U.S.-China Relations
NextFin News - The bond market has just logged its worst monthly selloff in two years, and this time the damage is spilling into stocks. The benchmark 10-year US Treasury yield spiked to a 19-year high above 5.27% overnight, up nearly 50 basis points through September, while Brent crude climbed to $106.60 a barrel with no breakthrough in the Middle East in sight. Asian equities slipped, the Nasdaq fell 0.9%, and China's CSI300 was pinned to a one-year low. Only a record $150 billion expansion of Nvidia's share buyback kept the rate-sensitive tech index from falling further.
The central question is no longer whether higher yields hurt stocks — they always do, eventually. It is whether this is a cyclical repricing that will fade when growth slows, or a structural break that resets the price of every asset for years. Three forces argue this time is different: fiscal deficits that need financing, inflation with a supply-side component that interest rates cannot fix, and an AI capital supercycle competing with governments for long-duration money. The era of cheap capital was the exception, not the rule.
The Double Squeeze: Oil and Yields Rising Together
For most of the post-pandemic rally, investors held one comforting belief: when stocks fell, bonds would catch them. That 60/40 logic depended on yields falling whenever growth fears rose. September inverted the relationship. Yields are climbing not because growth is surging, but because inflation is sticky, government borrowing is swelling, and oil — the original inflation tax — is back above $100 on geopolitical supply risk.
The move has run far, fast. The 10-year yield has gained nearly 50 basis points in September alone, its heaviest monthly selloff in two years, and now trades more than a full percentage point above where it sat a year ago. The two-year yield, which tracks the Federal Reserve's policy path most closely, has moved even further: up more than 57 basis points this month to the 5% threshold. Traders are pricing three more Fed rate hikes by the middle of next year, a sharp reversal from the cut expectations that dominated earlier in the year.
Equities spent the summer largely shrugging off the bond rout, but the tolerance is wearing thin. Overnight, only Nvidia's $150 billion buyback boost — the largest share repurchase authorization increase on record — held the Nasdaq to a 0.9% loss. Across Asia, bond markets in Japan, South Korea and Australia traded under pressure and most regional equity markets slipped. China's blue-chip CSI300 index was pinned to a one-year low after US plans to ban Chinese components from data centres hit the technology sector, with the Shanghai Composite down 1.6% and the CSI300 down 2.5%. Hong Kong's Hang Seng dipped 2%. The dollar was headed for a monthly gain.
Sovereign yields are the reference price for every risky asset — the discount rate applied to future earnings, the benchmark for mortgages, the cost of corporate borrowing. When that reference price resets higher, everything downstream reprices: governments pay more to service debt, companies face steeper borrowing costs, and households see mortgage and credit-card rates climb. The transmission is mechanical. The open question is whether it is durable.
Why This Time Is Different: The Structural Case
Is this a cyclical overshoot that will revert, or a regime shift that will not self-correct? A cyclical call requires a short-term driver and a demonstrated pattern of mean reversion. A structural call requires evidence of a permanent regime change — a rule, a technology, or an industry structure that will not unwind on its own. Three forces point to the latter.
First, fiscal arithmetic. Interest expense is becoming a growing line item in developed-market budgets, which forces governments to issue more debt into a market already demanding a higher term premium. That loop is self-reinforcing: more supply at higher yields, and higher yields feeding larger deficits. The term premium — the extra compensation investors demand for holding long-duration risk — is the market's fear tax on that dynamic, and it is no longer zero. History shows the pattern: after the Second World War, the US ran down its debt burden partly through financial repression, holding yields below inflation for years. Today's governments have the same incentive, but they lack the same control. Capital is mobile, inflation is import-driven, and central banks that spent a decade fighting deflation are reluctant to re-ignite it with yield-curve control. The result is a market-determined term premium that stays positive.
