02
Sep
2014
More EMH Silliness
I love this stuff. Scott Sumner scoffs at those thinking Robert Shiller contributed anything useful by warning of a stock bubble in 1996 or a real estate bubble in 2003. He quotes from Fama, who at first seems to blow Shiller up with some numbers. But I’m pretty sure I show that Fama’s arguments make no sense. Since he delivered them in his Nobel acceptance speech, that’s kind of a big deal.
Anyway, it’s all here in my latest Mises CA post.
Nice article.
Yeah, that comment: “The value to homeowners from housing services during those nine years from July 2003 to March 2012 surely exceeds 6.7% of July 2003 home values.”…wow.
Imagine saying something like that to explain away volatility in stock prices: “Surely the services from owning eco-friendly stocks during October 1987 exceeds the 25% price decline. No bubble.”
How can you make money spotting bubbles? If it was easy to do, bubbles would be arbitraged out. If a couple of people who followed the ABCT were making a killing, then wouldn’t others follow and iron out the fluctuations in the stock market? That doesn’t mean ABCT isn’t valid, only that it is impossible to say when a market is overvalued or undervalued. Even though the EMH sounds a little silly, it can be a useful way of looking at the world. It teaches you solid investment ideas like avoiding paying for active management, buying and holding indexes, not trying to time the market, etc. I think Austrians shouldn’t be so dismissive of the EMH because it reflects Hayek’s idea of dispersed knowledge being far more powerful than the expertise of one individual.
John Becker wrote:
If a couple of people who followed the ABCT were making a killing, then wouldn’t others follow and iron out the fluctuations in the stock market?
NO! Absolutely not. What would happen is that they would make a killing, and Scott Sumner would say they got lucky. I’m not kidding. If the stock market crashes next week, Sumner, Fama, et al. will certainly not say, “OK Shiller / Spitznagel / Soros, you were right…” They will say, “They couldn’t have KNOWN the market would crash, otherwise it wouldn’t have been so high in the first place.”
That doesn’t mean ABCT isn’t valid, only that it is impossible to say when a market is overvalued or undervalued. Even though the EMH sounds a little silly, it can be a useful way of looking at the world. It teaches you solid investment ideas like avoiding paying for active management, buying and holding indexes, not trying to time the market, etc. I think Austrians shouldn’t be so dismissive of the EMH
John I urge you to look at my introductory paragraphs again. I love the EMH. I think it is a great intellectual achievement. I also think it is empirically wrong, and disastrous in the hands of certain economists who write for the public.
Bob,
I think Sumner does make two convincing points that I haven’t heard you address: the fact that there has to be someone who gets really lucky like Warren Buffett (everyone chalks it up to skill) and the fact that housing markets in other countries like Australia heated up just as much as the U.S. housing market but never suffered a big decline. It’s possible that given different policy, there wouldn’t have been a home price decline like there was in the U.S. I still think you’re a much better economist than Sumner but he has interesting things to say sometimes.
“It teaches you solid investment ideas like avoiding paying for active management, buying and holding indexes, not trying to time the market, etc.”
But how do you know these are “solid investment ideas”? A: Because EMH says that they are. In other words, EMH teaches us investment ideas that it deems “solid.” Of course, by definition. These ideas work, except when they don’t. Investment implications of EMH are helpful to know and perhaps the best strategies for certain people at certain times. But is this the strategy hedge fund managers generally follow? Just because EMH says that timing the market is impossible, doesn’t make it so.
Wrong Brian. It’s because extensive empirical studies by people who do not believe in the EMH have shown an inverse correlation between what people pay for stock pickers and the total returns they get. Read one of John Bogle’s books to see some of the studies.
Are you saying John Bogle doesn’t believe in EMH?
But more to the point, the empirical studies show an inverse correlation. Yes, of course. That’s well known. As I said, these are useful strategies for many people in many situations (perhaps even “most”). But EMH (at least the “strong” version Bob is critiquing) claims that it’s never possible to beat the index – apart from sheer luck. Anything outside a certain average range is deemed mere statistical outlier. It’s the equivalent of telling Tiger Woods not to practice golf – 99.9% of people that play golf regularly never win a professional golf tournament. And those that do – sheer luck.
I might suggest Ken Fisher’s “The only 3 questions that count” that also gave me some great insights.
