The Oracle, Decoded: How They Reverse-Engineered Warren Buffett

For fifty years, Warren Buffett's record looked like magic — a once-in-a-century gift you either have or you don't. Then three quantitative researchers took that genius apart, piece by piece, and found that most of it could be copied. The part that couldn't is the part almost nobody wants.

This episode walks backstage on the famous "Buffett's Alpha" paper to ask the rudest question in investing: can the Oracle be reverse-engineered? The answer is humbling — and it changes what "be like Buffett" should mean for you.

In this episode:

The scoreboard everyone misreads. A dollar invested with Buffett in 1976 became more than $3,685 by 2017 — the best risk-adjusted record of any stock or fund that survived the stretch. The twist: he didn't take more risk. His market sensitivity (beta) was just 0.69. He got rich by being more efficient, not braver.

The trick explained. The standard four-factor model captured Buffett's style — cheap, big, steady, no chasing hot stocks — but left a giant pile of return unexplained. Add two more factors, "betting against beta" and "quality minus junk," and the magic shrinks to statistically indistinguishable from zero. Neither luck nor magic, the authors wrote, but a reward for leveraging cheap, safe, high-quality stocks.

The quiet half of the balance sheet. The part nobody discusses: financing. Roughly 1.6-to-1.7 leverage, funded by insurance float at about 1.72% a year — below what the US government paid to borrow — plus deferred taxes and the options he sold to investors who couldn't borrow. He wasn't just avoiding the crowd's mistake; he was getting paid by it.

The snow already skied. Those factors are now sold in low-cost funds. The edges crowded, Berkshire's size turned into an anchor, and a market carried by a handful of AI names is sprinting past exactly the strategy Buffett built to ignore. Even the greatest compounding machine ever built is sliding back toward the market's ordinary pace.

The strongest bear case — answered. If it's all factors and cheap leverage, was it ever genius, or just survivorship bias with a lab coat? Two answers: the "clone" only worked as a frictionless backtest, and the residual the equation couldn't explain — temperament — is the one thing survivorship bias can't manufacture. He sat through a 44% drawdown without flinching, and bought when everyone else was selling in 2008.

What it means for you. You can rent a rough version of the factor exposures today, for not very much. What you cannot rent is the float, the structure, or the temperament to do something boring, correctly, for half a century while everyone around you chases something shinier. That was never the part that looked like magic. It turns out it was the only real magic there ever was.

Read the written version at quietvelocity1.substack.com, the companion Substack to Conviction Bet.

New episodes weekly. Subscribe on Apple Podcasts, Spotify, YouTube, or Amazon Music.

Conviction Bet is independent investment commentary. Nothing in this episode is investment advice.