The Efficient Markets Hypothesis

In the first video in the Personal Finance section of our Principles of Macroeconomics course we pointed out that mutual fund managers do not beat the market on average. Why is this? In the second video, we take a look at the efficient markets hypothesis.

I’m rather fond of this video as it has some good story-telling elements, features Einstein in an important cameo (do you know why?) and yet also covers important ideas in an intuitive yet deep way.


First! An economist and a physicist go walking, and they come across a 20$ bill on the ground. The economist ignores it to the physicist's bafflement. The physicist asks why, and the economist says that someone has already picked it up.

Know anybody who makes a living walking around looking for dropped $20 bills?

My five year-old seems to think it's a career path. He wants to go on walks to look for dropped quarters.

The physicist should understand that perfectly.
The $20 is in a state of quantum flux.

No, Sham the Sam, it was not a $20 bill but a $20 TRILLION dollar bill. And nobody picked it up since they were racist.

The EMH hypothesis, interestingly enough, works when it's mostly but not entirely accurate - i.e. when it's 95% accurate versus 100% true. As long as that's the case, then we get what we have now.

I guessed wrong on Einstein. But three criticisms on the video:1) It talks about number of aging population, not proportion and even the graph shows % change - which is not very useful unless the aging are a large fraction of the total (compare to TC's previous post about Faroe Islands' birth statistics; with a base of only 47,000 comparing its birthrate to other countries is lame) 2) The Morton-Thiokol O-ring wasn't the problem, anymore than your car is the problem when you drive it off of a 10,000 ft cliff. At the temperatures (we actually don't know what the temps were, since no one bothered to monitor them at the critical location) of the launch, O-rings had a very good chance of failing, they were NOT DESIGNED to function at that temperature. This wasn't Morton's problem (although we only have anecdotal accounts of how hard they tried to wake NASA to the danger). Blaming Morton and not the criminally negligent NASA hierarchy is typical left wing "evil corporations" nonsense.3) Sociologists studying the US housing market believed that the sustainable new home construction rate was around 70k - 90k/yr. In 2007 the new home construction rate was above 200,000/yr. Public Information. The idea that the market has CORRECTLY adjusted for public information is simply wrong, wrong, wrong. There are plenty of examples, like the housing bubble, where the market ignores information. (Although I do agree that anyone who assumes that the market will continue to act irrationally on a specific issue or that they will be able to time the market's response/correction is likely in for a painful education...) 4) dumb = stupid, ignorant, or foolish. It's just not the case that you can't make a living on the stock market, if you have the right skill set. I've yet to see the report showing that the entire population of investors is absolutely limited to matching the same statistical distribution of returns which would be dictated by chance. Such an assertion, "You can't beat the Market", requires strong evidence. Why wasn't it cited here?

Yeah, I imagine that if I had the time to devote to studying a specific industry - like a full-time 60 hour a week job amount of time - I could beat the market. Very few people have that kind of time. A really good analyst would likely be going to industry conferences and meeting industry reps and sales people getting a really, really, really good idea of what products were coming down the pike and what was selling and which companies were well managed. And one would have to very narrowly specialize. I'm sure people like this exist of course, because I see their articles appear in the "news" ticker. But it's impossible to tell which ones have actually done the research and which ones are just faking it without actually doing the research yourself. The best I can do it hope to pick a managed fund that has analysts who actually know what they're doing.

I also imagine that IF you're a really good analyst, that it might be difficult to get rewarded and promoted up the management hierarchy because , due to random statistical chance, it might take years to become apparent that you're actually predicting things better. Why bother if you can get away with faking it?

The magazine, Institutional Investor runs an All American survey every year of Wall Street analyst, strategist, economist, quants, etc..

In virtually every category the same people seem to repeat year after year with only a few newcomers each year. Interestingly, most Wall Street firms give them big bonuses for making the list.

The people doing the ranking are the portfolio managers, etc., that use the research and recommendations of the various analyst.

That contradicts data that says every four years the top names of "hot funds" or "out-performing funds" are no longer top. Since the All American survey is not published except to subscribers, it's hard to verify your claim. The only way to explain your finding, if true, is either the All American perennial winners are heads of departments (i.e., figureheads who don't actually do any work but have hundreds of people working for them, and at least one of those hundreds has a hot hand), or they have inside information.

The All American list is not of portfolio managers. Rather it is of research analyst.
They are generally organized by industry, so it is a list of managers favorite auto, semiconductor,software, insurance, banks, drug companies, etc., etc, analyst. Supposedly it is a list of the analyst that provide managers the best buy and/or sell recommendations.So if year after year one person is voted the best bank stock analyst there is reason to believe that he has superior information and insights into the performance of bank stocks. Remember, roughly half of a stocks performance is due to the market,around a third is due to industry factors and under 20% if due to the individual company or stock.

