Yeah, but what have you done for me lately?
A few years ago, professors from the University of Nevada at Reno and the Wharton School observed gamblers at the roulette wheel. In case you’re not familiar with roulette, in short, gamblers place money on even or odd, black or red, or a number—one through 36, plus 0 and 00. A wheel is spun and the gambler is awarded money if a ball falls into a slot that carries one of the attributes he picked. (I’m sure there are other possible bets. Can’t remember the board right now.)
The researchers wanted to see how recent wins or losses affected the subsequent bets gamblers placed. And after watching 18 hours of surveillance footage on one roulette wheel, they got exactly what they were looking for.
Maybe this isn’t revelatory, but gamblers place a lot of emphasis on recent wins. If “black” won several times in a row, the number of bets on “red” increased, in the belief that the color was “due” for a win. If a certain player made a few great bets, other players started to mimic his bets, with the idea that he had hot hands.
The roulette wheel wasn’t rigged—that would make for a really fun experiment—so in reality, the ball went wherever it was going to go, without regard to the last result. So following either of those biases never consistently rewarded the gamblers with anything. Still with little data to bet on, the gamblers went with what they last observed.
Remind you of anything? I hate comparisons of the stock market and gambling. If you’re buying stocks for the really long term, company performance and dividends do matter. But it’s fair to say that in the short-term, price performance of stocks behaves more randomly. Despite the forced causation foisted on stocks in the press— “Stocks down 1% on profit taking” —there doesn’t have to be any reason for a couple percentage point move one way or the other.
The best predictor of where stocks go next
That brings us to the single, best predictor of where a stock price will most likely go in the next millisecond. It’s true for any market price for that matter, including that of homes, gold, and, uh, timber. It was nothing to do with price-to-earnings ratios, dividend yields, bond spreads, or tree diseases.
If you have no other information and need to guess the market’s next price, your best bet is to go with the recent winners. A stock that’s performed well in the last 12 months is likely to perform well in the next 12. Of course, overall and in the longterm, I think this is a bad strategy—one day the bet will be wrong and it will cost you a lot of money. But if you held a gun to my head, in the short-term, I’d never bet against momentum.
Our memory of recent events is simply much stronger than our long-term memory. This translates well into other aspects of life too. In the 1950s, a couple researchers looked at decisions made by judges to see if they could be predicted by whether the plaintiff or defendant made the most recent argument. Sure enough, if he last heard the plaintiff make his case, the plaintiff was more likely to win and vice versa. The effect disappeared if there was at least a week’s delay between the argument and the judge’s decision.
I’m sure this behavioral tendency has some sort of deep evolutionary cause. If the buffalo were grazing on the hilltop yesterday, they’re probably still near the hilltop, rather than by the river they were at two weeks ago. But today, it just means that we tend to bet on bubbles and run from crashes, even though the market is looking increasingly expensive or cheap.
So what’s happening now?
The thing is, the market’s done really well over the last 12 months. If own a S&P 500 index fund, your money would have returned 26.5% in 2009 and 3.8% this year. And yet, mom and pop investors are still pouring money into bond mutual funds.
Recency bias would suggest that the opposite should be happening. So what’s changing their minds?
Let’s set aside all the rational reasons for a moment. Sure, investors could be doing market analyses and coming to the conclusion that the market is overpriced. Maybe they’ve done an economic analysis that predicts a double-dip recession or are afraid of political meddling in the market or economy. For the sake of this argument, let’s just pretend that they’re not spending much more time on their investments than to look at how they’ve changed recently.
Remember the roulette wheel? If overall, the ball has an even tendency to land on red or black, gamblers start to bet more heavily on red if black has had a run.
For a long time, the stock market didn’t seem like an even bet. It was going through the greatest bull market ever. So all else being equal, investors naturally guessed it would continue to rise.
But the last ten years have been a lot different. The market’s spent as much time falling as it has rising. So even though we’ve had a great run since March 2009, I suspect a lot of investors are starting to think the market’s “due” for a drop.
Would I bet on it? Nah, these are just the musings of a casual observer. But it’s interesting to watch investors do exactly the opposite of what I would have predicted a year ago.
{ 5 comments… read them below or add one }
I don’t know where I read this (perhaps in a blog or in a book), but depending on where you were raised, your betting odds were different as well.
They put students from Asia and students from North America through the same test. When they saw a graph of something going up really high without dropping, Asian students assumed it was time for a drop and were pessimistic. North American students saw it as a trajectory that wouldn’t change.
And vice versa. They also observed that Asian students tended to be slightly more cautious and less risky, opting for a moderate increase with a low average (something like a straight line) theory, rather than wanting a lot of risk with high bumps and low valleys.
@FB That’s really interesting. I also read recently that those who are ideologically conservative are likewise conservative in their economic forecasts. I couldn’t find the link though so was hesitant to include it in here. I wonder if that North American bias will still be true if the market keeps see-sawing.
@Pop, regarding your statement “your best bet is to go with the recent winners”, do you have a good reference? The technique of picking stocks by looking solely at graphs — called technical analysis — was given a lickin’ in the book ‘A Random Walk Down Wall Street’, but a convincing counter-argument would be interesting to check out.
I have a more complicated explanation, Pop.
I think that we all suffer from cognitive dissonance re stock prices in bull markets. We “know” intellectually that the insane gains are not real but we also “know” emotionally that we very much want to pretend that they are real. So we continue investing heavily in stocks despite the extreme danger while also having a thought in the back of our heads that this is not going to turn out well.
When the crash comes, we let the thought in the back of our heads into the front of our heads. But we do this gradually, not all at once. It is too painful emotionally to take this all in at one time (out-of-control bull markets are so costly that they usually crater entire economies). We let reality in a little bit, then back off and returns to the fantasies a bit, then let in a bit more reality.
We let in a good bit of reality in early 2009. Then we got scared and pulled back. If this theory is true, then over the next few years prices will drop to about one-third of where they are today (with temporary gains mixed in at times).
Rob
@Derek – Sure, the random-walk hypothesis has been pretty well debated over the years, but the paper I’ve seen referenced the most is “Returns to buying winners and selling losers” by Jegadeesh and Titman. Basically, it found that buying a portfolio of “winners” that had gains over the last 6 months and selling “losers” that had losses produced gains.
However, I’d never tell anyone to actually do that. I was meaning to set up a little bit of a straw man…if you gave me a P/E, a dividend yield, a debt/equity, and a stock’s price performance, and asked me I could have one of those before guessing where the stock price moved next, I would ask for the most recent price move. That said, momentum holds up less well in stocks than it does for other, less liquid assets, like home prices. Robert Shiller has done a lot of work on how home prices are more subject to the momentum trend than stocks.