Safe Haven

Spitznagel runs Universa Investments, a fund that sells tail risk protection, and this book states the mathematical argument behind that business.

The central claim rests on the difference between arithmetic and geometric returns. A portfolio's compound growth rate is not the average of its annual returns but the geometric mean, which is reduced by volatility. A large loss requires a disproportionately large gain to recover, so avoiding large losses raises compound growth even if the avoidance costs something in most periods. Spitznagel therefore argues that a hedge should be evaluated by whether it raises the portfolio's compound growth rate, not by whether it has a positive expected return on its own or by whether it reduces measured volatility.

From this he attacks the conventional safe havens. Gold, bonds, diversifying assets and low-volatility strategies reduce risk mainly by reducing exposure to the assets that generate returns, so they lower compound growth as well as volatility. He calls this cost-ineffective mitigation. What he wants instead is an instrument that pays out very large multiples in the specific states where the portfolio suffers most, which costs little in normal periods precisely because those states are rare.

He develops the argument with simulations, with a treatment of the Kelly criterion and the St Petersburg problem, and with Austrian economic arguments about the distortion of capital markets by monetary policy.

The obvious objection is that Spitznagel sells the product his argument recommends, and the book does not disclose Universa's returns in a form that permits independent evaluation. Academic work on tail hedging generally finds that options protection is expensive over long periods and that its benefit depends heavily on the entry price and on execution. The mathematics about geometric returns is correct and uncontroversial; the claim that his particular implementation delivers on it is not established here.