Volatility Shock – A Concern That Is Real, but Should Be Tempered By VIX Dynamics

| | Volatility

We believe that a volatility shock will generate a feedback loop that will force equity prices lower. High leverage tied to volatility targeted risk management will mean that any increase in volatility will lead to portfolio rebalancing and position cutting. This negatively correlated leverage effect between equity returns and change in the VIX is real, as measured by researchers at the New York Fed, see “The Low Volatility Puzzle: Are Investors Complacent?“.

 

 

 

 

 

 

 

 

 

 

 

 

 

The key question is the degree of mean-reversion in the VIX. If there is a strong mean-reversion and the seeds of a volatility shock exist in the price dynamics of the VIX, there is a large investor problem. If there is limited mean reversion, the likelihood of a volatility increase is reduced. The researchers at the Fed have also opined on this issue, in their blog piece, “The Low Volatility Puzzle: Is This Time Different?”.

There is mean reversion but it is slow. Low or high volatility will not persist but the adjustment period is long. There is also limited evidence that there will be a jump in volatility when it is low. Finally the volatility term premium has persisted to be higher for longer horizons which suggests that the market still views volatility will be higher in the longer-run. The data show that a volatility spike is not embedded in fact that current volatility is low.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The argument for a rise in volatility has to be focused on some event that will serve as a volatility shock catalyst. The market will need an economic surprise that will change the dispersion in investor expectations. These shocks by definition will be unanticipated. An initial shock may lead to the feedback loop we have described in the past which may turn into a non-linear event unlike what the data has predicted in the past.

See,