Buffett and Li Lu's Sortino ratios differ by 0.002. Every other number about them doesn't

Warren Buffett's published Sortino ratio is 0.3967. Li Lu's is 0.3948. That is a gap of 0.0019, on a figure Tenachine has never applied to investor data before this guide, only to individual backtest tickers. Two managers whose disclosed records have disagreed on nearly every other measure this site has checked land almost exactly together on this one.

Where these two normally split apart

Tenachine's guide on Buffett's Mastercard trade already runs the Kelly criterion on both investors and gets 65.8 percent for Buffett against 62.0 percent for Li Lu, close in that case but for a different reason: both payoff ratios happen to be large. Tenachine's guide on median holding period and return skew finds real separation instead, Buffett's median hold of 3.125 years against Li Lu's 2.625, and a skew of 4.78 against 1.22, more than three times apart. Tenachine's guide comparing expectancy does not compare Buffett and Li Lu directly, but its own published expectancy figures, 14.14 percent for Buffett against 40.61 percent for Li Lu, are nearly three times apart. On almost every dimension this site has checked, Buffett and Li Lu read as two different kinds of investor. The Sortino ratio is the one place they do not.

What Sortino counts that the other numbers do not

Sortino ratio divides a return figure by downside deviation, the spread of only the losing outcomes, rather than the full spread of every outcome the way a Sharpe ratio does. Tenachine's guide on DHR's Sharpe and Sortino ratios covers this same distinction on backtest data: a ticker whose volatility comes mostly from large winners scores much better on Sortino than on Sharpe, because Sortino stops charging a strategy for swinging up. Buffett's and Li Lu's disclosed records both lean the same way by this measure, most of their spread sitting on the winning side, even though the two records disagree sharply on how often they win, how large the typical winner is, and how long a position sits before it closes. Downside deviation, on its own published figures, is 0.1197 for Buffett and 0.1941 for Li Lu, not identical, but the return figure each investor's downside deviation gets divided into moves enough in the same direction to land the final ratio this close.

Burry's number, and why it does not match either

Michael Burry's Sortino ratio is 0.7190, roughly 1.8 times Buffett's and Li Lu's. His downside deviation, 0.2548, is the largest of the three, which on its own would suggest a worse ratio, not a better one. What pulls Burry's Sortino above both other investors is the return figure sitting on top of that downside deviation: Tenachine's guide on Burry's median IRR already shows his typical position compounding at 18.3 percent annualized, ahead of all but one of 76 tracked investors, well ahead of Buffett's 4.7 percent and Li Lu's 7.7 percent on the same measure. A bigger downside deviation did not cost Burry on this ratio, because the return in the numerator grew by more.

None of this ranks the three investors against each other, and a close Sortino ratio between two of them is not evidence their strategies work the same way; the guides above already show they do not, on nearly every other measure checked. It is evidence that one specific ratio, computed the same way Tenachine computes it on backtest tickers, happens to land two very different disclosed records within two thousandths of each other. Tenachine does not publish the exact downside-deviation window or return basis behind this figure beyond the published inputs above, so this is a description of what the published numbers show, not a claim about why the two happened to converge. Past performance does not predict future results, for any of the three investors covered here, and a ratio this close today carries no promise of staying close in the next disclosed quarter.