Professor Brad Cornell: A Skeptic's Look at the Cross Section of Expected Returns (EP.151)

Episode Summary

Executive Summary: Brad Cornell challenged mainstream factor investing, arguing markets are non-stationary and many “premiums” are just characteristics that can disappear or reverse. He was skeptical of past-performance-based fund selection, emphasized DCF and present-value logic over factor labels, warned about narrative-driven bubbles in Tesla/EVs and ESG, and said investors should use lower, implied forward return assumptions rather than historical averages.

Main Topics: Characteristics vs. risk factors: Cornell distinguished observable stock traits from true priced risk factors, arguing that many academic/practitioner strategies confuse a correlation with compensation for systematic risk. Non-stationarity and skepticism toward historical regularities: He argued that financial markets change over time, so patterns found in one era may not persist; this undermines confidence in cross-sectional return studies and factor models. Bubbles, narratives, and the “big market delusion”: Cornell attributed dramatic price runs in sectors like EVs and AI to sentiment, narratives, and competitive underestimation rather than fundamentals alone. Fund manager selection and past performance: He said recent winners are often the least attractive forward-looking bets because performance is often strategy-driven and mean reversion can punish recent outperformance. Value investing, DCF, and profitability: He rejected simplistic valuation ratios as outdated, preferring price-to-estimated-value via DCF, while viewing profitability as useful input for forecasting margins. ESG, climate policy, and expected returns: Cornell argued ESG is hard to define, may be more “sounding good” than doing good, and that investors accept lower expected returns when they bid up green assets. Equity risk premium and forward return expectations: He stressed that current prices imply lower future returns, and investors should base planning on implied equity risk premium rather than historical averages.

Key Arguments: Many stock “factors” are just characteristics; only a true risk factor justifies a risk premium under asset-pricing theory. Markets are non-stationary, so relationships observed in the past can break or reverse, as Cornell believes happened with value after 2008. Narrative-driven sector booms often ignore competition; even if a market grows, only a few firms may capture most of the value. For concentrated themes like EVs, diversification is safer because investors cannot know which company will win. Recent fund winners may underperform going forward because their outperformance often reflects a strategy that has already become expensive. Past performance reveals strategy exposure more than manager skill. Value investing is not dead, but it should be based on price versus your own estimate of intrinsic value, not on static ratios like price/book. Profitability matters mainly because it improves DCF inputs such as margins. ESG can lower a green firm’s cost of capital, but that implies lower expected returns for investors buying the green asset. Current equity valuations imply lower future returns; planning with historical return assumptions is likely too optimistic.

Data Points: Episode number: 151 - This interview was introduced as episode 151 of the Rational Reminder Podcast. Academic output: 150+ referenced articles - Cornell was described as having published over 150 referenced articles. Books authored: 4 books - He was noted as the author of four books. Value effect turning point: around 2008 - Cornell said the value effect was highly documented until about 2008, then disappeared and reversed. Tesla valuation target discussed: $300+ per share target challenge - Cornell said Tesla struggles to get above $300 even assuming Porsche-like margins and Toyota-like output. Tesla alternative price target referenced: about $3,000 - He referenced Cathie Wood’s updated Tesla price target as an example of bullish narrative-based valuation. SP500 level used in example: 4,150 - He used an S&P 500 level of about 4,150 to illustrate implied forward return calculations. Implied equity risk premium (U.S.): about 4% over Treasuries - Cornell cited Aswath Damodaran-style calculations to estimate the current market’s implied ERP. Implied equity return example: about 5.5% - He said 4% over Treasuries implies an equity return near 5.5%. Low ERP scenario: 3% - He said if the ERP fell to 3%, the S&P 500 could rise to about 5,600. High ERP scenario: 6% - He said a reversion to a 6% ERP would imply the S&P 500 near 2,600. Tesla implied ERP example: about 1.5% - He said Tesla can be rationalized only with an extremely low cost of capital / ERP around 1.5%. Blue/red marble example: 60% blue / 40% red - Used to illustrate learning probabilities only if the underlying process is stationary. Pi digits anecdote: 768 digits / six nines - He recounted Feynman memorizing the first 768 digits of pi and noticing six consecutive 9s as an example of false pattern detection.

Pivotal Quotes: "financial economics is not physics" — Brad Cornell: He used this to stress uncertainty, non-stationarity, and the limits of treating empirical patterns as laws. "I don't believe value investing is dead, but it should be interpreted properly. It's price to value." — Brad Cornell: He contrasted intrinsic-value-based valuation with simplistic factor-style ratios like price/book or price/earnings. "If you're not planning for that, you're not being rational. You're praying for a miracle." — Brad Cornell: He said investors must plan for lower implied future returns rather than assuming historical performance will repeat.

Implications: Listeners should be cautious about factor claims, ESG narratives, and recent performance chasing. Cornell’s framework favors DCF, diversification, and forward-looking return assumptions over historical averages or catchy stories.