TIP 063 : Quant Investing with Patrick O'Shaughnessy (Business Podcast)

Episode Summary

Executive Summary: This episode explores quantitative value investing through the lens of Patrick O'Shaughnessy and Toby Carlisle. They explain how systematic investing blends value, momentum, and buyback signals, why backtests must be treated skeptically, and why human behavior—not pure risk—likely explains factor premia. The discussion also covers leverage, interest rates, Amazon’s valuation, and practical limits of macro forecasting.

Main Topics: Patrick O'Shaughnessy's background and path into quant investing (Priority: 5/5): Patrick explains his philosophy background, self-taught finance skills, CFA work, and early career entry during the 2007-09 crisis, showing how a nontraditional path led into systematic investing. Share buybacks, conviction, and debt (Priority: 5/5): The hosts debate whether buybacks destroy value or create it. Patrick distinguishes low-conviction repurchases from large, undervalued buybacks, and notes that leverage is harmful only at extremes. Backtesting pitfalls and research discipline (Priority: 5/5): Patrick warns that most backtests fail due to data issues, survivorship bias, and overfitting. He argues backtests are useful but often create false confidence and must be paired with robust intuition and out-of-sample testing. Value, momentum, and factor blending (Priority: 5/5): The conversation explains why value and momentum can complement each other despite seeming opposite. Patrick cites a 70/30 value-momentum blend as strong in sample and emphasizes that mixing factors reduces drawdowns. Behavioral explanation of factor investing (Priority: 4/5): Patrick argues value works mainly because investors overreact and overextrapolate trends, causing mean reversion. He frames factor investing as a story about human psychology, not just compensation for risk. Amazon, growth stocks, and valuation discipline (Priority: 4/5): The hosts use Amazon as an example of a high-priced growth stock whose valuation depends on future profitability and margin expansion. Patrick treats it as a lottery-ticket style stock rather than a core value opportunity. Interest rates, macro uncertainty, and portfolio diversification (Priority: 4/5): The final section discusses the difficulty of predicting Fed policy, the effects of ultra-low rates, and why diversified factor exposure is preferable to macro timing. Patrick notes value has historically performed well in rising-rate regimes.

Key Arguments: Quant investing is not necessarily high-frequency math; it can mean systematizing sound fundamental ideas into repeatable rules. A philosophy background can be an advantage because investing rewards reasoning about human behavior and incentives. CFA study helped bridge technical gaps, but most of the credential reflects disciplined effort rather than retained theory. Buybacks are not uniformly bad; their effect depends on whether management repurchases stock at a discount or premium to intrinsic value. High-conviction buybacks tend to be associated with stronger performance than low-conviction repurchases. Companies with extreme leverage tend to perform poorly, but the middle of the leverage distribution often outperforms both zero-debt and highly levered firms. Backtests are frequently misleading because of survivorship bias, data errors, p-hacking, and weak statistical significance. Value and momentum work best together because they capture different time horizons and often outperform in different market regimes. Momentum is strongest in established bull trends and can suffer severe factor crashes at major market bottoms. Value’s edge is likely behavioral: markets overreact to bad news and underreact to modest improvements. Pure quality screens are less powerful than many investors assume; avoiding the worst companies is more useful than chasing the best-quality names. Macro forecasting is difficult and often politicized; investors are better off focusing on robust, diversified strategies. Rising interest-rate environments have historically favored value stocks, while low-rate environments have helped momentum and aggressive leverage strategies.

Data Points: Buyback outperformance: 3.3% - Referenced as the excess performance of companies that buy back shares most aggressively in Patrick's research. High-conviction buybacks: 10% to 30% of shares outstanding - Patrick describes large repurchases as a strong signal of management conviction. Hold period: 5+ years - Patrick says some systematic portfolios hold positions for five years or more. CFA charter: CFA - Patrick notes he completed the CFA program and holds the charter. Peak buybacks: Early 2008 - Patrick says aggregate buyback dollar values peaked in early 2008, a poor time for repurchases. Value outperformance in rising-rate periods: 14 of 17 periods - Patrick cites historical rising-rate regimes where cheap stocks beat the market. Value outperformance in rising-rate periods: ~4% annualized - Average outperformance of value during the 17 rising-rate periods Patrick discussed. Momentum/value blend: 70% value / 30% momentum - A sample-optimal long-only blend based on Sharpe ratio in Patrick's testing. Momentum lookback window: 3 to 9 months - Patrick says strict price momentum over this window is the most predictive. Momentum crash period: Post-bear-market bottoms - He highlights severe momentum factor crashes after major market lows such as March 2009. Amazon earnings growth required: ~55% per year - Preston estimates the growth Amazon would need over 10 years to justify its valuation. Probability of such growth: 0.3% of the time - Patrick says this level of earnings growth historically occurs only about 0.3% of the time. Book club size: ~5,000 people - Patrick's free book recommendation email list has grown to about 5,000 subscribers. Books read per year: 80-100 - Patrick says he currently reads around 80 to 100 books annually. Books read per year previously: 150 - He says before having children he read about 150 books per year. IDC benefit estimate for Vanta: $535,000 per year - Sponsor ad copy mentions estimated annual benefits from Vanta customers. Vanta customer count: >10,000 companies - Sponsor ad copy cites the number of global companies using Vanta. Public.com transfer bonus: 1% uncapped - Sponsor ad copy offers a 1% bonus when transferring a portfolio.

Pivotal Quotes: "There's a lot of things that fall under the quant umbrella or descriptor... The better way to describe what we do is a sound investing strategy that's been systematized." — Patrick O'Shaughnessy: Explaining that quant investing is broader than high-frequency trading or heavy math. "The market is a discounting mechanism. It's effectively making predictions about each individual stock's future." — Patrick O'Shaughnessy: Describing why value works through behavioral mispricing and mean reversion. "No one's ever seen a bad backtest." — Patrick O'Shaughnessy: A cautionary remark about the tendency for backtests to look better than reality.

Implications: Listeners should view quant investing as disciplined factor selection, not prediction or magic. The episode argues for diversification across styles, skepticism toward backtests, and patience through factor cycles, especially when value or momentum temporarily fall out of favor.