Arnold Kling  

Can you Beat the Market?

War Economics... Economics and its Enemies...

Financial columnist James Glassman touts a study by Joshua D. Coval, David A. Hirshleifer, and Tyler G. Shumway that shows that some investors were able to beat the market consistently. This conflicts with the efficient markets hypothesis. The authors write,

A rolling-forward strategy of going long firms purchased by previously successful investors and shorting firms purchased by previously unsuccessful investors results in excess returns of 5 basis points per day.

Glassman points out that the study covers only the time period from January of 1990 through November of 1996. Nonetheless, Glassman ends up endorsing the study and endorsing a policy of choosing a portfolio that is not well diversified.

I disagree with that advice. Over a short time horizon, all sorts of risky strategies can appear to be profitable, particularly in a rising stock market. Look at all the people who were successful day-traders in 1998-1999, and who lost everything when the market started to go down.

The overwhelming body of empirical evidence suggests that buying an index fund and holding it is both safer and more profitable than trying to beat the market.

For Discussion. Why would day-trading tend to be more profitable in a rising market than in a falling market?

Comments and Sharing

COMMENTS (14 to date)
Bernard Yomtov writes:

Is that James "36,000" Glassman? Why should anyone take him seriously?

The excess return obtainable from emulating successful investors is 5 basis points a day, according to the study, for one-week holding periods. For one-month holding periods it's zero. I didn't read the paper carefully, but the conclusion Glassman draws sounds vastly overstated. Surprise.

Jim Glass writes:

At the risk of straying a bit off topic, one can wager on beating the market's guess at how long Saddam is going to last at

Osama too.

Place your bets while you can.

Yasser Mawji writes:

"Why would day-trading tend to be more profitable in a rising market than in a falling market?"

One reason is the uptick rule: stocks can only be sold short on an uptick, thanks to SEC regulations. Buying puts isn't a very good substitute because implied volatility rises when the market falls, causing option premiums to balloon. More importantly, though, day-trading is less profitable in falling markets because people are conditioned to buy stocks instead of selling short. After all, the stock market tends to go up in the long run, and bear markets have almost always been short-lived. It's difficult to switch gears when the bull market comes to an end.

* * *

I've never been a big fan of the Efficient Markets Hypothesis because it isn't compatible with human psychology. Human beings tend to make systematic errors in judgment (e.g., irrational loss aversion) which militate against market efficiency. Daniel Kahneman has done some good research on this, for which he was of course awarded the Nobel prize.

The other problem with EMH is the existence of "noise traders". The smart money will be reluctant to play its role if noise traders (daytraders, chartists, etc.) can cause shares to decouple from their fair value for long periods of time. There's some good info about noise trader risk on Brad DeLong's site.

Andrew Lo of MIT has also done some empirical research that seems to disprove EMH. His book, 'A Non-Random Walk Down Wall Street', leads to some rather startling conclusions - for ex., technical analysis does have some predictive value, and this can't be attributed to chance alone.

Bernard Yomtov writes:

What evidence is there that day-traders made money in 1998-99?

That seems implausible.

Arnold Kling writes:

Bernard writes, "What evidence is there that day-traders made money in 1998-99?

That seems implausible."

I haven't seen evidence, but there certainly were a lot more of them back then, so at least some of them were making money. My thinking is that the typical day trader is net long during the day, so that day trading is like leveraged investing, but with a lot of transactions costs. If the market is moving up fast enough, you can make money doing that (although you would make much more just buying and holding).

Yasser Mawji writes:

Surely there's a problem of aggregation that's inherent to index fund investing. If most investors are blindly purchasing index funds without any regard for valuation, then those stocks which constitute the index will become overvalued.

Back in 1999-2000, a stock would rocket 20% or more on the announcement that it was being added to the S&P 500, because investors anticipated large-scale buying from index funds who would soon be forced to buy the stock. So, stocks that were added to the index fund tended to be very overvalued. To make matters worse, the S&P favors stocks with the largest market caps -- which also skews the index towards overvalued issues (assuming, of course, the market isn't perfectly efficient to begin with).

It was no surprise, then, that when the S&P 500 began it's precipitous decline a few years ago, the losses were concentrated in the index, as opposed to the broad market. In fact, the advance-decline line for the broad market was flat or rising for a long time.

For these reasons, index fund investing isn't a terribly good strategy, IMO.

Bernard Yomtov writes:


All very well, but where is the evidence that individual investors can do better than index funds? Managed mutual funds don't, and the managers claim to be experts, have considerable resources to apply to the task, and work full time. As I recall, there are some studies that suggest that the very few managers who are superior investors tend to capture their excess returns themselves, by charging high fees.

