Patrick O'Shaughnessy – O'Shaughnessy Asset Management (First Meeting, EP.01)

Episode Summary

Executive Summary: Ted Seides interviews Patrick O'Shaughnessy about OSAM’s evolution from a rules-based quant shop into a research-driven investment platform. They discuss the firm’s core factors, why momentum, value, quality avoidance, and shareholder yield work, how machine learning and data are changing research, why quants are reshaping markets, and how Patrick’s podcasting and public learning loop have influenced business building, manager selection, and external partnerships.

Main Topics: Origins of OSAM and Patrick’s path into investing (Priority: 5/5): Patrick describes growing up in a curiosity-driven household, stumbling into OSAM through timing and an internship, and falling in love with the research side of the business. OSAM’s core quantitative framework (Priority: 5/5): The firm’s portfolio construction is built around four core concepts tied to sources of equity return: valuation, momentum, quality avoidance, and shareholder yield. Why factors work and how they behave over time (Priority: 5/5): Patrick explains the economics behind the factors, especially the difference between valuation’s long runway and momentum’s short, fast alpha decay. Research process, machine learning, and the role of labels (Priority: 5/5): OSAM’s research is split into incremental factor improvement, hypothesis-driven new signals, and machine-learning work focused more on data construction and prediction of stationary variables than direct return forecasting. How quants affect the market and how OSAM differentiates (Priority: 4/5): He argues quants are steadily commoditizing many traditional active strategies, but OSAM differs through non-linear modeling, tail-focused portfolio construction, low benchmark overlap, and a direct alpha objective rather than tracking-error optimization. Podcasting as a learning loop and business builder (Priority: 4/5): Patrick frames learning, building, sharing, and repeating as a compounding loop that powers both his podcast and OSAM’s research culture, while also generating ideas and relationships. Investing in external managers and private markets (Priority: 4/5): Patrick discusses using the podcast as a discovery mechanism for managers and notes early, promising but still limited applications of quant methods in private investing.

Key Arguments: Patrick’s upbringing emphasized curiosity and self-reliance rather than explicit investing instruction, which shaped his research mindset. OSAM’s premise is to extract signal from data and build disciplined models; process matters more than ad hoc decision-making, especially in difficult markets. The four main return sources in equities are business growth, multiple expansion, and return of capital; OSAM’s factors map to these sources. Value works by buying depressed businesses where multiple expansion can occur over a long horizon, while momentum works because strong recent price performance predicts near-term operational growth. Quality is most useful as a negative screen: avoiding the worst balance sheets, accounting, and capital allocators matters more than buying the highest-quality companies outright. Momentum has a much shorter alpha half-life than value; it pays off quickly and then reverses, so turnover and timing discipline are essential. OSAM’s research agenda has three silos: improving existing factors, testing new hypotheses, and using ML to build better data sets and more targeted predictive models. Machine learning is less effective at predicting raw returns because returns are non-stationary; it is more useful for forecasting specific components like dividend cuts or earnings growth. Most public quants optimize for information ratio and tracking error; OSAM instead targets excess return directly and uses non-linear, tail-focused portfolio construction. The firm’s biggest competitive edge is not just better models but better questions, better data work, and a willingness to let the tails drive decisions. Quants are increasingly eating away at strategies that can be replicated from data; discretionary managers retain an edge in deep fundamental analysis and concentrated portfolio construction. The podcast functions as a curiosity machine, a networking engine, and a structured way to learn, build, and share ideas in public. Sharing research can improve it through outside feedback, especially when the source of edge is behavioral rather than purely informational. OSAM has applied software/platform thinking to asset management through tools, research partnerships, and opening parts of its architecture to external researchers. The firm sees promise in quant approaches to private markets, but believes relationship-building, deal structure, and domain expertise still matter enormously. A good quant manager should have a large graveyard of failed ideas and obsessive data hygiene; the quality of their failures and data work reveal their research strength.

Data Points: OSAM holding period: 18–24 months - Patrick says the firm is not a short-term stat-arb shop and typically holds equities for a long time. Value factor payoff window: up to 10 years - He says value can deliver marginal monthly alpha for many years after entry. Momentum alpha window: first 1 year - Momentum alpha is earned quickly and then tends to reverse. Worst decile avoidance: bottom 10% - OSAM removes the worst decile of certain factors and quality measures from consideration. Typical US large-cap value portfolio size: about 60 stocks - He says this is at the low end for the firm’s strategies. Largest portfolio size mentioned: mid-100s stocks - A different strategy may hold around the mid-100s names. Research partners: 7 people - Patrick says the firm has seven research partners in the program. Public podcast scale at launch: 571 listeners in the first week - He recalls the early response to the first Invest Like the Best recording. Podcast episode count mentioned: roughly 130 episodes - He says one out of every 10 or so guests became a business partner or close connection. Quantitative study on data cleaning: 3 years - OSAM spent three years cleaning and scrubbing its primary data set. Ownership data use: maintained as a live production process - Even when a study does not produce immediate alpha, OSAM preserves the data set for future use. Sharing effect on inbound content: more inbound books and guests - Patrick says sharing his book club and podcast created a strong inbound feedback loop.

Pivotal Quotes: "Look it up." — Patrick O'Shaughnessy: Describing his parents’ approach to encouraging curiosity and self-teaching. "Past glories are poor feeding." — Patrick O'Shaughnessy: A quote from Isaac Asimov that he uses to emphasize continuous learning and improvement. "learn build share repeat" — Patrick O'Shaughnessy: OSAM’s operating mantra and the framework he uses to describe his learning loop.

Implications: For investors, the episode shows that durable alpha comes from disciplined process, better data, and clear economic intuition—not generic factor mimicry. It also suggests quants will keep expanding into more of active management, while human judgment still matters most in concentrated, domain-rich, relationship-driven opportunities.