Recency Bias and Investor Pitfalls…
I’ll make this brief. I hate the chart below. It shows a comparison of 2011 vs 2012. The implication is simple – since the market performed in a certain way in 2011 and this performance somehow correlates with the first 7 months of 2012 then that must mean it’s somehow useful in deciphering how the rest of this year might play out. There are a lot of problems with this sort of analysis, but the most blatant issue is that it’s a horrid case of recency bias. Let me explain.
I once had a friend in college who said that he had devised a winning strategy for gambling. He was a major in some sort of environmental studies so give him a break here. Anyhow, he said if you could count the number of spins at a roulette wheel over a long number of spins and find out when the spins became uneven (for instance, 60 blacks vs 40 reds) that you could establish a winning percentage going forward by assuming that the wheel will generate a 50/50 split over the long-term. Hence, because the wheel has recently been spinning black, it now makes more sense to bet on red. But each spin at a roulette wheel is a totally new and independent event. The odds are always 50/50 (well, not exactly thanks to those sneaky casino bastards who throw the single or double red slot in there!). So each spin has the same odds and past spins have zero impact on future spins. This is a case of recency bias that often plagues stock market investors.
While the equity market is not 100% independent of past events, each new year in the equity markets is impacted by totally new events that are largely independent of past events. The past might rhyme, but it doesn’t just repeat. Yes, there are trends in place (some longer than others), but don’t be fooled into believing that you can predict the future purely by looking at the past. I prefer to use technical analysis and charting tools as a supplement to a good fundamental understanding of current and future events. Remember, you have think like a chess player. And while it might benefit the chess player to understand past history, he becomes a winner by understanding potential future moves. Not merely looking at the past and expecting his opponents to repeat prior moves.
(chart courtesy of Bespoke Investments)
Cullen Roche
Mr. Roche is the Founder of Orcam Financial Group, LLC. Orcam is a financial services firm offering research, private advisory, institutional consulting and educational services.
Follow Me:
11 Comments
I completely agree. And try convincing your chartologist/risk mgr of this.
However, the one thing I will say is that the mkt is not really independent like a roulette wheel…it is possible for enough people to believe that Aug is a horrible month that they act on it and pull their money, creating a self-fulfilling prophecy. In roulette, how the players bet doesn’t affect the wheel. Not so in the mkt.
Still, your point is well taken.
Yeah, not a perfect analogy, but close….
“Reflexivity”, according to Soros.
Recency bias is one of many things that keeps the mkt inefficient, and it’s not going to change anytime soon. Good post…makes me wonder if the “pain trade” is higher. What do you think?
Cullen, your analogy actually has a name: Gambler’s Fallacy
http://en.wikipedia.org/wiki/Gambler's_fallacy
And there’s a very, very fine line between the Gambler’s Fallacy and a phrase we hear all over the investment universe: Reversion to the mean.
some elements always revert to the mean over the long run, chiefly margins, or as grantham put it: “Profit margins are probably the most mean-reverting series in finance, and if profit margins do not mean-revert, then something has gone badly wrong with capitalism. If high profits do not attract competition, there is something wrong with the system and it is not functioning properly.”
@ jaymaster
Reversion to the mean has a totally different rationale to Gambler’s fallacy. Take for example corporate profits, if Corp X can make a 100% profit margin selling in market Y, this is a very juicy target, others invest, new entrants compete, and lo and behold margins for Corp X and the whole market sector erode in time. Ultimately oversupply occurs and profits plunge. Even Apple is not immune to this inexorable market force.
The commodities cycle over the last 200 years of industrialisation is a good long term proof, boom then bust. Ignore these facts at your own peril!
Well, it is certainly difficult to draw conclusions from just looking at patterns on a chart. However, there is a clear similarity here: the market responds to monetary morphine, and monetary morphine has a certain lifetime, after which its effects begin to wane.
Great graph! I am going short!!! Ha
100% agree in the case of longer time horizons like one year. Doing mathematical time series analysis on equity data tells you this. There is however a larger dependency among returns (autocorrelation) and especially volatility (heteroskedasticity) for shorter horizons like a day, which i guess enables trend followers (the good ones) to make money in short term trading.
So Cullen, what is your next chess move, what is the algo saying?
Hi Cullen,
Dont think buying this chart and recency bias as outlined by you are similar. Saying that rest of 2012 may chart out close to 2011 is not same as saying since the market on last 10 days have moved up so the next move is likely to be down than up – which I think is recency bias. Correct me if I am wrong.
There may be some merit in analysing charts this way – heard few people call it congruent markets etc. If we look at the chart of 2011 and consider it a proxy for markets reaction to certain prevalent economic,political,social conditions etc then we can expect similar recation this time around too. Is this not similar to pattren recognition.