One of the most popular sentiment measures is the put/call ratio. Extreme readings in put and call options volume are commonly used as a contrarian signal and are considered by many to be one of the most reliable indicators of future market direction. But does that actually hold up under closer scrutiny?

The underlying idea is that, when options volume becomes significantly skewed towards speculative call buying, investors are becoming euphoric. The market gets too bullish and conditions become ripe for a reversal. On the opposite side an upside bounce is expected when investors panic and buy put options hand over fist. A strategy using the indicator is based on the principle of reversion to the mean.

At the moment there is quite a buzz around this indicator, because in an epic stock market rally from the March 2020 bottom we are now seeing extreme readings, that signal speculative excess:

Let´s take a closer look at how the put/call ratio actually works in practice and whether it has any useful predictive power in signaling reversals.

Issues with mean-reversion indicators

A general problem with mean-reversion indicators (the way most sentiment and overbought/oversold indicators are commonly used) is  that extreme readings can stay extreme or become even more so for quite a while. The momentum effect is a strong market anomaly, that counteracts mean-reversion tendencies and makes trading against the currently prevalent market direction often quite painful.

There are two additional elements I will consider while analyzing the put/call ratio to see, if they can be used to improve the outcome of a mean-reversion signal:

1. Time frame: are mean-reversion or momentum effects likely to dominate over specific time horizons?
2. Trend: are mean-reversion strategies more likely to work when they incorporate a trend filter or signal?

Time frame

Ever since they were described in a seminal momentum study in 1993 these time frames over which either mean-reversion or momentum dominate have continued to be valid:

• Short-Term Mean Reversion (1-month) – exhibits a reversal in returns
• Intermediate-Term Momentum (3-15 months) – exhibits a continuation in returns
• Long-Term Mean Reversion (1 to 5 years) – exhibits a reversal in returns

Over the medium- and long-term time horizons the momentum and value factors have been extensively studied and validated. Short-term mean reversion around a one month time window should be the time frame to look at for an optimal use of the put/call ratio as a trading signal.

The trend is your friend

Strong evidence for relative and absolute momentum effects suggests that trend strategies work well over time; for example by aligning trades with the direction of moving averages based on 50- to 300-day periods (or 3- to 15-months).

This implies that a short-term mean-reversion strategy may be most likely to profit, if trades are taken in the direction of the longer-term trend. Momentum effects should then prove supportive rather than a headwind. For the put/call ratio I want to check what happens, if we filter trades by trend (I am using the 200-day MA slope in my tests):

• Go long on pessimism in an uptrend
• Go short on euphoria in a downtrend

Enough with the general thoughts, let’s get to it!

Backtesting the Put/Call Ratio

Getting good data

Free data sources for the put/call ratio are limited:

• CBOE Equity Put/Call Ratio concentrates on speculative activity by excluding major hedging activity (Index Put/Call Ratio): Stockcharts; Y-charts; CBOE
• Alternative composites: Index Put/Call Ratio & Total Put/Call Ratio

For my backtests I used Stockcharts and additional data by the excellent SentimenTrader service, that provides longer (to 1997) and more granular versions of the ratio data.

Setting signal thresholds

The decisive factor of how good the put/call indicator works will likely be how extreme the positioning is. Rather than optimizing the parameters I looked at the ratio values over the last 5 years and randomly selected the following levels:

• frequent occurrences above and below the noise: 0.85 / 0.51
• rare occurences: 0.98 / 0.47
• extreme spikes: 1.13

At first glance the ratio looks a bit like the VIX with extreme spikes (panic put buying) to the upside and much tighter clusters of frequent to rare extremes below the noise (speculative call buying) – it´s a typical mean-reverting series.

Base rates for comparison

To interpret the results in any meaningful way the context matters: backtests need to differ significantly from the usual return distribution of the S&P500 (Its base rate according to historical probabilities. Return analysis from 1928 – 2019).

Selected Backtests

Long trades

Running the parameters above for long trades (excessive put buying) initially is pretty disappointing. Probabilities for a positive return are only slightly higher than random, except for the most extreme spikes which saw an increased probability for a bounce over the short and medium term:

With a trend filter things are looking better – as was to be expected. Almost all of the extreme spikes are filtered out, because they tend to occur after major trends have already been broken. But both frequent and rare put buying extremes in a bull market significantly increase probabilities for positive returns over all time frames.

Short trades

On the short side basic probabilities are against us, because stocks generally move up over time. Rather than trading explicitly on the short side, increased probabilities for negative returns could also be used to time hedging activity for a long equity portfolio. As put options, a convenient hedging instrument, are sensitive to time decay and volatility, increased short term probabilities and cheap option prices could greatly reduce the cost of hedging or even make this a net positive short strategy.

Again elevated and extreme call buying levels did not influence future return probabilities very much at all. It becomes apparent, that in the late 1990s bull market put/call ratio extremes were very common and often lasted for a long time. Only when filtering for bear market environments are the results more convincing and significant even when considering S&P500 base rates in bear markets (they are close to neutral showing neither a distinct up- nor down-bias):

Contrary to my initial theory, that the ratio signal is likely to work best around one month, the bearish edge of speculative call buying is strongest over the medium- and long-term (3-12 months). It is a good indicator for a market top, specifically (as I filtered for downtrending markets) the top of a bear market rally.
An independent analysis confirms this idea, that extremely depressed put/call ratios may spell longer term trouble in all market environments.

This might prove important in today’s market, but stands in contrast to current strongly positive indications coming from momentum and breadth indicators.
My interpretation is that we are in the decisive area for a market regime change: either we continue to recover quickly to a new bull market or we fall back into a prolonged bear market environment.

Small trader speculation

SentimenTrader provides an interesting additional indicator that focuses on small trader option activity (isolated option buys of 10 contracts or less)

When small traders make extreme bets on a rally with more than 40% to 45% of their total volume on buying call options, this activity is overwhelmingly clustered around major market tops. In the last 15 years call volumes over 45% led to significant losses every time over the next weeks. 

Conclusion

I have seen more convincing indicators then put/call ratio trading signal, but most mean-reversion signals show fairly lackluster results.
The addition of a trend filter makes a profitable trading strategy based on the put/call ratio certainly possible. Excessive call buying in bear markets provides good opportunities for short entries or cost effective hedging of a buy-and-hold portfolio in difficult times – as we are seeing at the moment.

Especially when we consider the speculative excesses of small traders.

Thank you for reading!

David

Browse Recent Blog Posts

If you like what you are reading, please consider subscribing – thank you!