David Rosenberg and Rich Bernstein on What Ends the AI Trade — And What They Own Instead

Episode Summary

Executive Summary: Richard Bernstein and David Rosenberg revisit lessons from the financial crisis, then argue the current market is defined by extreme AI-driven speculation, narrow U.S. equity leadership, and misallocated capital rather than healthy broad-based growth. Bernstein sees bubble-like conditions and wants international diversification; Rosenberg is more deflationary on the macro backdrop, skeptical the Fed should hike, and expects gold and non-U.S. assets to remain attractive over time.

Main Topics: Crisis-era lessons and analyst resolve (Priority: 4/5): They reflect on the 2008 housing/credit crisis as a formative period that sharpened their conviction, humility, and willingness to hold a view without becoming stubborn. Fed policy and the Taylor rule debate (Priority: 5/5): Bernstein argues conservative economists are inconsistent about the Taylor rule and notes multiple versions imply hikes, while Rosenberg argues the Fed should not hike because underlying inflation is easing and growth is below potential. AI as a speculative bubble and capital misallocation (Priority: 5/5): Both see AI as driving a concentrated, potentially late-cycle boom. Bernstein emphasizes bubble-like misallocation versus housing and real economy needs; Rosenberg says AI spending is crowding out the rest of business capex and will likely roll over via credit markets first. Inflation, productivity, and economic fragility (Priority: 5/5): Rosenberg argues inflation is decelerating, productivity is doing most of the work, and the U.S. economy is more fragile than markets imply. He sees real demand below potential supply and limited reason for tighter policy. Gold as a portfolio hedge and central-bank demand (Priority: 4/5): Rosenberg views gold as driven mainly by real rates, the dollar, and persistent central-bank buying, not just retail speculation. Bernstein frames gold as a 'spare tire'—a hedge, not a trade—used to protect against uncertainty. International stocks versus U.S. concentration (Priority: 5/5): Bernstein argues non-U.S. equities now offer better value and growth than commonly assumed, while Rosenberg warns the U.S. market is in a bubble-like, highly concentrated state and that diversification internationally is the best defense. Sentiment, positioning, and market-cycle warnings (Priority: 4/5): They lean on sentiment indicators, positioning, and historical analogs—especially Bob Farrell’s rules—to argue that crowded bullishness, low cash, and speculative behavior are late-cycle warning signs.

Key Arguments: The Taylor rule is being selectively invoked; multiple versions can be used to argue for hikes, but the real macro case does not support a tightening cycle. Inflation is cooling on underlying measures, especially once volatile shocks are stripped out, while real GDP growth is running below potential. AI spending is highly concentrated and may be inflating asset prices while starving broader parts of the economy of capital. The equity market’s strength is narrow and concentrated, unlike the broad-based late-1990s boom; that makes the current cycle more fragile. Credit markets are likely to signal any AI trade reversal before equities do, similar to the housing crisis. Gold remains a strategic hedge because central banks are still buyers and real rates/dollar moves explain much of its short-term volatility. Non-U.S. markets are becoming more attractive on valuation, dividends, and now increasingly growth, making global diversification more compelling. Extreme bullish sentiment, low cash, and high equity allocation suggest investors are heavily exposed if the AI trade reverses.

Data Points: Fed policy model references: 30 versions of the Taylor rule - Bernstein says the Atlanta Fed publishes many variants, all of which can be interpreted as implying hikes. U.S. real GDP four-quarter average: 1.4% - Rosenberg cites expected growth after third-quarter GDP data, below potential. Fed estimate of potential GDP growth: 2.0% - Used by Rosenberg to argue demand is below supply and disinflationary. Dallas Fed trim mean inflation: 2.2% - Rosenberg says this is below last year’s 2.7% and at a five-year low trend. Dallas Fed trim mean inflation one year ago: 2.7% - Rosenberg uses the decline to argue underlying inflation is cooling. Unit labor costs: 0.5% year over year - Rosenberg cites this as evidence of weak wage-driven inflation pressure. Real disposable income: Fractionally negative year over year - Rosenberg says real incomes are weak despite headline inflation narratives. AI share of business capex: About 50% - Rosenberg says roughly half of business capex is now AI/AI-related. AI-related capex growth: About 18% real growth - Rosenberg contrasts it with declining capex elsewhere. Non-AI capex growth: Negative year over year - Rosenberg says the rest of the economy’s capex is shrinking. Second-quarter GDP growth: 1.5% - Rosenberg says that is all the economy is getting despite AI spend. CAPE ratio: Above 41 - Rosenberg says the S&P 500 valuation is the highest since 2000. Historical frequency of current U.S. valuation: Top 1.5% of history - Rosenberg says the U.S. market is this expensive only 1.5% of the time. Household asset mix in equities: 73% - Rosenberg says household exposure to equities is at a record. Household asset mix in bonds: 7% - Rosenberg uses this to illustrate risk-on positioning. Mutual fund cash: 1% - Rosenberg says portfolio managers are extremely fully invested. Bearish recession expectation among global PMs: 2% - Rosenberg cites Bank of America survey data as evidence of complacency. Gold peak call: $6,000/oz - Rosenberg says his upside target for gold has not changed. Gold speculative peak cited: $5,500/oz - He references the prior parabolic move that looked overextended. Historical central-bank bullion share in reserves: Over 70% in 1980; about 10% by 1999; about 25%-30% now - Rosenberg uses reserve composition to explain gold’s secular bull case. Gold supply growth: About 1.5% annually - Rosenberg notes mined supply is stable and predictable. Gold demand growth: About 2%-3% annually - He attributes much of this demand to central banks. Non-U.S. equity vs venture capital: Non-U.S. stocks have outperformed venture capital over the last five years - Bernstein cites this as evidence that market leadership is shifting. Projected earnings growth cohort: About 200 companies globally with 25%+ projected long-term earnings growth - Bernstein says growth is not limited to the U.S. Magnificent 7 presence in that growth cohort: Only 1 company - Bernstein uses this to argue growth opportunities are broader than the Mag 7.

Pivotal Quotes: "I think it's been the era of the bubbles, right? We've seen very, very loose monetary policy again and again." — Richard Bernstein: Bernstein frames the last several decades as a repeated bubble cycle shaped by loose policy and capital misallocation. "The financial markets are here for capital formation... The prediction markets are here to make a bet." — Richard Bernstein: He argues the current environment has become hyper-speculative and confused about the purpose of markets. "I think the U.S. is in a huge bubble." — David Rosenberg: Rosenberg explicitly characterizes the U.S. equity market as bubble-like, while emphasizing valuation and sentiment extremes.

Implications: Listeners should expect continued leadership from non-U.S. equities, gold as a strategic hedge, and heightened risk that AI-driven U.S. concentration reverses through credit before equities. The pair’s message is to stay diversified and skeptical of crowded narratives.