Whenever the market takes a nasty fall, debate inevitably rages anew over whether such dips represent an opportunity to buy or a warning to sell. Efficient market fans will of course argue that there is no information about future price movements embedded in current or historical prices, but they are demonstrably wrong.
The average return on the stocks has averaged 0.44% per month during the years from February 1871 through February 2013 with a standard variance of 4.1%. The probability of losing money in any given month has been 42%.
If we define a "dip" as any decline of more than 1 standard deviation, or 3.7% in any one month, returns in the subsequent month are typically down 1.72% and the probability of further losses rises to nearly 60%. There is typically some very modest relief in the second, third, and fourth months following such dips, but at the end of 12 months, stocks on average are still 1.77% below where they would have been if there had been "no information" about future prices in the dip.
The opposite, known more fondly as the "rip", offers even stronger embedded information about future prices. In the month immediately following a rip, stocks are typically up 3.03% and the probability of a loss drops to less than 20%. Moreover, following such rips, stocks typically continue to hit new highs almost every month, with very little interruption, for another 12 months. Of course, these are only the average returns of hundreds of individual events. There is a great deal of variance in these returns with each individual event, but the divergence created by seemingly trivial pieces of information appears be significant and exploitable.
The most likely explanation for this apparently irrational result is that investors consistently under-react to news unusually consequential news. For example, if new information comes to the market that would add more than 10% of new value instantaneously, it would be rational for investors to react cautiously, and add, say, only 5% to prices. Then it would still be necessary to add another 5% over subsequent months to complete the adjustment.
The main problem with using such a strategy effectively is that dips, and rips, by definition are rare occurrences. On average, each happens once per year, but in practice, they tend to cluster so that you frequently see 2 or more within the span of a quarter, and many years without any. The last rip was in July of 2013 and whatever information was contained in that rip is by now, almost fully dissipated. The last dip was in August of 2011.
Counter-intuitively, rips are slightly more likely to happen when the market is over-valued, and dips more likely when the market is under-valued, as defined by the standard deviation bands around the Shiller price/10-year average earnings. But the difference in the message being delivered by rips and dips while under-valued and the rips and dips while over-valued is phenomenal.
When the market is over-valued, any excess volatility should be taken as a warning sign, while any excess volatility during a period of under-valuation is bullish. During periods of over-valuation, a rip is still more positive than not, but the average total price appreciation even 12 months after a rip under over-valued circumstances is only 2.6%. This is as good a time as any to gain excess income by selling covered calls. If the market exceeds the strike-price during such periods, you should be happy to unload you positions and collect the premiums while waiting for a better entry point.
On the other hand, a dip during periods of over-valuation is a clear sign to head for the hills. The average decline in the following month is 3.8%, and while there is often a subsequent bounce, the market is down another 13.5% on average during the subsequent 12 months.
Conversely, any excessive volatility while the market is under-valued is likely to be a clear sign to get as aggressive toward stocks as you possibly can. Especially when the market rips, extraordinary gains can follow. On average, the very next month is up almost 9% and the odds of losing money are less than 12%. Stocks on average gain another 1.5% in the second month following a "Rip While Under-valued" and are up on average 24.8% at the end of 12 months.
Dips are almost as bullish, but require a little more finesse. The first month following a "Dip While Under-Valued" is usually not very pretty. Stocks drop another 1.3% on average, and the odds of a loss are more than 60%, but thereafter, gains tend to be even stronger than those following rips. At the end of 12 months, stocks are up on average 20.7%. While this is not as high at the total of 24.8% gained on average in the 12 months following rips, remember that a very substantial percentage of the post-rip gains are earned in the first month. In the 11 subsequent months, post-rip gains average only 16%, while post-dip gains average 22%.
The Aston Research asset allocation model already takes advantage of this information indirectly, in that it's weighting toward equities is based primarily on the Shiller valuation regime, adjusted for a momentum signal that, while not based explicitly on rips or dips, is likely to be triggered by either. Thus, nothing really needs to be changed. The market has been fairly quiet for an unusually extended period and is due for an interruption any time. Now we know what to do when the next big one comes. The market is over-valued on the basis of the Shiller methodology, and nothing good has ever been sustained under such circumstances. Any big rip from this point onward would only be mildly supportive, while any dip would be a sign to short the market immediately.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.