Second, the inflation shock has a supply side that central banks cannot fix with rates. Oil above $100 on Middle East disruption is not a demand phenomenon; it is a capacity-and-risk phenomenon. Rate hikes can crush demand, but they cannot drill new wells in a conflict zone. That is why the combination of rising yields and rising oil is so uncomfortable: it is stagflationary pressure, the one mix central banks are worst equipped to handle. The 1970s analogy is imperfect — today's economies are far less energy-intensive — but the mechanism is the same: a supply shock that raises costs while slowing output leaves policymakers choosing between inflation and recession rather than defeating both.
Third, the AI capital supercycle is absorbing savings at a scale that competes with sovereign issuance. The prospectus for US AI giant Anthropic illustrates the size of the gamble: the company is targeting a $2 trillion valuation while planning to spend $518 billion on computing and infrastructure. To put that in perspective, $518 billion is larger than the annual defence budget of most major powers and rivals the entire GDP of countries such as Sweden or Belgium. That is private-sector demand for long-duration capital colliding with public-sector borrowing. When both the state and the most capital-hungry industry in history are bidding for the same pool of savings, the clearing price of money rises. And unlike the dot-com boom, this capex is already hitting the national accounts — data centres, grids, chips — making the demand for capital real rather than promised.
"The way to look at expected returns and overall bond yields going forward (is) we're coming to a new environment," said Angus Hui, head of fixed income at Fullerton Fund Management in Singapore. "Interest expenses are increasing as a part of the government budget in a lot of developed markets, which means stretching sovereign finances and perhaps limiting a recovery for bonds should the global economy slow down. Hence we think bond yields are unlikely to go back to the very good old days when bond yields were very, very low."
That final sentence is the structural thesis in one line: the "very good old days" of near-zero yields were an anomaly produced by a unique mix of demographics, globalization and central-bank balance-sheet expansion — not a baseline to which markets naturally return.
The Counter-Argument: Stocks Have Shrugged Off Bond Rout Before
The strongest case against the structural-break read is simple and data-backed: stocks have not cared yet. Through much of the summer the S&P 500 hovered near its all-time high set on August 13, even as the 10-year yield marched higher, because AI-driven profit growth and economic resilience kept investors buying every dip. Second-quarter earnings for S&P 500 companies were expected to have grown 53% year-on-year, or 49.5% excluding energy, according to LSEG data. Profits were forecast to jump 35% in 2026, sharply above the 14% growth of the prior year. As long as earnings grow faster than the discount rate rises, multiples can compress without prices falling.
There is also a correlation argument. Analysts note that a positive stock-bond correlation emerged after the pandemic inflation shock, meaning stocks and bonds can move in the same direction — which removes the case for bonds as a hedge and lets equities keep rising even as yields climb. Strategists at HSBC argued in mid-September that "sovereign bonds no longer diversify or hedge risk assets as effectively," so allocations to bonds have been falling while investors turned to short-term hedging strategies.
This counter-thesis is real, and it explains why the Nasdaq's 0.9% drop looks more like a wobble than a rout. But it rests on one assumption: that earnings growth can outrun the discount rate indefinitely. That is the gap the structural view attacks. Earnings forecasts were set in a low-yield world; if the 10-year yield settles structurally above 5%, the earnings multiple compresses faster than estimates get revised down. The positive correlation that protects stocks today is precisely what removes their hedge tomorrow — when yields rise on inflation rather than growth, stocks have nowhere to rotate.
There is also a historical check on the "stocks shrug it off" view. The summer resilience came while the Fed was still on hold and while oil spent periods below $90. The current setup — a hiking cycle underway, oil above $100, and the 10-year at a 19-year high — has no clean analog in the 2023-2026 period. Each previous leg of the bond selloff was met with a dovish pivot or a growth scare that forced yields back down. This time, the pivot is being priced out, not in.
The Second-Order Trade: What the Market Has Not Priced
The first-order effect — higher yields, lower equity multiples — is fully priced. The second-order effect is not. It runs through the currency and policy channel, and it is where the squeeze tightens.
