The idea that returns from financial assets are normally distributed underpins many traditional financial theories, but the reality is that many (even most) assets do not conform to this. Instead, empirical distributions exhibit higher peaks and fatter tails—returns are mostly clustered within a small range around the mean but extreme moves occur more frequently than a normal distribution would suggest.
The observation of fat-tails is not new, but there is evidence that the frequency of extreme moves is increasing across both equity and fixed income markets. The table below shows that daily moves greater than three standard deviations have become more frequent (a normal distribution would suggest that this happens less than once a year).
Endogenous risk refers to the idea that changes in asset prices can be exacerbated by the behaviour and dynamics of other market participants instead of being driven solely by changes in external factors. Essentially, the risk is of negative feedback loops developing whereby sharp declines in an asset price, rather than bringing in new buyers and establishing a new equilibrium, actually encourage further, often indiscriminate, selling.
This risk materializes in so called ‘flash crashes’ or VaR shocks which can move asset prices to levels beyond those suggested by fundamentals—even at such levels there may be limited participants able or willing to step in and stop the cycle.
Ex-ante endogenous risk is difficult to identify but there are certain features of current markets which appear to be contributing to this risk and the incidence of extreme market moves.
Central banks. Prolonged accommodative policies have simultaneously dampened volatility; decreased the universe of positive yielding assets and encouraged leverage. In response, investors have migrated to riskier parts of the market with evidence of increased herding. The fact that easy monetary policy underpins the valuations of assets across the board means that correlations between assets have increased and what used to be referred to as “risk-on risk-off”, can now be more generally characterised as a system-wide heightened sensitivity to central bank action. Perceived changes in the future path of rates which otherwise might be innocuous can lead to synchronized investment actions across multiple markets.
Volatility based risk budgeting. The introduction of Solvency II means insurance companies now follow an explicit VaR based measure for capital similar to that which is prescribed to banks in the Basel regulations. Very broadly, such frameworks can increase endogenous risk because increased volatility can require liquidation of positions to keep VaR within limits.
But beyond the regulatory focus, and perhaps more significant, is the increase in AUM within systematic based investing vehicles such as risk-parity funds and CTAs. This means there is now a meaningful part of the fund industry with explicit (self-imposed) links between changes in volatility and investment action. Furthermore, the construction of many of these systematic investment vehicles increases the connection between otherwise disparate assets and markets, with potentially abnormal impacts to correlations.
Liquidity constraints. The issue of declining bond market liquidity has been reviewed in detail on this blog. As far as endogenous risk is concerned, this generalised decline in liquidity serves to amplify market moves and increases the probability, and severity, of negative feedback loops.
These are not the only developments which are contributing to a build-up in endogenous risk. The increasing use of ETFs and high frequency trading strategies have both been shown to increase volatility creating a system that is less stable and subject to fatter tailed returns.
Although endogenous risks are difficult to quantify, there are ways to recognize and mitigate them. Analysis of flow data and correlation can provide insight into crowding and cross asset dynamics. Stress testing can help quantify potential tail losses, and hedging via non-linear products such as options can help protect against the risks. Conservative cash management provides optionality so when such forces push asset values beyond fundamentals, active managers can actually capitalise on any mispricing. Finally, a longer term investment horizon can provide insulation from some of the shorter term vagaries of the market.