Why This Economy Refuses To Break | David Cervantes

David Cervantes identifies the AI buildout as the single biggest macro driver right now, with capital expenditure running at approximately one trillion dollars and still rising. The funding sequence has moved through three phases: hyperscalers first redirected free cash flow away from buybacks, then tapped debt markets around December through February, and are now issuing equity, with Google's roughly 80 billion dollar equity issuance cited as evidence of that progression. Federal deficits running at 6 to 7 percent of GDP, a level last seen during World War II and far above the Reagan-era military buildup deficits of 3 to 4 percent, combine with AI capex to create a continuous flow of money through the economy that makes recession very difficult to engineer. Cervantes frames public deficits as private sector surpluses and argues the economy has already absorbed the 2022 rate shock, the tariff war, kinetic war, and an oil price shock without tipping into recession.

Unlike prior episodes of market exuberance, Cervantes argues current equity valuations are being driven by earnings expansion rather than pure multiple expansion, though he flags a potential bubble in earnings expectations and questions whether those earnings will materialize fast enough to fund the next phase of the buildout given hardware obsolescence cycles of roughly five years. He identifies profit margin expansion as the most meaningful forward-looking signal and describes the current productivity boom as the first pro-cyclical productivity boom since the 1990s. The manufacturing PMI printing at 54 reflects a restocking cycle he attributes to an echo of the COVID supply chain shock and tariff-driven distortions rather than the AI buildout directly. The structural shift from just-in-time supply chain optimization toward resiliency and buffer stock management is, in his view, a permanent change that raises working capital demands and will eat into long-run profitability, though it cannot be reversed quickly once capital has been committed.

Cervantes builds a three-legged framework for consumer resilience. The first leg is a stealth intergenerational wealth transfer from baby boomers to adult children covering expenses like vacations and childcare, which he acknowledges rests partly on conjecture since informal transfers below the roughly 13,000 dollar annual gift tax exclusion threshold go undocumented. The second leg is the wealth effect from asset appreciation, with the S&P 500 returning approximately 300 percent since 2009, though he notes this effect is concentrated in the upper 50 percent of the population since the bottom 50 percent holds little to no equity wealth. The third leg is that somewhere between 40 and 60 percent of the population no longer carries a mortgage, freeing up significant spending capacity. He argues that commentary about the savings rate falling below 3 percent in the most recent PCE release is misplaced because it ignores the large rise in household wealth, even as credit card delinquency rates are accelerating meaningfully.

On March 1st Cervantes called zero Fed rate cuts by year end, based on a broadening inflationary impulse he identified before any oil shock. He points to January CPI coming in at 0.42 month over month as an early signal and argues the disinflationary tailwind from declining rents has largely been harvested. The labor market has tightened since January with unemployment falling from approximately 4.4 percent and potentially reaching 4 percent by year end, in part because presidential immigration policies have mechanically shrunk the labor supply. The ISM prices paid component is running at levels not seen since 2022, leading Cervantes to reject the view that current inflation is simply an oil story.

Cervantes is skeptical of Kevin Warsh, describing him as not independent, and criticizes his push to substitute the Dallas Fed trimmed mean PCE for core PCE as the Fed's inflation benchmark. He argues the Dallas methodology is asymmetric, trimming the top 34 percent of high price observations but only the bottom 20-something percent of low price observations, which mechanically produces a lower inflation reading. The Dallas trimmed mean currently sits around 2.5 percent, the Cleveland Fed's symmetric alternative is near 2.9 percent, and traditional core PCE is at 3.3 percent year over year. Cervantes predicts Warsh will struggle to build FOMC consensus, noting Beth Hammack has stated publicly that persistent inflationary prints will shift the conversation and that Christopher Waller appears to be moving in a hawkish direction. He expects rate hike pricing in the curve to accelerate as markets front-run policy before any actual hike occurs, citing the September 50 basis point cut as an example of a move large enough to shift markets beyond what was already priced in.

The recent sell-off in the long end of the bond market was driven more by rising term premium than by rising real rates, which Cervantes attributed to uncertainty about which inflation target the Fed is actually using. The 10-year Treasury yield peaked at around 4.60 percent approximately three weeks before the recording, below the 5 percent level it tagged in October 2023. He argued that higher nominal rates and rising equity markets can coexist when the factors driving rates are not harmful to the economic cycle, pointing to corporate profit margin expansion, accelerating earnings, and the AI data center buildout as tailwinds more powerful than Fed policy or bond market levels.