Patrick O'Shaughnessy: great CEOs become capital allocators — Tony Xu is one of the greats emerging
Sep 30, 2025 with Patrick O'Shaughnessy
Key Points
- Patrick O'Shaughnessy argues that great CEOs master capital allocation over decades, citing Tony Xu's Singleton Award as evidence Xu will deploy DoorDash cash into aggressive buybacks when stock is undervalued.
- Apple's $1 trillion shareholder return through buybacks over a decade ranks as the single best investment by alpha basis, setting the template for how Xu should manage DoorDash's capital once cash generation accelerates.
- SaaS incumbents under AI pressure face a business model problem, not a technology problem, making transformation harder than switching technology stacks, per Brett Taylor of Sierra.
Summary
Patrick O'Shaughnessy argues that the defining skill separating great CEOs over decades isn't product intuition — it's capital allocation. The catalyst for the conversation is Tony Xu receiving the Singleton Award, named after Henry Singleton of Teledyne, whom Buffett credited as a formative influence. Singleton's playbook: spend the first 15–20 years acquiring roughly 200 businesses, then spend the final 30 years buying back approximately 95% of his stock — producing one of the best shareholder return records in market history.
O'Shaughnessy's read is that Xu is following a similar trajectory. DoorDash turned cash flow positive several years ago, and O'Shaughnessy believes Xu will deploy that capital aggressively into buybacks when he judges the stock undervalued. The Apple comparison anchors the argument: O'Shaughnessy says quantitative research identified Apple's buyback program as the single best investment by dollars-returned-on-an-alpha basis, with Apple returning over $1 trillion to shareholders in roughly a decade.
The broader point is structural. Once a company starts generating cash, the CEO's job shifts from operator to allocator — and almost nobody is trained for it. The contrast O'Shaughnessy draws is Larry Ellison at Oracle versus the dilution patterns at companies like Salesforce. Aggressive buybacks signal conviction about intrinsic value, and historically that signal has been reliable.
The conversation turns briefly to the SaaS AI transition. Mark Leonard of Constellation Software reportedly signaled he was turning bearish on vertical SaaS shortly before stepping down. Barry Diller, whom O'Shaughnessy spoke with recently, said he wouldn't try to ride the AI wave — too different, too hard, better left to others. Brett Taylor of Sierra is quoted as saying it's easier to change your technology stack than your business model, which frames the core challenge: the SaaS incumbents under pressure aren't facing a technology problem so much as a business model problem, and that's harder to solve.
O'Shaughnessy's bottom line on Xu is unambiguous — he calls him one of the greats emerging, and says the market is still pricing DoorDash as a food delivery app while Xu is building something with exponential rather than linear trajectory.