Patrick O'Shaughnessy on building OSAM, Canvas, and Invest Like the Best

Executive overview

O'Shaughnessy Asset Management grew from a 1987 quant research firm into a software business that delivers customised portfolios through a platform called Canvas. Most asset managers sell a fixed menu of strategies; OSAM lets advisors design bespoke portfolios for each end client, then handles all trading and reporting. The same curiosity-driven openness that built a podcast empire also became the firm's go-to-market engine.

Consistent, radically public learning compounds into durable competitive advantage across both investing and media.

From quant research to asset management

• OSAM's origin: Jim O'Shaughnessy found that stupidly simple rules (e.g. Dogs of the Dow) consistently beat the market when applied with iron discipline.
• Core insight: it was never value investing per se — discipline and buying cheap were the two pillars.
• Pioneering data work involved hand-collected microfiche going back to the 1920s; differentiated data beats sophisticated modelling.
• The firm passed through Bear Stearns (via acquisition), survived the 2008 crisis by spinning out in July 2007 — what they called OSAM Independence Day.
• Netfolio (late 1990s) was a proto-robo-advisor: the right idea, too early.

The Canvas platform

• OSAM rebuilt its internal quant infrastructure over ten years; Canvas is that infrastructure offered as a service to financial advisors.
• Advisors configure bespoke strategies for each client — tax preferences, sector exclusions, concentration limits — rather than choosing from a fixed menu.
• Analogy: Shopify for investment strategies, with advisors as merchants and OSAM as the platform.
• Built in three months because the underlying infrastructure already existed; five engineers produced a product that looks like the work of thirty.
• Pricing is dynamic: the more the portfolio deviates from a vanilla index, the higher the fee — up to a ceiling still below most long-only manager rates.
• Counter-positioned against the quant hedge funds that could theoretically replicate it: those firms monetise proprietary research and would never open-source their edge.

Invest Like the Best — origin and growth

• Started in 2016 as Patrick's personal search for applicable business ideas after reaching the limits of his quant research ability.
• Named after his father's book: extract portable lessons from the best practitioners and make them accessible.
• Episode one was an audio companion to a friend's book; episode two with Michael Mauboussin had product-market fit immediately.
• Never marketed beyond a tweet; growth was organic with step-function jumps after high-signal guests (the "Mauboussin bounce" ≈ 10% permanent audience increase per appearance).
• Key rule: curiosity drives the guest list, never business objectives — the audience detects inauthenticity instantly.
• The show became OSAM's best marketing asset without ever being designed as one.

Seven Powers applied to OSAM and Invest Like the Best

• Brand: high quality, low variance output over time — the barrier is simply time, which cannot be bought.
• Counter-positioning: radical transparency in research is the opposite of the "logo + info@ email" digital velvet rope most elite investors use; positive-sum venture fund extends this by externalising the diligence process publicly.
• Cornered resource: the Research Partners programme gives independent researchers free access to OSAM's full data library in exchange for IP ownership of what they produce, attracting people like the anonymous analyst "Jesse Livermore."
• Network flywheel (not a formal power but real): better guests grow the audience; a larger audience attracts better guests; the flywheel accelerates and cannot be bought.

Playbook lessons from podcast guests

• Chetan Pudigunta: go slow to go fast in enterprise software — mature the product with a small early cohort before scaling; fragile systems need to evolve, not be airdropped.
• Bill Gurley: evaluate marketplaces by plotting supplier penetration against consumer benefit — if the curve flattens early, it is not a platform business.
• Josh Wolf: follow the directional arrow of progress and jump to its logical endpoint; Robinhood did this with commission costs going to zero.
• Katrina Lake: legacy e-commerce competed on speed, price, convenience; the future is personalisation.
• Charlie Songhurst: find talent in uncompetitive markets — Silicon Valley is the worst place to look.
• Matt Ball: a metaverse requires interoperability, not walled gardens — the value is in the portable layer.

Growth without goals

• Patrick explicitly rejects five- and ten-year goals; he uses "boss battles" — near-term, concrete objectives — instead.
• The single thing he actually does: ask better questions of the right people; everything else (Canvas, the fund, the podcast) is a byproduct.
• Managing multiple concurrent initiatives works because they are all the same activity expressed in different formats — scouting and routing knowledge, not originating it.
• Solve problems with technology rather than people; build systems that let you do more of the one thing you are good at.