Jim O'Shaughnessy hosts Cliff Asness, co-founder and Chief Investment Officer of AQR Capital Management, one of the world's largest quantitative investment firms. The conversation explores the emotional reality of investing, the evolution of quantitative strategies, and critiques of popular investment trends.

The discussion covers AQR's experience through major market crises including the dot-com bubble, global financial crisis, and the 2018-2020 challenging period. Asness shares insights on how quantitative firms adapt to changing markets while maintaining disciplined approaches to factor investing and risk management.

The Emotional Reality of Quantitative Investing

Despite being quantitative investors, both Asness and O'Shaughnessy experience prospect theory firsthand - 'good years feel about a third as good as equivalent bad years feel bad' - Asness

During bad periods, quant firms get more aggressive in explaining their strategies, especially when losses are valuation-based rather than fundamental breakdowns

AQR's correlation to basic value strategies is only about 0.2, but during value disasters like the tech bubble and COVID period, they still suffer alongside traditional value investors

Valuation Extremes and Market Manias

The value spreads at the end of 2020 were wider than the dot-com bubble, representing 'the craziest thing in 50, 60 years of good data' - Asness

During the dot-com bubble, O'Shaughnessy wrote 'The Internet Contrarian' in April 1999, predicting 95% of internet companies would be 'carried out feet first or be 90% low'

The GameStop phenomenon and meme stock trading created a 'death cult' mentality where retail investors couldn't distinguish between FanDuel and Robinhood - both just apps for gambling

Asness received significant backlash after calling AMC investors 'crazy' on CNBC, despite AQR holding only 12 basis points short position in the stock

The Private Equity Critique and Volatility Laundering

Asness coined the term 'volatility laundering' to describe how private equity artificially shows low volatility (6.5%) compared to public equities (17%) by simply not marking positions to market

Private equity is 'active levered equity' with beta greater than one, but investors don't have to acknowledge losses until forced sales occur

Pioneering Portfolio Management by David Swensen originally justified private equity through the illiquidity premium, but if illiquidity becomes a 'feature' rather than a 'bug,' expected returns should be lower

The democratization of private equity to 401(k)s is 'a pretty bad idea' because regular investors lack domain expertise and are being misled by mark-to-market practices

Machine Learning Evolution in Quantitative Investing

AQR has moved from requiring roughly 50% story and 50% evidence to approximately 67% evidence and 33% story when incorporating machine learning techniques

Natural language processing now analyzes corporate statements as vectors of numbers rather than simple good/bad word tables, creating better sentiment analysis with some opacity

'If artificial intelligence, machine learning, if I understood every step of what it's doing, what's it adding to my world?' - Asness on accepting necessary opacity in ML

The challenge is that ML improvements make explaining losses during bad periods more difficult compared to traditional factor-based strategies

Hedge Fund Reality and Beta in Alternatives

AQR's 2001 paper 'Do Hedge Funds Hedge' found the average hedge fund delivers zero alpha after fees, despite being 0.8 correlated to the S&P 500

Active management is 'an inherently arrogant act' because Sharpe's arithmetic proves the average cannot beat the average, so managers must believe they're above average

Beta-one alternatives make sense for most investors funding out of equities - 'don't think of this investment as a beta of one, think of it as we're not charging you for an index fund' - Asness

AQR originally launched with only 22% volatility products, which was 'theoretically correct and practically disastrously wrong' due to client psychology around large drawdowns

ESG Investing and Stated vs Revealed Preferences

ESG constraints are costly by design - they can only help the world by raising cost of capital for 'bad' companies, meaning ESG investors should expect lower returns

O'Shaughnessy's experience with ESG led him to study stated versus revealed preferences, discovering polling methods that accurately predicted Trump's victory months ahead of traditional pollsters

'You can't tell people you're gonna make them even more money if you insist it's ESG' - investment constraints have a maximum expected value of zero

Academic vs Practitioner Perspectives

O'Shaughnessy wrote What Works on Wall Street to democratize quantitative research, but faced academic criticism despite extensively citing their work in the bibliography

The small firm effect largely disappears when adjusting for market beta and implementation costs, with most outperformance coming from the illiquid first decile

Credentialism cuts both ways - 'experts are on average right' but it's a disaster to assume they're always right or always wrong

Fooled by Randomness by Nassim Taleb was 'doing the Lord's work' in helping people understand randomness, despite the author's pretentious writing style