A Real Time Experiment In Contrarian Market Timing

Includes: SPY
by: Robert Keyfitz

A few weeks ago, I got interested in contrarian market timing - especially when I discovered there seems to be something to it (see A Guide to Contrarian Market Timing). Over the past decade, pessimistic investor sentiment as measured by a low value of the American Association of Individual Investors (AAII) bullish sentiment index has been a statistically significant predictor of above average returns on equities.

A look at the data suggested the following contrarian strategy: After a low reading (< 25%) of the AAII bullish sentiment index, buy SPY and hold it for 6 weeks; then sell it and do something else with the money until it happens again. There have been 25 occurrences of bullish sentiment < 25% in 2002-12, following which the average 6 week return on SPY was 4.5%, or 46% at an annual rate. By contrast, buying SPY without regard to the value of bullish sentiment had an expected return of only 0.2% (Figure 1).

Figure 1 - Value of $1 invested in SPY after AAII Bullish Sentiment < 25% compared to the unconditional expectation, 2002-12

The experiment: On May 17, the bullish sentiment index fell to 23.6%. At the time I had some SPY shares under the mattress which I had planned to sell (the shares, not the mattress). But, after seeing the results outlined above, I decided to hang on and see what would happen.

I charged the experiment for 300 shares of SPY at $130.86, the closing price on May 17, 2012, the publication date of the bullish sentiment index. Six weeks later, on June 28, I sold them at the closing price of $132.79. In addition to the capital transaction, I credited the experiment with dividend earnings on an accrual basis using the 2% dividend yield of the most recent quarterly dividend. Table 1 summarizes the P&L. Over the 6 week holding period, the contrarian strategy netted $670 or 1.7%, equivalent to 15.8% at an annual rate.

Table 1 - Profit and loss statement for the contrarian experiment

What was the experience like? This is really what I wanted to learn about, and it was a bit of a roller coaster ride. My investment bounced around from 2% down to 4% up (Figure 2), and there seemed to be no shortage of things to worry about (Greek elections, Spanish banks, JP Morgan's internal controls,…). Meanwhile, a string of pundits all seemed to be trying to out-bearish each other - this one predicting the market could go down to 1200, that one to 1150. In other words, this wasn't a relaxing way to earn 1.7%. Of course, that's exactly what you'd have to expect: there will never be a contrarian buy signal at a time when things are looking rosy. As I wrote earlier, contrarian market timing is especially suited to more risk tolerant investors since the exceptional returns are partly compensation for greater risk.

Looking back on the experience I think I'd do it again, as in this case on a limited scale. An alternative might be to shed some of the downside risk (with protective puts or long calls), though in such a bearish environment that could be expensive.

Figure 2 - Expected value of $1 invested in SPY and actual outcome for the 6 week period beginning May 17, 2012.

A final note: The day after I cashed out, the markets had their best day in 6 months. So did my contrarian strategy let me down after all? Fortunately, the answer is no. I've been working on a more sophisticated version incorporating sector rotations and I used the proceeds from the SPY sale to buy a blend of sector ETFs. Last Friday they did almost as well as SPY, and with luck over time they'll do better. I'm not quite ready to publish that one yet, but I hope to soon. I promise, it will be my last word on the subject.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500. Become a contributor »
Problem with this article? Please tell us. Disagree with this article? .