Arnold Kling  

Efficient Markets Theory

PRINT
Behavioral Philosophy?... A Pacifist Syllogism...

Megan McArdle writes,


It doesn't matter which version of the EMH is correct. It doesn't matter if the behavioral finance guys are correct. You--adorable, clever, hardworking little you--are mathematically just as likely to underperform the market as outperform it. You would do better to go to Vegas and sit down at the $25 blackjack table with a firm resolve to walk away as soon as probability has varied a few hundred dollars in your favor.

And the guy you're paying to manage your money? Same deal. Statistically, in fact, he will give you a lower return than a broad market index, because of his salary and trading fees.


She refers to an article by Michael Lewis, which I assume is this one. The article is too long for my taste.

But the point is important that the typical investor should act as if the efficient markets hypothesis is true. One of the most widespread biases in finance is hubris. I could not believe it when Mike Moffat sided with Patri Friedman's argument to use mortgage leverage to buy stocks. In an efficient market, that is just plain incorrect.


Comments and Sharing





COMMENTS (23 to date)
Dennis Mangan writes:

Any account of the EMH will have to deal with Warren Buffett's speech, The Superinvestors of Graham-and-Doddsville, which notes the long line of successors to Benjamin Graham who have consistently beat the market.

John Thacker writes:

Any account of the EMH will have to deal with Warren Buffett's speech, The Superinvestors of Graham-and-Doddsville, which notes the long line of successors to Benjamin Graham who have consistently beat the market.

Not so long as it applies to the individual investor, it doesn't. Even if small groups of super-talented people investing limited sums of money can beat the market when they spend all day developing technology and working to do so, that doesn't mean that anyone who has another job can do so. And don't say that they can just find someone to beat the market for them, because anything that an average person can invest in isn't going to beat the market-- because if it does so too much, then the arbitrage will go away, the EMH will take over. Even for hedge funds, with limited amounts invested, the problem of competing new hedge funds using exactly the same strategy has destroyed many arbitrage systems and turned their hedging into leveraging.

All the people out there working to beat the market are what make the market efficient. Those returns by the super-investors? That doesn't make it more likely for the average investor to beat the market-- it makes it less likely. They've already taken the arbitrage, and their have the superior technology and time and effort spent in order to make their trades before you can, and to take advantage of new information before you can.

John Thacker writes:

Any account of the EMH will have to deal with Warren Buffett's speech, The Superinvestors of Graham-and-Doddsville, which notes the long line of successors to Benjamin Graham who have consistently beat the market.

If they all used substantially the same strategy, which is likely if they all studied from the same master, then it's simply not nearly as unlikely as he claims that they all did well. Their chances of success were not independent at all, but he claims in his speech that they were when arguing how unlikely "all" of them succeeding was. That strategy was simply one that happened to work over the particular time period. What does that tell us about the future?

Mike Moffatt writes:

"One of the most widespread biases in finance is hubris. I could not believe it when Mike Moffat sided with Patri Friedman's argument to use mortgage leverage to buy stocks. In an efficient market, that is just plain incorrect."

Well, if you're surprised that I took a contrarian position that is likely to be wrong, you clearly don't know me well. :)

EMT can hold and this could still be a good idea, if you're less risk-averse than the market. So it's not "wrong", per se, the same way that having a stock heavy portfolio isn't wrong.

For someone with the same risk-profile as the market, if EMT holds then it really makes no difference one way or the other. But I suspect EMT might not hold.

More on that on my site.

Dennis Mangan writes:

John Thacker: Wrong. For one thing, the average investor can invest in Berkshire Hathaway (or Leucadia National, or Brookfield Asset Management, or Third Avenue, to name just a few companies run by superior management). Furthermore, none of these superinvestors uses "technology", like say Jim Simons of Renaissance does. Assets that the "average" investor can invest in are *more* likely to be able to beat the market, simply because big money can't touch them. Anyway, as you point out, efficient markets are predicated on investors trying to beat it, so someone has to, and as Buffett showed, a certain group of investors did so consistently over decades. His point was that, since they all followed a specific investment style, namely Grahamian value investing, the result was very unlikely to be a coincidence.

Mike Moffatt writes:

"More on that on my site."

Or at least, there would be more on the site if my blog was replicating, which it doesn't seem to be. Hopefully the post will show up soon.

spencer writes:

Mike Moffet -- your problem is that you ar relying on averages to reach your conclusion.

It is like the story of the six foot economist that drowned fording a river with an average depth of three feet.

If you are leveraging your investmnet in the stock market there is always a good possibility that even though over time you get the average return at any time your portfolio could go to zero and you have no assets left to participate in the market rebound.

Moreover, Patri Friedman was taking that risk twice -- once with her home and once with her stock portfolio. That is why I commented that she did not understand the risk she was taking.

Risk is not about averages. It is about the unusual event.

Mike Moffatt writes:

"Risk is not about averages. It is about the unusual event."

Well, yes, that's why it's called risk.

