Do Put-Call Ratios Tell You Anything?
02/13/2014 8:00 am EST
A popular sentiment indicator, the put-call ratio is used by traders to gauge the mood of the market, and here, Jared Woodard of CondorOptions.com investigates its validity.
The intuition around ratios of option trading volume is that large deviations from normal trading volume may indicate a significant change in investor sentiment, and especially a “capitulation” among traders near the end of a significant trend. In 2013, I wrote about the research on put-call ratios using equity, equity index, and VIX options, and concluded that, whatever their value in earlier periods of market history, these indicators don’t appear to have any significant predictive value now.
During the market dip in early February, VIX option volume (shown above) set a new daily record on Feb. 3, with 2.4 million contracts traded. In light of the spike in VIX activity, some readers wondered whether the most recent market decline would offer any unusual data points compared to the study I did last year. One immediately notable fact about recent activity is that the increase in VIX put volume was large enough to offset the surge in VIX call buying: 1.8M calls traded on February 3, but more than 530k puts traded as well, bringing the ratio of calls to puts to just 3.47.
If anything, the interesting story where VIX option activity is concerned is the higher ratio values from late December and through much of January, where on average, 3.3 times as many calls traded as puts (Dec. 18-Jan. 21).
So what has recent market activity meant for future returns? The plots below show VIX call-put ratio values and S&P 500 Index returns one day, one week, and one month after a given VIX option ratio. Data is shown since the beginning of 2013, with values since Dec. 18 shown in black and regressions run against the full data set shown.
- In 2013 and early 2014, at daily and weekly horizons, there was no apparent change in the predictive value of the VIX call-put ratio compared to the longer-term study we did before.
- At a one-month horizon, the indicator appeared to have slightly more predictive value (Adj. R2: 0.03), but in the “wrong” direction vs. the normal intuition that traders use as the basis for the indicator. That is, higher ratio values were more likely to precede poor one-month returns, instead of marking the short-term bottom in the market.
Extremely high ratio levels (>5) look consistent with positive expected returns at each horizon, but the number of observations is also very low.
By Jared Woodard of CondorOptions.com