Yeah, I imagine it's a very arcane world. It's funny how everyone knows who the top basketball performers are, but nobody knows or cares who is actually good at doing market research, which is worth real money.

Frank Rosenthal's biography (partially made into the movie Casino), describes exactly this approach to placing bets on college basketball games. Lots of research into what the conventional wisdom is and into what is really happening (he paid off a lot of students to spy on star athletes). When a large difference between reality and CW is seen, you call people who have money and tell them to go in big.

He protected his reputation by not betting on matches, or advising people to bet on matches, that he didn't thoroughly understand.

Good point about knowing what the conventional wisdom is.
You don't just have to know what's really happening, you have to know how what's really happening is different from what people think is happening. Only place bets when your sure that the CW is wrong, based on real data.

Thing is there is always information to be gleaned by putting pieces together, and only someone who is doing real research is going to be able to collect all the pieces. That's the only way to have information that isn't public.

" The best I can do it hope to pick a managed fund that has analysts who actually know what they’re doing" [SNIP]

Cowen: "In the first video in the Personal Finance section of our Principles of Macroeconomics course we pointed out that mutual fund managers do not beat the market on average"

'F' for comprehension. But, hey, its your life savings!

Analysts, not managers.

Instead, huge sums go into obtaining and mining data, instead of trying to understand what's going on.

If you look at the crude birth rate data for the Faroe Islands, in almost every year between 2000 and 2014 the Faroe Islands has a higher crude birth rate than Norway, Denmark, or Sweden, despite the Faroe Islands' skewed sex ratio. There is more variation if you look at the year to year change in CBR, but its not so dramatic as to make the measurement totally unreliable:

Good stuff. Nicely done.

Cool story bro.

u mad bruh?

No bro. The dog vomit visual did not upset me, we cool.

So, is this suitable for 16 year olds, in light of a certain non-profit putative start-up's long range goal of forming forming partnerships at the high school level?

Ok, I'm waiting for the part of the series that explains bubbles and crashes.

If stock prices adequately reflect all public information, what information is it that causes the entire market to suddenly reverse itself in 2008? Maybe "information" is not well defined. Some information upon which the market price is based can be both incorrect and widely believed mythology. Such as (say) passive index funds will always outperform managed funds. It's those large mythological stories that cause systemic risk. If everyone believes in Peak Oil, then Oil is GOING to have a price bubble. If everyone thinks that housing markets are uncorrelated then people will be wrong about the risk of MBSes.

What we really need is a theory of popular mythology. Why do large numbers of people come to believe the same things at the same time?

The EMH (or EMT, as I was taught) has three components: weak, semi-strong, and strong form efficiency. If you subscribe to semi-strong (I know almost no one who believes weak form efficiency is false), then the availability of private information in the marketplace (see strong form) may be impacting prices.

I think Fama, et al are correct re weak and semi-strong as almost no one believes you can beat the market using old information and as the video simply demonstrates, new information is priced into the market almost instantly, but strong form efficiency is harder to prove. In fact I think if they stripped strong form from their hypothesis, it would be more readily accepted as a theory.

1987 is probably a better example than 2008. In 2008 before the crash, even most highly informed people underestimated how closely Lehman Brothers' fortunes were tied to the housing market. Lehman Brothers was a major participant in derivatives markets and in the commercial paper market so once knowledge of its failure became public, investors correctly feared that other banks, non-financial firms that relied on commercial paper and anyone who owned a money market account were also at risk and that a worldwide recession could follow.

Knowledge of Lehman's failures were public for a long time before the crash. Bear Stearns was before them. The crash happened when the government didn't bail out Lehman. But I think also, even before that, there must have been some point where investors realized that mortgage backed securities (and the derivatives associated with them) weren't safe investments. How did that happen?

There is a video of Peter Schiff giving a speech to a mortgage bankers association meeting in 2006. I think it was in Phoenix, probably as ground zero to the collapse as anywhere could be.

He followed another analyst who talked of continued growth and that the slowdown at that time was temporary and it would revert to trend.

What information? In retrospect it is obvious. But at the time Schiff was (and still is) the wierd goldbug Austrian crank.

The stock market is a closed box with almost no opportunity for information arbitrage. also a favorite of economists because there are reams of data.

One investor who saw the collapse and profited dug up information. One bit was that the majority of sales in some Florida markets are to builders and construction workers. Two flags, speculation and if it corrected the default level would be far higher than estimated.

Was that common knowledge? It was available, as were the prospectus' of the structured finance products. Opaque and ripe for information arbitrage.