Nobody is expecting that stocks should have the same return as bonds or mortgage interest, etc. But unless you have a fairly high level of risk aversion, Patri's play seems like a good one to me.*

* With the caveat that I may very well be mis-reading the figures.

John Thacker writes:

His point was that, since they all followed a specific investment style, namely Grahamian value investing, the result was very unlikely to be a coincidence.

That point is utterly wrong. They all followed a specific investment style, out of a universe of many investment styles. It is absolutely guaranteed that some investment style would have outperformed all others, looking back. That ten people, or a hundred, or more following that strategy succeeded is no more unlikely than one person following that strategy succeeded, given that that one strategy happened to be one that outperformed over the given time period.

One might as well imagine a tout that said that you should bet on the Red Sox to win the World Series this past year. Low and behold, everyone who followed his advice won. Would it make sense to say, "It would be one thing if only one person who followed my strategy of betting on the Red Sox to win it all one, but everyone who bet on them one. That so many were able to win following my strategy makes it unlikely to be a coincidence?"

No, it wouldn't. So long as the strategy worked, it would work for everyone following it. Therefore, it adds no evidence the more people who follow exactly the same strategy succeed.

Berkshire Hathaway's money is also largely from insurance and buying entire companies these days, not stock picking in any case. The share price also discourages individual investors from buying it, I would think.

John writes:

I'm sorry, but stock returns aren't normally distributed and the Fed has found evidence of autocorrelation in stock returns. I think it's more correct to say the market isn't efficient, but it's getting more efficient everyday. Regardless, the average person who isn't willing to spend their life studying finance and investing probably will probably be better off investing in an index fund or copying the investments of famous investors.

8 writes:

EMH only works if everyone is exactly the same, with the same exact brain, same age, same risk tolerance, same everything. Or, if people think and behave differently, it creates random variation. EMH assumes the variation is unknowable and impossible to predict.

But if EMH were true, there would be no fashion industry.

Brandon Berg writes:

I could not believe it when Mike Moffat sided with Patri Friedman's argument to use mortgage leverage to buy stocks. In an efficient market, that is just plain incorrect.

Efficient or not, isn't it generally acknowledged that an equity premium does exist? And with mortgage interest deductible at a higher rate than capital gains are taxed, the government is essentially subsidizing homeowners who arbitrage the equity premium.

Dennis Mangan writes:

Even Fama and French have come round to the idea that low price to book and small cap stocks outperform the market, and have done so since good data have become available, in 1926. I don't know how long a period of time or how large a data set you would need to forgo the claim of coincidence; it's clear that if value stocks outperformed for 300 years, someone would still be able to say, "Well, the next 300 years could easily be different."

Independent George writes:

...since they all followed a specific investment style, namely Grahamian value investing, the result was very unlikely to be a coincidence.

I actually disagree with this point. Warren Buffet was spectacularly successful following Graham's advice, but what's lost is the fact that it's still Warren Buffet. Value investing doesn't follow some magic formula that never changes; requires a lot of analysis and knowledge on the part of the investor. Two people can follow the same strategy and still lose if they're not smart or diligent enough. The key questions are: (1) Are you (or your manager) smart or diligent enough to take advantage of value investing? (2) How can you tell?

A good comparison is the NFL. For a while, it seemed like every team was hiring a Bill Walsh disciple and installing the West Coast Offense. It turned out the WCO worked best when Steve Young or Brett Favre were running it. It was somewhat less effective when run by the likes of Doug Pederson, Joey Harrington, or Cade McNown.

In other words, it doesn't matter if 100% of the people beating the market consistently are Graham disciples, if they represent 1% of the population of all Graham disciples. It doesn't matter if some managers are able to beat the market consistently if you can't be sure who they are until after the fact (and, even then, not always).

Mike Moffatt writes:

"Efficient or not, isn't it generally acknowledged that an equity premium does exist?"

Yes - I make this exact point on my blog entry, which has *finally* replicated and is online. Anyhow, it appears great minds think alike. :)

What Patri is essentially doing is buying stock and taking a short-position on a quasi-bond. Given the equity premium puzzle, I think this, in some cases, can be a smart move.

While Mr. Kling is correct in suggesting I'm ignoring the efficient market theory, I think he's being too glib in rejecting the equity premium puzzle, given that in the 25 or so years that economists have studied it, they still don't have a satisfactory answer on why stocks have a higher risk-adjusted return than bonds.

Patri Friedman writes:

Sorry Arnold, but as Brandon points out, you are wrong. For two reasons: 1) Equity Premium, 2) Tax Deductibility of mortgages (even under AMT!).

(2) is the real killer. We can completely set aside (1) as unnecessary, because the fact that mortgages are tax deductible for individual homeowners means that the EMH does not apply. No hedge fund can go arbitrage away the difference between tax-advantaged home loans and the performance of the market. Because they don't get the tax advantage.

Risk-tolerance affects the mix of stocks/bonds I should buy with the money, but I don't think it affects the basic point.

