Thoughts on Safe Haven by Spitznagel?

Hi all,

Recently read this book and found it super interesting as well as quite controversial in terms of modern finance theory, specifically Modern Portfolio Theory (MPT). I believe the quote from Taleb was "This is his monumental f*** you to the investment industry". 

I will try to give a brief summary of the main message for those who have not read it (excuse any inaccuracies and feel free to correct/ add in the comments).

Backdrop

Spitznagel runs Universa Investments, which is a tail-risk hedge fund. They offer what they call "explosive" tail-risk insurance. In short, during extreme outlier events where the market (S&P500 used as "market" in the book) is down a lot, they return an absurd amount. I think there is a quote from a WSJ article where they state a 20% downturn would yield between 4000-7000% returns on invested capital. The rest of the time, they lose most of their invested capital -- "bleeding" is the term used. Importantly, they do not try to predict the market, as the outlier events they are targeting are so-called "Black Swans", i.e. unknown unknowns. Their proposition is then that from a portfolio perspective, a small allocation (low single digits) of capital to such a strategy is optimal risk-mitigation. 

The central idea

Spitznagel states: "People of risk mitigation as a liability, as a trade-off against wealth creation, because it usually is. ... Risk mitigation doesn't have to be viewed that way. Risk mitigation can and should be thought of as being additive to portfolios over time -- with the right risk mitigation, that is." He lays out the case for the case for an insurance-based strategy similar to the one implemented at Universa versus other common risk-mitigation strategies (60/40 portfolio, gold, etc.). He argues that you pay an "arithmetic cost" of 0.2% (i.e. lowering the arithmetic average of returns by 0.2% versus the S&P) for a "geometric net" of 0.5% (i.e. raising CAGR by 0.5% versus the S&P), whilst the other strategies lower both arithmetic and geometric returns. He then defines a strategy to be cost-effective when its net portfolio effect (i.e. effect on CAGR) is positive. 

From what I understand, his key contention with modern finance theory is that diversification is not "the only free-lunch", that you are essentially just diluting risk, not mitigating it. From MPT we would expect that any change in strategy that lowers volatility more than average returns, thus raising the Sharpe ratio, is good.  Spitznagel argues that in extreme outlier events, correlations break down as people rush to the exits. The effect of diversification then breaks down. Also, as investors often apply leverage to low-return high-Sharpe strategies, the effect of the breakdown in correlation is then amplified. 

Thoughts?

There are so many details and ideas not covered here, but I would be very interested to hear from practitioners on this forum regarding the ideas presented in this book. I am very new to the industry, and found it quite shocking as it opposed much of what you learn in basic FE courses.

Thanks.