As short-term rates settle at their highest levels in years, the global carry trade unwinds unevenly. The Australian dollar was steady at $0.7012 with a Reserve Bank of Australia rate hike fully priced and another expected by February; the yen hovered at 157.31 per dollar after Japan's top currency diplomat warned traders to heed the united concern from Tokyo and Washington over yen weakness; the euro held at $1.1367. The Reserve Bank of Australia was expected to lift the cash rate 25 basis points to 4.60% at its September 28-29 meeting, with NAB, Commonwealth Bank and Westpac all calling the hike. Australia's inflation had eased to 3.5% in July, but the trimmed-mean measure remained stuck at 3.6%, well above the central bank's 2-3% target band.
But there is a trap in that consensus. Westpac's rates strategists, Damien McColough and Uma Choudhury, struck a note of caution: "We question whether the governor can be sufficiently hawkish to frank the markets current mindset, especially if the decision is not unanimous." The logic cuts both ways. If central banks cannot be hawkish enough to match what markets have already priced, yields keep rising and the tightening does real damage to debtors. If they are too hawkish, growth breaks and credit spreads widen. It is a narrow path, and the bond market is betting it cannot be walked.
The same trap faces the Federal Reserve. The September dot plot moved the median policy path higher across the next several years, with officials signalling one more hike in 2026 and a funds rate around 4.1% in 2027. But if those hikes arrive because inflation is supply-driven rather than demand-driven, each 25 basis points buys less disinflation and more recession risk. That is the second-order consequence the market is not fully grappling with: the policy response to this inflation mix is less effective and more costly than the policy response to the 2021-2022 demand shock.
The transmission to the real economy is already visible in the most rate-sensitive sectors. Higher mortgage rates cool housing turnover, which hits real-estate services, appliances and home improvement. Higher corporate borrowing costs push leveraged buyouts and refinancing to the sidelines, which hits private equity and the banks that syndicate their debt. Higher discount rates compress the valuations of any business whose earnings sit far in the future — which is most of the technology sector. The bond market is not a sideshow; it is the plumbing, and the plumbing is clogged.
Conclusion: Winners, Losers, and the Signal That Would Break the Thesis
A structural call has concrete implications. The beneficiaries of a higher-for-longer rate world are short-duration assets, money-market funds, and companies with net cash positions that earn more on their balances than they pay on debt. Financials with floating-rate loan books can benefit if credit costs stay contained. The exposed are equally clear: long-duration growth stocks priced on distant earnings, highly leveraged companies and households with variable-rate debt, and the sovereigns whose interest bills are compounding fastest.
The outlook splits by time horizon, and collapsing them into one verdict is the most common error. In the short term, sentiment and liquidity dominate — a Middle East de-escalation or a soft inflation print could trigger a sharp relief rally, and the AI earnings machine keeps delivering beats. Over the medium term, fundamentals take over: if 5%+ yields persist into earnings season, estimate revisions will start to lag multiple compression, and that is when equity downside accelerates. Over the long term, the regime question decides everything — if fiscal deficits and AI capex keep competing for capital, the neutral rate has moved up and will not come down on its own.
Three scenarios frame the path. The base case is yields grinding between 4.8% and 5.4% while equities churn in a narrower range, with stock selection mattering more than index beta. The upside case is a growth scare that forces the Fed to pause — yields fall, but only because recession risk has risen, which is not the relief rally bulls imagine. The downside case is an inflation re-acceleration on oil that pushes the 10-year through 5.5%; at that level, the equity multiple has nowhere to hide and earnings estimates begin to crack.
The falsifying signal is specific and observable: if the 10-year yield falls back below 4.5% on a growth scare while equities hold their gains and credit spreads stay tight, then this was a cyclical overshoot, not a regime change, and the "new environment" thesis is wrong. Below 4.5%, the market is pricing a recession without inflation — a very different world from the one described here. Watch that level, and watch whether the next Fed meeting delivers hikes or a pause.
The bond market is not asking whether stocks are expensive. It is telling investors that the price of money has changed — and that cheap capital was the holiday, not the job.
Explore more exclusive insights at nextfin.ai.
