About That Recent Market Volatility
About That Recent Market Volatility
Note: I revised this piece slightly on November 5th
Note: I revised this piece slightly on November 5th
In what follows, I’ve updated my follow up on John Cochrane’s blogpost earlier this year called Tremors, which raised concerns about the volume of outstanding U.S. Treasury debt, after which he co-authored a Washington Post op ed. Those posts give me an opportunity to discuss an ominous empirical regularity that in the past has often provided a leading indicator of investor flight to safety from stock market crashes or financial crises, because the empirical regularity occurred once again at the end of last year, and was followed by the mini-crash in early February. The question is, will there be more to come?
The leading indicator in question measures how sensitive is the volatility of the stock market as a whole with respect to the current level or “price” of the market as a whole. This elasticity’s usually negative, which reflects what has been called the “leverage effect” in that when the stock market goes up, volatility tends to go down, and when the market goes down, volatility tends to go up. This tendency is associated with normal aggregate market behavior. It’s when you see the stock market and stock market volatility moving in the same direction that you tend to see subsequent market crashes or financial crises.
The Estimates
You can estimate the elasticity using a simple ordinary least squares regression of the natural log of the volatility against the natural log of the stock market quarter (or month) using daily data over a short horizon. For the discussion here, I’ll estimate the elasticity between the S&P 100 index and the “Old VIX”, a measure of future volatility implied by options on S&P 100 index options. The figure below depicts the elasticity that I’m referring to from Q1 1986 through Q3 2018 for the relationship between the S&P 100 and the Old VIX, estimated from daily data between January 2nd, 1986 and September 30st, 2018, which you can get from http://finance.yahoo.com:
I published a paper (the working paper’s here) documenting this empirical regularity some time ago and the graph above updates those findings. [The inspiration for this exercise came from the applied mathematician, Fischer Black, who mentioned this in passing in a short article in the context of the 1987 crash; see page 272]. You get similar findings if you apply this exercise to the S&P 500 and the VIX, or to similar data that’s available for European markets.
The sign switch occurred in Q1 and Q3 1987, leading up to the October 1987 crash. It happened again in Q1 1989, which preceded the October 1989 mini-crash.
The 1990s were a particularly volatile period. The next two sign switches in Q1 1992 and Q1 1993 could be related to the troubles in Japan or the British Pound crash. Then came the sign switches in Q1 1995, Q2 1995, in this case after the Tequila Crisis that began in 1994 and got resolved by March 1995. The sign switches happened again in Q1 1996, Q4 1996, Q1 1997 and Q2 1997, prior to the Asian Crisis that began in July 1997 and continued on through early 1999.
No sign switch occurs during the IT sector crash in 2001, which could reflect the fact that the S&P 100 tends to have a lower IT exposure. The next occurrence happened in Q2 2007, after which many of the events concerning the Subprime Crisis began to unfold. The most recent occurrence happened in Q4 2017, and then we had the mini crash in early February.
Interpreting Results
To understand the intuition behind the graph take a hypothetical rational investor’s point of view. A typical investor doesn’t like market volatility, at least not without compensation, but the sign of the elasticity between the market and market volatility tells you whether a market move is adding to, or subtracting from market volatility.
In ordinary times, a negative elasticity indicates that as the market goes up, volatility goes down, so you’d expect investors to allocate more wealth to the risky stock market. But when the sign switches, an increase in the market adds to market volatility, and the only way for investor exposure to market volatility to fall is if investors allocate less to the risky stock market.
The most recent sign switch happened in Q4 2017, and the S&P 100 experienced a mini crash of 3.8 percent on February 6. For the S&P 100, that’s a far cry from the 23.8 percent decline on October 19, 1987, and smaller than the mini-crash of 6.5 percent on October 13, 1989.
In a previous op ed, I mentioned this empirical regularity in the context of explaining how officials don’t necessarily need lots of information to see a crash or crisis looming on the horizon. After all, Title I of Dodd Frank called for the creation of the Financial Oversight Council (in Subtitle A) and the Office of Financial Research (in Subtitle B). Title X called for the establishment of the Bureau of Consumer Financial Protection, and officials there have sometimes mentioned the Bureau’s role in preventing a repeat of the last crisis. Then again, no two crises seem alike, and crises are hard, if not impossible, to predict consistently.
Conclusion
The elasticity does not suggest a policy, but simply suggests an oncoming flight to safety in the face of a perceived increase in risk. To the extent that the sign switch may signal flight to safety, and considering John Cochrane’s Tremors post and the fact that these sign switches have occurred less frequently since the 1990s, the question remains, will there be more to come? Let me conclude by saying “THIS POST IS NOT INVESTMENT ADVICE.”