AQR’s Antti Ilmanen – US Exceptionalism: Growth Story or Valuation Trap? | #607

Episode Summary

Executive Summary: Meb Faber and AQR’s Antti Ilmanen discuss a series of papers on expected returns, arguing that investors are overly influenced by the “rearview mirror.” They contrast objective, yield-based return estimates with subjective, extrapolative expectations, warn that U.S. equities are priced for exceptional growth, and emphasize diversification—especially liquid alternatives and trend strategies—over binary market timing.

Main Topics: Rearview mirror vs. forward-looking expectations (Priority: 5/5): The conversation opens with the idea that investors often extrapolate recent winners too far into the future, especially after long U.S. outperformance. Ilmanen contrasts that with objective, valuation/yield-based expected returns that are grounded in history and market starting points. U.S. exceptionalism and valuation richening (Priority: 5/5): Ilmanen argues that U.S. equity strength has been driven partly by real growth advantages, but a larger share of recent outperformance came from valuation expansion. He warns that current U.S. valuations imply an overly optimistic growth outlook. Valuations vs. earnings growth as return drivers (Priority: 5/5): The papers highlight that decade-long equity returns are driven much more by valuation changes than by earnings growth alone. High starting valuations predict lower future returns, though the relationship is noisy and can be distorted by structural regime shifts. Sentiment, analyst forecasts, and extrapolation bias (Priority: 4/5): The discussion shows how bullish sentiment often coincides with high valuations, and how analyst earnings forecasts tend to be optimistic and extrapolative. Ilmanen says analysts can be informative, but incentives and over-optimism reduce their predictive value. Why bonds think differently than equities (Priority: 4/5): Ilmanen argues bond investors are more contrarian because they quote and think in yield terms, which naturally anchors expectations. Equity investors focus on prices and recent returns, which encourages momentum in beliefs and stronger extrapolation. Portfolio construction, diversification, and illiquids (Priority: 5/5): They discuss why diversifiers have looked unattractive during strong equity markets, but may become more valuable now. Ilmanen favors liquid diversifiers and trend strategies, while criticizing private markets for accounting smoothing and hidden equity beta. Practical implications for the 2020s (Priority: 5/5): The episode ends with a cautious outlook: U.S. and tech expectations are too benign, bond-risk premia are thin, illiquid alternatives look less attractive, and trend-following may be one of the best defenses against a persistent equity bear market.

Key Arguments: Investors systematically over-extrapolate recent returns and growth, especially after extended bull markets, leading to rearview-mirror forecasting errors. Objective expected returns based on yields/valuations are more useful than subjective survey-based expectations for long-horizon forecasting, even though they are noisy. U.S. equity outperformance has reflected both genuine growth advantages and a large valuation rerating; the market now appears to be pricing in too much of the former. Across long history, valuation changes matter more than earnings growth for decade-long equity returns; rich starting valuations usually imply lower future returns. Analyst forecasts are often over-optimistic and become more bullish after good performance, which makes them poor standalone forecasting tools. Bond investors appear more mean-reverting and contrarian because yields are a forward-looking anchor, while equity investors are more prone to price-chasing. Diversifiers should not be judged as standalone line items; their true value emerges at the portfolio level, especially when combined with equity beta efficiently. Private equity and other illiquid alternatives often exhibit smoothed returns and hidden market beta, which can create the illusion of diversification. Trend-following remains one of the strongest historical defenses against deep, persistent bear markets and may also capture outlier upside in inflationary or regime-shift environments. The most prudent takeaway is not an all-out market exit, but lower return expectations, greater humility, and more emphasis on diversification and risk management.

Data Points: U.S. vs. non-U.S. CAPE valuation ratio: ~1.8x to 2.0x - Ilmanen says U.S. equities are currently almost twice as expensive as non-U.S. markets on a relative CAPE basis. Post-GFC U.S. earnings growth edge: ~3% per year - He notes many investors extrapolate the post-GFC period, when U.S. EPS growth looked about 3% higher than the rest of the world. Long-run U.S. earnings growth edge: ~1% per year - Over the long run, U.S. earnings per share growth has exceeded the rest of the world by about 1% annually. Correlation: inverse CAPE vs. future returns: ~0.5 correlation / ~26% R-squared - Used to argue valuations have predictive power over long horizons, though the signal is noisy. Correlation: last decade returns vs. next decade returns: ~-0.3 correlation - Illustrates rearview-mirror extrapolation is negatively informative over long horizons. Correlation: 10-year real returns vs. 10-year valuation changes: ~0.9 correlation - Shows valuation changes are the dominant driver of decade-long equity returns. CAPE ratio in the U.S.: Around 40 - Ilmanen cites current U.S. CAPE levels as extreme by historical standards. Historical U.S. CAPE average shift: ~15 to 25+ - He says the CAPE regime shifted upward in the 1980s/1990s, complicating mean-reversion forecasts. Relative return impact from valuation richening: ~7 percentage points per year - A cited example of how valuation rerating drove the gap between good and bad stock-return decades. Best real equity decades starting CAPE: ~11 - In the reverse-engineering example, the best 10-year real stock returns began with low valuations. Worst real equity decades starting CAPE: ~22 - The worst 10-year real stock returns began with much higher valuations. 60/40 Sharpe ratio: ~1.3 to 1.4 during the 2010s - Used to show how strong traditional 60/40 performance can make diversifiers look unattractive. Long-run 60/40 Sharpe ratio: ~0.4 - Compared to the unusually strong 2010s experience. Forecast relationship in bond markets: Markets/economists often expected rising yields during the 2000-2020 decline - Used to show bond forecasters were often mean-reverting/contrarian but missed the long structural downtrend. EPS growth forecast optimism: Typically 8% to 20% real, or double-digit nominal - Ilmanen says analyst forecasts are persistently too optimistic relative to long-run realized growth.

Pivotal Quotes: "The real trouble with this world of ours is not that it is an unreasonable world, nor even that it is a reasonable one. The commonest kind of trouble is that it is nearly reasonable, but not quite." — Antti Ilmanen (via G.K. Chesterton / Peter Bernstein): Used in a footnote to emphasize humility in forecasting markets and the danger of overconfidence in neat models. "We are trying to tell that there is something useful in these things that don't throw them away, especially if throwing them away means that you are just going to do even more this rear view mirror of the last decade or so." — Antti Ilmanen: Explains why valuation/yield signals remain useful despite their limitations and recent underperformance. "I do think that the rear view mirror looks much too benign for overall market, especially for US and maybe tech." — Antti Ilmanen: Summarizes his caution on current U.S. equity and tech valuations going into the rest of the 2020s.

Implications: Listeners should expect lower long-term returns than recent U.S. experience suggests, avoid binary market timing, and focus on diversification, especially liquid alternatives and trend strategies, as protection against valuation mean reversion and regime shifts.