Yeah, they made a movie about that one guy--Batman (C. Bale)--the screen adaptation of M. Lewis' The Big Short. Good movie, and showcased in a brief scene the up-and-rising (literally, at least for me) star M. Robbie, who plays H. Quinn in the comic book movie Suicide Squad.

Yeah, I saw The Big Short. It was good, but I think still missing pieces of the puzzle. IIRC it got the part about the ratings agencies being asleep at the switch correct, tho. Been a while since I saw it.

@Hazel: some stock pickers refer to fundamentals as information and everything else as belief. The house price/rent ratio was very bad in 2007. These metrics/information were and are public but not reflected on prices because there are more compelling narratives: own the place you inhabit, the house according to your taste, etc.

There is a big problem with fundamental values, the conditions in the real world that control them evolve over time. Some technological change may change everything and make it difficult to make a new valuation of a certain stock. However, housing markets are well studied and understood by now......the problem is the people that belief consciously or unconsciously that this time is different. Very few people stop actions to ponder on "why am I taking this decision?".

Also, from my peasant level of awareness I see friends dedicating more time to choose the paint color or which curtains will look terrific in the new house compared to reading the mortgage contract. It's not stupid people, it's just that they are looking at the sky while they should be looking at the next footstep.

"Positive upward drift" -- even better than a "permanently high plateau".

I wonder if it's all that informative to ask the question about individual performance versus various aggregate institutional results. If we asked the question "can you beat the market" in the sense of getting a better than average price for a new car, or a can of soup, how well that that play out. I suspect one reason we don't look at the individual performance and how efficient markets are in these contexts is largely premised on the idea we al buy at the market clearing price rather than there really being a bunch of infra-marginal/off margin transaction (though clearly there are).

Perhaps relevant to this general approach I heard a few years back that the guys that ran the SPY fund didn't really invested in all the 500 securities, they always invested the funds in the top performing subset and that was how they were really making their money running the fund, not off the fee. I would assume there's a lot of funds flow in their activity so seems to imply that they must be betting the SP500 on a pretty regular basis if they are really earning their required return via management fees for the ETF.

It's far harder to look at EMH in consumer goods because there's a whole lot more variables that make comparisons difficult. For starters, different people have different ideas of value, and, as such, you can't really compare across items: If my value function is very different from most people's when it comes to cars, it might look like I am buying cheap. There's also far more work doing arbitrage: The cost of buying an item that is heavily on sale, waiting a few days/weeks/months, and then unloading it at a higher price is not something anyone can do cheaply and efficiently, but in practice, that's what we do with stocks.

Arbitrage is indeed the main factor. EMH only works, if it works at all, in very large, very liquid markets with almost instant arbitrage.

Counterpoint: Josh Barro points out that his NYTimes article on taxi medallion prices that relied solely on public information plus a simple email exchange with city officials in Chicago caused a NASDAQ-traded stock price to decline by 7%.

Exactly. In short, the Fractal Markets Hypothesis beats the Efficient Market Theory. But, since it's impossible to figure out which Levy-Mandelbrot curve fits the stock market at any given time (there are many variants, depending on the parameters chosen, and in fact the Gaussian 'normal' distribution is a special case of a Levy-Mandelbrot curve), the EMT is popular since you can easily write a computer program for it, as they did for the "value at risk" models, which work well in 'normal' times, but not in chaotic times.

New information affects stock prices for sure, but the idea that stock prices quickly reflect the correct magnitude of the information is clearly wrong and the video doesn't make the difference clear. Its an extremely low bar to say that new information causes some impact on prices. Even in the case where the correct company fell more than the incorrect companies, that still isn't anywhere close to saying that the magnitude of the fall was correct. Was there a continued downward drift in said company all the way through the public hearings?

+1. Momentum.

I also question how closely many companies are followed, as well as noting that there is much "technically" public information that in practice may be more like non-public information. Listen to any conference call and you'll often hear mgmt answer and analyst question by saying "we'll get back with you on that offline." There are conversations going on that are not public. Additionally, mgmt may meet and present at a conference and unless you happen to be an analyst there, you likely won't hear the conversation, although sometimes you can find the slides. So I guess part of the question becomes how quickly does this type of information leak into public consciousness.

From my perspective, I also doubt many investors have a precise value estimate of the shares they hold or evaluate that on a very frequent basis. I think it's much more fuzzy.

I think everybody detracting from the EMH is kind of missing the point. There will always be instances where there are post hoc inefficiencies. But there is a giant profit motive to get the price right. SO the market is efficient in the sense that getting the answer right will make you more money i.e. there isn't market failure.

Comments for this post are closed