Efficient Market Hypothesis as a Useful Benchmark, Not a Literal Law

# Why is efficient market hypothesis still unanswered and no one really seems to care about it?


# Why is efficient market hypothesis still unanswered and no one really seems to care about it?












I have a background in mathematics and have been interested in trading for some time. Apart from some empirical studies, I found little (theoretical) discussion about the efficient market hypothesis in the literature. And yet, as I understand it, this question is the key to everything in this field. I mean, if the Emh is actually true, all those papers on short term trading, portfolio optimization, sentiment analysis, etc. are just a complete waste of time. And all the universities and financial institutions that fund research on this are wasting unimaginable amounts of money, as they would made exactly the same profit than just investing in e.g. the S&P 500 Index. On the other hand, if the Emh is wrong, virtually all economic and quantitative models would be fundamentally wrong, because the assumption that financial market prices are normally distributed is wrong. Nevertheless, at least from my (inexpertly) point of view, there is little to no research on the Emh and (as far as I know) no valid contradiction proof to the Grossman-Stiglitz paradox or the question of why Warren Buffet, for example, has been able to consistently beat the market. So what am i missing here?

In a similar post, this was answered by saying that the research is primarily about risk management. But as I understand it, this is also a violation of the Emh. I mean if there would be a strategy with which you could constantly have less risk but the same return, you could simply buy derivatives (e.g. call options) and also beat the market in the long term, without insider information.

## Answer by Chris Taylor (score 16)

https://quant.stackexchange.com/a/81696

Short answer

The premise of your question is incorrect. The EMH has been resolved, and it is false.

Long answer

Nearly everyone (including the most famous academic proponents of the EMH) believe that the EMH is not literally true. For example, here is an interview with Eugene Fama, who first coined the phrase "efficient market" in a 1965 paper -

https://www.ft.com/content/ec06fe06-6150-4f39-8175-37b9b61a5520

The relevant quotation is

> The efficient market hypothesis is just “a model”, Fama stresses. “It’s got to be wrong to some extent. The question is whether it is efficient for your purpose. And for almost every investor I know, the answer to that is yes. They’re not going to be able to beat the market so they might as well behave as if the prices are right,” he argues.

Fama argues that the EMH is not literally true, and that markets are not actually efficient, in the sense that it is impossible to extract profits in excess of the broad market return. However, he says that for most investors (including professionals) the market is "efficient enough" that it would not be worthwhile for them to seek to beat it, since they are unlikely to have the skills to do so, and the return on effort would not be worth it.

The most common viewpoint among professionals and academics that I know is that the EMH is a useful starting point, from which it may be interesting or profitable to look for deviations. Finding these deviations (also known as "anomalies" or "inefficiencies") has a cost, as does the ongoing exploitation of the inefficiency, and the question is whether the profit available from exploiting an inefficiency exceeds the cost of finding it and maintaining the infrastructure necessary to exploit it.

Many successful businesses have been built around finding these inefficiencies. People often mention Warren Buffett or George Soros, but these examples are less convincing for me than more recent examples of firms which have more consistently outperformed the market, such as Renaissance Technologies, DE Shaw, Millennium and Citadel (the hedge fund). If the EMH was true, it would be close to impossible for these firms to have generated the profits that they have, so their existence and track record is a proof that the EMH is false.

In a comment, Richard Hardy suggested that I should address whether the track record of these firms can be attributed to luck. For example - there are many hedge funds. In any given year some of them will do well, and some will not. Over any span of time, there will be some performance outliers. How likely is it that Citadel's performance is this kind of outlier?

It is not easy to get a track record for a firm like Citadel, as they only distribute their performance to their investors. However, some information does leak. I found a 2021 article in the FT [1] containing annual performance from 1991 to 2020 in the form of a bar chart, and I supplemented this with information from Google searches of the form "Citadel hedge fund returns YYYY" for each year from 2017 to 2024, to cross-check the information in the chart and extend it to include 2021 to 2024. I then subtracted the average rate on three month Treasury bills to get an excess return, and calculate the average arithmetic return (17.8%), volatility (18.9%) and Sharpe ratio (0.95) for this 34 year period. Citadel operates a market neutral investment strategy (i.e. they target zero market beta) so a cash benchmark is appropriate here (as opposed to e.g. an S&P 500 benchmark, which would be appropriate for a long only equities investment fund which was fully allocated).

For a fund to achieve a Sharpe ratio of 0.95 over a 34 year period by chance alone would be very unlikely. For example, if we assume excess returns are normally distributed with mean zero, the chance of this happening is $1.5\times 10^{-8}$ i.e. it would be expected to happen once in 65.9 million trials. Needless to say, there have not been 65.9 million hedge funds that have ever existed (there are currently around 30,000, and there have probably been under 300,000 ever, including all extinct funds).

