Some interesting clips from the Financial Analysts Journal Q2, 2018 article entitled: Everybody’s Doing It: Short Volatility Strategies and Shadow Financial Insurers.

A defining characteristic of most short volatility-contingent strategies is that their hedging and unwinding trades are destabilizing. When traders hedge and unwind in response to changing market conditions, their trades tend to accelerate those changes, which results in more unwinding trades. This positive feedback may prove to be dangerous. A severe market environment could expose everyone to rapid and destructive bouts of unwinding. The short volatility ecosystem shows the classic property of a complex system: “the possible occurrence of coherent large-scale collective behaviors with a very rich structure, resulting from the repeated nonlinear interactions among its constituents: The whole turns out to be much more than the sum of its parts” (Sornette 2003, 11).

Could a Volatility Cascade Lead to a Correlated Asset Market Crash?

The possibility that the participants we have discussed might act in concert is alarming because uncoordinated but correlated behavior could trigger a significant volatility event. Rises in implied volatility would likely cause many of those traders to sell securities to adjust their hedges. Such selling would increase implied volatility, which would lead to more asset sales. A crash would occur if this feedback loop exhausted the normal liquidity that stabilizes markets. The following are some issues that should elevate concerns.

The level of assets under management in volatility contingent strategies is large. Adding implicit volatility sellers, such as risk-parity funds (estimated at $500 billion), volatility-targeting funds ($350 billion), risk premium harvesting funds ($300 billion), and trend followers ($300 billion), to explicit sellers, such as pension overwriting funds ($50 billion), dedicated option funds ($10 billion), ETPs (exchange-traded products) and ETNs ($3 billion), and VIX-related strategies ($3 billion), yields a total invested in short volatilitycontingent strategies of over $1.5 trillion. Recent estimates by credible market sources suggest that this sum is large enough to present a significant risk should these parties all trade in the same direction. These estimates are imprecise; they may be off by a factor of 2. But they show that exposure is extremely large and that a potential simultaneous deleveraging by many investors could be a significant event that would challenge the liquidity of the markets.

The assets in short volatility-contingent strategies continue to grow. Low levels of realized volatility make low levels of implied volatility appear reasonable and allow managers to justify selling volatility at low prices. Low yield levels also drive growth in short volatility strategies as managers seek to meet yield targets. Low realized volatility also makes long volatility positions less attractive for delta hedgers who might otherwise offset volatility selling.

Confidence in selling volatility continues to grow.  Managers base their volatility-selling activities on backtests and on many academic studies that show that selling volatility is a positive-expected-return activity. Such research provides cover to managers engaged in risky risk premium harvesting strategies. The absence of recent significant volatility also emboldens investors. Many investors now undoubtedly confuse low volatility with low tail risk, despite that fact that different factors determine the middle and tails of the return distribution.

Asset class diversification has broadened the scope of volatility selling. Volatility-selling strategies are now widespread across asset classes as short volatility investors seek diversification. As a result, implied volatility has collapsed across all assets. If we think of selling financial insurance as a shadow insurance operation, then just as in a multiline insurance company, diversifying across different lines of insurance business makes rational economic sense.

All short volatility strategies are similar.  Regardless of investment horizon, the inverse of volatility is the main factor driving the dynamic portfolio rebalancing associated with short volatility strategies. In a volatility shock, with one small exception that we will discuss later, each strategy will respond in the same direction, so that the response of the whole will be larger than the sum of the parts and larger than most participants would expect based on analyses of only their own strategies. This self-similarity is an important factor in inducing endogenous longrange correlations between participants at different time horizons.

Investors are not generally aware of the extent to which their strategies are correlated. Each participant believes that he or she has some edge or specific mechanism to control downside risk. However, the success of these strategies depends on the liquidity available to them. Traders who do not recognize that they will compete for liquidity with investors who are trading strategies that are seemingly different but essentially the same will oversize their positions.

Mechanized trading is common. Many managers use machine-driven algorithms to implement their volatility-trading strategies automatically. The mechanization ensures that reactions to market moves will be tightly coupled, quick, and price insensitive—three properties that greatly increase the probability and severity of market crashes.

Participants continue to sell volatility despite declining prices and obvious risks. Several reasons may explain the persistence of short volatility strategies: low yields elsewhere (substitution), the need for relative performance compared with peers (herding), increasing expected returns as the futures term structure steepens, and a belief that the economy somehow is now different than it was before.