The Bond Market Is Pricing In Something Worse Than Inflation | Weekly Roundup

Hot CPI, including core inflation, and a very hot PPI initially pushed bond yields higher, but yields then fell sharply as investors moved toward safety and worries about weakening growth—especially an AI slowdown—overtook the inflation data. Quinn said record short positioning in bonds left the market vulnerable to a countertrend rally. He thinks the economy is still holding up and does not expect the AI trade to be allowed to implode, but believes a sustained decline in yields would require clear growth deterioration or an equity-market crash. He thinks the cyclical yield peak may be behind the market, while allowing for another high. Quinn briefly shorted long-end bonds and bought SOFR after the inflation release, then sold SOFR later that day; the other speaker was flat and hesitant to bet on further yield declines.

Despite two hot inflation reports, the market ended with October rate-hike odds around 25–30%, roughly where they began. Those odds had been about 70% two or three weeks earlier and briefly approached a full hike before retreating. Quinn said comments from Williams, Waller, Bowman, and Jefferson helped shift expectations away from an October move. With no Fed meeting in November, more data will arrive before the December decision. The speakers argued that tightening is already reaching the economy through energy and diesel prices, long-term yields, mortgage rates, and wider high-yield and hyperscaler and AI credit spreads. The 10-year yield had remained above 5% for more than a month, while the 2-year and 5-year yields had stayed above 4.5% for longer, with the 5-year around 4.7%. They said the effects of tighter financial conditions could take one to four months to show up in economic data.

One speaker described AI as the main support for top-line growth while much of the rest of the economy was weak. Anthropic’s third-quarter annual recurring revenue was still rising, but growth was below expectations and slowing. The speaker said that could matter for data centers, CoreWeave bonds, semiconductors, and memory. AI financing is also a major source of demand for debt capital, so slower expansion could reduce companies’ access to financing and the expected supply of AI-related debt. That might ease pressure on yields and other parts of the economy, but it would also expose how much recent growth and investment depend on the AI trade.

Weakness outside large technology stocks has received less attention, the speakers said. The Russell had fallen for eight or nine consecutive weeks, while regional banks, housing, and restaurants had been under pressure for two months; by contrast, three or four days of technology selling could provoke a crisis reaction. One speaker saw room for a bounce in areas already down 20–40%, but was uncertain beyond the next few weeks and did not expect a renewed bull market. Real wages had been negative for seven consecutive months, and the speaker characterized the bottom half of the economy as squeezed. The economy is also exposed to a reversal in AI-linked wealth: the top 10% own most assets and account for more than half of consumer spending.

CoreWeave debt has repriced sharply, but one speaker cautioned that falling bond prices do not mean an overnight default. Capital-intensive firms may need more debt, equity dilution, or both to finance projects. The speaker leaned toward higher-quality CoreWeave and Core Scientific debt being money-good, while acknowledging limited expertise. Hyperscalers might support infrastructure firms to protect assets they need, potentially helping creditors even as equity holders take losses or dilution. A government backstop alone would not make AI infrastructure equity attractive: it could sustain financing even where natural market demand is insufficient. The speaker saw power, energy, resources, and raw-material businesses as cleaner beneficiaries of government-driven demand than frontier AI labs.

The speakers disagreed about the likely policy path, with one saying both rate cuts and hikes remained plausible this year. Hikes could follow an inflation-expectations shock linked to financing-supply pressures, rather than current inflation alone. They said bond-market resistance to Treasury expansion and fiscal stimulus had helped contain inflation, but refinancing and financing vulnerabilities could weaken that restraint and encourage more inflationary support. A Treasury buyback reportedly received $40 billion in offers and accepted $6 billion, while the economy was still running deficits of about 5%. One speaker cited a chart suggesting hikes have had less economic effect as debt and deficits have grown, and argued a standalone 25-basis-point increase would do little. They cautioned that bonds are not a dependable safe haven, correlations can change, and brief positioning moves do not establish a lasting trend; further volatility and possible re-escalation remained risks.