So maybe you are suggesting that if I'm smart, and spend money on information gathering, and work hard, I can beat the indexes, maybe. But then why not invest in the index and get a job? Suppose I can outperform the index by 2% a year, or can earn $75,000 in a job. I would need to have a $3,750,000 portfolio for this to be a breakeven proposition. And that doesn't consider risk, fringe benefits, etc.


Yes, there were lots of day traders then. But why assume they were making money and then quit when the market turned and they started losing? Why couldn't it be that lots of people were attracted by the idea because of the market's sharp rise, tried it, lost money, and quit?

I fail to see the point of discussing the cause of something that we have no good reason to assume ever happened.

Yasser Mawji writes:

Bernard, your point is well taken. The market may not be perfectly efficient, but market-beating returns remain beyond the grasp of the average investor, and indeed most fund managers.

Aside from EMH, though, I think that lackluster performance on the part of fund managers can be attributed to the incentive structure. Returns are benchmarked against an index (usually the S&P 500), and investors can always bolt at the slightest sign of underperformance. This encourages managers to mimic the index instead of taking risks. Hedge funds aren't beset by any such problems because they target absolute returns, and investors are locked in for a pre-determined amount of time. Perhaps that explains why hedge funds have beaten the S&P over the past 10 years, despite the vicious bear market (according to CSFB Tremont).

At any rate, despite the allure of index funds I don't believe that indexing will supplant active fund management anytime soon, because of the pathologies discussed previously.

Yasser Mawji writes:

With regard to the daytrading issue, I would be shocked if most daytraders didn't make money during the bull market. Those were some heady times: it wasn't uncommon for technology stocks to go up 20% in a day. Even a parrot trained to squawk "buy" would have done exceptionally well.

Based on my own experiences in the market and cursory readings of various trading-related message boards, my feeling is that most traders made money during the boom, but lost money when the market collapsed because they failed to embrace short-selling. Even the experienced traders who survived the winnowing have folded their tents because volatility has seriously contracted. And the small-time traders who hadn't yet lost everything were forced out by the SEC's recent $25K minimum balance requirement. (Last I heard, these guys had switched to other, more esoteric types of trading -- forex, gold, cocoa, pork bellies, etc. I guess some people never learn.)

Bernard Yomtov writes:


If "I would be shocked if..." is evidence, then let me say that I would be shocked if most day traders made money during the bull market.

On a theoretical level, they shouldn't have been able to make money, especially when you consider transaction costs. So I want to see the non-anecdotal evidence that they did.

As far as hedge funds go,I haven't seen the report you mention, but remember that it is always necessary to adjust for risk, and that unless very carefully done, these sorts of studies suffer from heavy survivorship bias.

jack writes:

Hi Bernard,
The day traders were probably net long for most of the time so would have benefitted even with entirely random positions. Outperforming is a different matter but it also wouldn't be surprising if many of them were highly geared.
Hi Yasser,
the effect of joining the index is probably a distortion but it should be taken in perspective. A company joining the 500 is more or less the 500th largest company. A 20 per cent premium is at most 0.08% which is less than you would save on management charges. In reality that is a grossly conservative upper bound out by two orders of magnitude or more. With a smaller index that could be a problem but not with a big one like the S&P.
As for the article, I have not yet read it but it is very difficult not to be "fooled by randomness", tautologous or biased by survivorship. In any case it would require rather longer to determine the statistical significance of the effect and be genuinely sure that it was the result of a persistent skill.
In general I'm a strong believer in a weak form of the EMH but I am often discomfitted by the flip side -- persistant lack of skill. It is quite possible, judging by the abstract alone that the outperformance is due to the short positions rather than the long. For some reason I find the idea harder to dismiss.

Bernard Yomtov writes:


You make a good point about random long positions in a rising market. I suppose I was thinking in terms of outperforming rather than just making a profit.

I also don't doubt that many day traders were leveraged. But mightn't that hurt rather than help? If your bets are large relative to your capital that increases your chance of going broke. If you can play a game tilted in your favor a very large number of times isn't capital preservation a good strategy?

I don't agree that EMH, in weak form at least, implies any lack of skill. All it says is that price history tells us nothing about future prices. When you get to stronger versions there do get to be issues.

Bernard Yomtov writes:

One more point, a bit "legalistic" perhaps. If you devote your labor to daytrading, and don't outperform the market, you are really not making a profit, even if you have gains. It looks like a profit only because you are excluding the opportunity cost of selling your labor elsewhere and investing passively.

Of course, if you enjoy day-trading more than any other job the labor argument cuts the other way, since you derive non-financial income from your activities.

Jack writes:

Hi Bernard,
I think my original note was confusing. I mean that while it might be hard to tell if a fund manager was good by looking at their past performance it is harder to believe that bad paerformance is also hard to spot. This report
seems to find something like that.

Comments for this entry have been closed
Return to top