- Fixed Income Portfolio Manager
United States, Institutional
Change
The views expressed are those of the author at the time of writing. Other teams may hold different views and make different investment decisions. The value of your investment may become worth more or less than at the time of original investment. While any third-party data used is considered reliable, its accuracy is not guaranteed. For professional, institutional, or accredited investors only.
We asked everyone the same two questions,” said Vinny. “'What is your assumption about home prices, and what is your assumption about loan losses.’ Both rating agencies said they expected home prices to rise and loan losses to be around five percent — which, if true, meant that even the lowest-rated, triple-B, subprime mortgage bonds crafted from them were money-good.
While most people consider the start of the global financial crisis (GFC) to be the collapse of Lehman Brothers in September 2008, the lead-up began much sooner. Mortgage delinquencies and defaults began to materialize in early 2007. Loose US lending standards and a galloping housing market had led banks to issue loans to millions of borrowers who could ill afford them. Lax regulations, meanwhile, had allowed markets to create complex mortgage-backed securities that included risky credits. When the housing market faltered, loan losses piled up. The prevailing narrative quickly shifted from “house prices can’t fall” to “loan losses will be contained” to “some mortgage bonds are still ‘money-good,’” meaning that although securitizations were taking losses, ultimately the highest-rated bonds would not lose principal. The rest, of course, is history.
That history is now rhyming with the current AI spending narrative.
Is the Jevons paradox the new “money-good”?
It may seem strange to relate AI-model competition to the direction of Treasury yields but increasingly, the former is having an impact on the latter. In my yen carry trade article, I discussed how international balance-of-payments dynamics explain foreign-capital financing of US technology companies. Today, it is worth looking at the dramatic impact the AI spending by these same companies is having on the US economy.
Estimates vary, but many economists think AI infrastructure spending will contribute 0.75% – 1.5% to US GDP this year.1 Until recently, there was little reason to believe that would not materialize. Now, however, the Chinese company DeepSeek claims it was able to create a top-tier AI model with significantly less resources and training than its US rivals. The company also announced plans to provide its model via open-source to developers. If both statements are true, then today’s prevailing narrative that massive capex for AI-model development — which has been fueling US growth —may face scrutiny, particularly as models become commoditized and free.
Following the DeepSeek news, US tech companies have not announced plans to reduce capex, nor has the market drastically penalized them for that. In my mind, there are two reasons for this. First, over the past few years, the stock market has rewarded companies with the highest earnings momentum. In an elevated capex cycle, earnings momentum should be strong, as a company’s return on investment can show up immediately in earnings while capex can depreciate over many years. That explains substantially higher earnings growth, compared to free-cash-flow growth.
The second reason has to do with the Jevons paradox, which holds that as technology becomes more efficient, consumption counterintuitively increases, often with unintended consequences. A couple of months ago, no one was thinking about the Jevons paradox in the context of AI capex. Tech companies were spending massive sums on AI to develop new cutting-edge models. Post-DeepSeek, they are now arguing that major capex is necessary because of the widening model usage and the need to fund further investment in inference and data-intensive agentic AI. So far, they have generated an attractive return on these investments, so perhaps they are correct. But this is a significant shift in the narrative, similar to that with mortgage bonds throughout 2007 and 2008.
The Fed’s AI spending conundrum
This brings us to the Treasury market and the Federal Reserve (Fed). People often ask me what rising bond yields means for stocks, but the better question is what do rising stock prices mean for bond yields? The Fed’s job has become more difficult with the increase in AI spending, because that has been uncorrelated with interest rates.
Typically, when the Fed raises short-term rates, growth slows in interest-rate sensitive parts of the economy such as housing and manufacturing. The decline in aggregate demand from investment then causes inflation to decelerate. Today, the seemingly existential need for companies to invest in AI has decoupled this spending from interest rates. (The stock market has also made financial conditions quite easy for the tech sector by continuing to reward its capex.)
That is the Fed’s conundrum. It has insisted that recent interest-rate hikes are restrictive, and they are, if you want to buy a house or a car or if a small business wants to borrow money. For a tech behemoth aiming to raise its next round of funding, however, it makes little difference if the Fed’s policy rate is 0% or 8%.
Another challenge for the Fed is that long-term interest rates are rising despite the 100 basis point cuts to short-term rates. Figure 1 shows the recent performance of cyclical stocks relative to defensive stocks compared to the US 10-year Treasury yield. A relatively tight correlation had existed for most of the past two years, which makes sense: Yields generally decline as the market begins to worry about growth, and as cyclical stocks start underperforming defensive stocks.
Figure 1
Starting in 4Q24, a meaningful break in that pattern occurred. Some might point to the markets’ worries about deficits in the wake of the election, but I take a different view. I believe that at least part of the disconnect is because despite high policy rates, the US economy continues to chug along, with investors rewarding AI beneficiaries. This has created a wealth effect that has boosted consumption and GDP, albeit amid deceleration among some interest-rate sensitive and cyclical segments.
So what effect might the DeepSeek news eventually have? If declining model-training costs continue to justify large-scale capex for development of the best, most powerful new AI applications, then it is likely that US growth will continue to surprise to the upside, keeping bond yields elevated. But if the Jevons paradox is this cycle’s version of “money-good” and investors start penalizing companies for what they perceive as wasteful outsized spending, then yields could decline materially from here.
1Beth Kindig, “AI spending to exceed a quarter trillion next year,” Forbes, 14 November 2024.
Stay up to date with the latest market insights and our point of view.