I would also recommend it to all who believe the EMH rules out the existence of bubbles. In my view the book is able to show that it does not.
To use the EMH to define the possibility of bubbles away, in my understanding means not understanding the building blocks on which the EMH is actually constructed.
The EMH should only say you cannot make above average gains on already priced in knowledge/expectations. However it doesn’t say anything if the priced in knowledge/expectations are correct.
Since no one will dare deny that lots of expectations about the future are wrong, there are lots of opportunities to bet against. And if a huge part of the market believes stuff that turn out to be wrong for whatever reason and are betting hard on it then people call this a bubble. Combine this with the “trend is you friend/herd mentality” and you should know all you need to know how bubbles are able to form, you only need some convincing trend setter (Like Central Banks for example).
What is a “rational” strategy for betting in Poker?
Answer is that nobody knows. Computers are notoriously bad at Poker, and not because of a shortage of processing capacity, but because fairly rapidly human opponents will discover a way to game the computer’s strategy and beat it. If we had a proven rational strategy then the computer would win (or the biggest processor would win).
And yet, some humans are better at Poker than others, so clearly strategy is involved, and it may well make sense for a group of people to pool their money and hire a really good player rather than all try playing it themselves (presuming they can agree on a mutually advantageous way to divide the profits).
It’s tricky. I am closer to Sumner on bubbles, but agree it’s hard to test.
If you told me that Robert Shiller (1) correctly predicted the last 3 superbowl upsets or (2) that he was managing a mutual fund that consistently outperformed the market, my default hypotheses would be that (1) he got lucky and/or identified a temporary inefficiency, and I’d still be better off betting in favor of the vegas odds rather than Shiller’s predictions in future superbowls or (2) he’s either lucky or has hidden some risk somewhere that I don’t see. I agree that those hypotheses are both capable of being mistaken and very difficult to falsify, but I don’t see a clean solution.
If you told me that Robert Shiller (1) correctly predicted the last 3 superbowl upsets…my default hypotheses would be that…he got lucky and/or identified a temporary inefficiency, and I’d still be better off betting in favor of the vegas odds rather than Shiller’s predictions in future superbowls…
That’s a pretty good analogy, J Mann. I agree with you here. However, notice that strictly speaking, that’s not even what we’re arguing about in this post, regarding Fama and Sumner. It would be as if Shiller predicted 3 of the last upsets, and then Fama said, “But actually, since there were some horrible calls in those games, I’m not really sure he predicted the upsets after all.”
Thanks – that’s helpful.
If I’m being as charitable as I can to Fama, then if Shiller claims there’s a bubble in equities, and equities somewhat underperform other assets over the next 5 years, that’s a different story than if Shiller calls a bubble and we end up with a sustained crash from which we never really recover.
Of course, to be charitable to you, that’s still better for Shiller’s side than if he calls a bubble and notheing happens.
J Mann wrote:
If I’m being as charitable as I can to Fama, then if Shiller claims there’s a bubble in equities, and equities somewhat underperform other assets over the next 5 years,.
Here’s what I think happened:
(1) Shiller claimed a tech bubble.
(2) A few years later, tech stocks crashed hard.
(3) People said, “Shiller called it, he’s a prophet.”
(4) Fama pointed out, “Wait a second, tech stocks went way up after Shiller called bubble. So we need to see what the crash looked like, compared to when he first called bubble, not compared to the peak.”
If the above (4) steps were all Fama/Sumner had done, then they would be great. I would congratulate them on their caution before rewarding people with kudos for calling “bubble.”
However, Fama went further:
(5) He looked at the market bottom, saw that it was higher than the price when Shiller first called bubble, and concluded that Shiller had in fact been an idiot.
So that’s what I was objecting to.
Further, the stuff with housing was just complete nonsense. With perfect certainty and a static equilibrium, you would expect house prices to be flat at a price that capitalized the annual rental equivalent at the market rate of interest. If your house price went down 6.7% or whatever over a 9-year stretch, you wouldn’t say, “Oh that’s OK, I got to live in my house for free.” Fama totally ignored the opportunity cost of the funds you invested in your house in 2003. He stated something matter-of-factly about the preferences of the home buyers that only makes sense if you ignore interest over 9 years. That’s kind of a big deal, especially in a Nobel acceptance speech pertaining to the subject matter for the award.