally, the risk could run the other way. If AI returns disappoint and hyperscalers cut back on spending, borrowing pressure would ease and yields would fall, but the single brick holding up the index would crack with it. In that scenario, we would be right about the fragility, but wrong about where it comes from.
AI tools were used to research, validate, and review this memo, however all ideas and views expressed herein are those of The Brief's Editor-in-Chief. This is not investment advice.
Markets, macro data, tech and geopolitics, distilled into one email. Free.