If you agree that EMH does not apply to tax-advantaged home loans, I'd appreciate it if you'd update the post with that correction.

eccdogg writes:

Isn't the fact that you should use leverage instead of stock picking to amplify your returns is you are willing to accept more risk one of the key elements of Markowitz and the EMH.

So borrowing against your home to invest in stocks is not crazy if you are willing to take the risk. It is just a simple way of getting leverage.

Lord writes:

There is really nothing wrong with mortgaging your home to buy stock if you have a reliable source of income to pay it. The problem is if you lose your job and can't replace it when the market tanks. A secure position makes this easy to take, while an insecure position makes it difficult. Though equities return around 10% during accumulation, they can often support only 4% during decumulation. So the question is how secure is your job?

Geoffrey writes:

I once used the example of Warren Buffet on a Bryan Caplan midterm as an example of an exception to the EMH.

He marked my answer as wrong saying something to the effect - the EMH doesn't exclude outliers.

Richard writes:

Small nitpick to Spencer: Patri should be "him."

fundamentalist writes:

8:

EMH only works if everyone is exactly the same, with the same exact brain, same age, same risk tolerance, same everything. Or, if people think and behave differently, it creates random variation. EMH assumes the variation is unknowable and impossible to predict.

Good point, 8! There aren't enough good Austrian economists reading this site. EMH isn't a fact. It's a theory that's based on very faulty assumptions. In addition to those you expressed, everyone would have to have the same information and know everthing. The reason some people can earn more than the market indexes is that some people know more than others, Austrians for example.

As for Buffet and others simply being lucky, or outliers, that applies only to single instances, a few instances at most. But when someone consistantly outperforms the average, then you know that something systematic is happening. That's all very basic quality control statistics.

But we don't want the majority of people giving up on the EMH, for as long as most investors are slaves to EMH, they'll be blind to arbitrage opportunities. Like the guys who sees a $10 bill on the street and his EMH buddy tells him it doesn't exist.

For the average investor who doesn't want to work at his investments at all, indexes are a good choice. But for those who want to put a little effort into it, stock clubs are probably the best way to go. Many stock clubs outperform the market by a wide margin over many years, for too many times for anyone to claim their simply outliers.

fundamentalist writes:

PS: Great investment strategies are far more than simplying seeing arbitrage opportunities; great investing is seeing arbitrage opportunities before they appear, and before others see them. It's clear that if an arbitrage opportunity exists, many people will see it and it will vanish. Great investors can see those opportunities before they become apparent to others, especially those blinded by the bright light of EMH.

Tracy W writes:

EMH only works if everyone is exactly the same, with the same exact brain, same age, same risk tolerance, same everything. Or, if people think and behave differently, it creates random variation. EMH assumes the variation is unknowable and impossible to predict.

No it doesn't.

The theory behind EMH is that if there exists a reliable way of making above-average returns on the market, then investors who are after the money will enter into the market and compete away that method. For example, if the stock market always dips down unjustifiably in January, smart investors will buy stocks in January when they are underpriced. But if lots of smart investors buy stocks in January then the price of the stocks in January will rise, and thus the unjustifiable dip will disappear.

This does not depend on all investors being alike. Consider investors with different levels of risk tolerance. Highly-risk averse investors will only engage in arbitrage opportunities - which are risk-free by definition. A less risk-averse investors may invest on a 90% correlation, extremely risk-loving investors may invest based on speculative theories. So what? They will all compete away whatever price opportunities available.

Or take investors of different ages. A 60-year old investor may have more of their assets in low-risk, low-return government bonds than a 25-year-old investor due to age. But how does that translate into opportunities to beat the market?

As for people having different information and different theories about the world - EMH does not assume that all the variation in their thoughts is unknown and impossible to predict. EMH argues that the prices will adjust to those bits that are possible to predict. For example, people buy warmer coats for colder weather. People's tolerance of the cold varies greatly, for example my husband has been spotted wandering around in snow in shorts and a t-shirt. People's level of forward planning varies greatly, some people buy warm coats when the shops first start selling them, others only realise they are heading for cold weather when they get off their Miami-to-Chicago flight, exit the airport and then turn around immediately and go back in to find a clothing store. Now say you are a clothing retailer trying to decide when to stock coats and how many. You have ample past information on coat-buying habits throughout the year. You have weather forecasts. If you really want to match coats-in-your stores to customers (either out of the goodness of your heart or to maximise profits), EMH predicts that you will do so well that the only errors in your forecasts will be the unpredictable ones. For example, you may not know the exact timing of a cold snap in early autumn. But you will manage to stock your coats to take account of the surge in buying when it starts to turn cold, and also have some on hand for later in the season for people who just flew in from Miami, or had their coat eaten by the puppy.

To summarise, EMH does *not* assume that people are identical, nor that they vary randomly. EMH assumes that people are smart enough that what is left after what can be predicted is the unpredictable stuff.

A lot of people have gone broke trying to disprove EMH.

Comments for this entry have been closed
Return to top