Obviously there are many flaws to this model (e.g. hedge fund excess returns are not normally distributed, and using a fat tailed distribution would make outliers more likely). On the other hand -

- My Sharpe ratio estimate is probably too low, given other credible information that I have about Citadel.

- Hedge fund returns are correlated, so we do not have 300,000 independent funds flipping coins - it is a smaller number when you take correlations into account.

- Citadel are not the firm with the highest Sharpe ratio or longest track record, they are merely the firm with the best Sharpe ratio and longest track record where I could get reasonably credible performance numbers. Private trading firms like Jane Street, Susquehanna International, Hudson River Trading, Headlands Technologies, XTX Markets or Renaissance Technologies most likely have even more impressive track records, but their performance is not public.

[1] https://www.ft.com/content/25e6100d-4cdd-45d0-aaab-6f9b77b14257

## Answer by lehalle (score 7)

https://quant.stackexchange.com/a/81686

The EMH comes from a time economists tried to make toy models to understand the principles of price formation and investment. For instance in the 80ties Grossman and Stiglitz (Nobel prizes) wrote a celebrated about exactly the opposite: Grossman, Sanford J., and Joseph E. Stiglitz. "On the impossibility of informationally efficient markets." The American economic review 70, no. 3 (1980): 393-408.

What to conclude: all that are toy models to illustrate large principles, they say

- (EMH) investors, who are trading on a double auction game, are not buying or selling at prices that are too much in contradiction with the information they have (when Pfizer patent the vaccine for the covid, its price goes up),

- (GS'80) when processing information has a cost, the persons who are spending money to process it have to be rewarded that for, and hence the adjustment is not immediate.

You can conclude a lot of useful "principles" using this: for instance that while text processing become cheaper (compared to 20 years ago, we have far more machine readable news, open source libraries and Large Language Models), the less rewarding it is to process texts for investment.

## Answer by John Bentin (score 2)

https://quant.stackexchange.com/a/81708

Efficiency is an ideal that is approached more closely in some markets than in others, and at some times rather than at other times. Generally, the greater the population of participants, the more efficient the market.

There are times when a market that has performed well for a few years (for whatever reason) sweeps up a wave of novice investors who believe (and are encouraged to believe) in “momentum”. This belief is, to a considerable degree, self-fulfilling—until it collapses under the patent absurdity of its consequences. A clear example of such a market is Tokyo in the late 1980s.

At other times, society is gripped by a political fervour (e.g. the UK in 1974) that can have disastrous consequences for markets. However, often enough, people return to their senses. At such turning points, markets offer extraordinary value.

In more normal times, when not affected by extremes of sentiment, the major markets are fairly efficient, in the sense that only the wisest investors can outperform the index. Even in such times, though, there are segments of the major markets where inefficiency can be found—in particular, in the stock of small companies. We'll look in more detail at why this should be so.

The general force that drives a market toward efficiency is that wiser investors are more successful and become rich. Thus the market gets to be increasingly dominated by the participants who have the deepest understanding of it. Warren Buffett is a prime example of such an investor. However, because of their success and the consequent huge size of funds that they manage, such super-investors can no longer practically trade in the shares of small companies to an extent that would be meaningful to them. Hence the market for such shares is left to minor players, many of whom have a poor understanding of investment principles. In such markets, fools can easily lose all their money, while the only moderately wise, adopting relatively simple rules for investing, can achieve index-beating, albeit not spectacular, returns.

One particular sector of the market—the stock of car-manufacturing companies—is outstandingly inefficient, in the sense of assuring loss of capital to investors in most companies at most times. Perhaps this is because cars arouse sentiment more than most products. As a current example, Tesla Inc., presently trading at a PE ratio of 116, is priced at the equivalent of about 14 years of total revenue.

All that said, the best advice is to treat the market as efficient and buy an index fund. Anyone who refuses such advice is either wise and brave or (as is more often the case) merely brave and will probably lose their money.

## Answer by KaiSqDist (score 0)

https://quant.stackexchange.com/a/81724

Not an answer, but more of a quote from a book I've read recently that I think is relevant to this question.

From Misbehaving by Richard Thaler, a book on behavioral economics. In chapter 17, the author quotes a line made by Robert Shiller at a major conference about debates between rationalists and behavioral economists -

"I tend to view the study of behavioral extensions of these efficient market models as leading in a sense to the enhancement of the efficient market models. I could teach the efficient market models to my students with much more relish if I could describe them as extreme special cases before moving to the more realistic models."