I noticed Noah Smith and John Authers on Twitter discussing how great the Efficient Market Hypothesis is because it explains why indexing works. I responded saying that the EMH really has nothing to do with why indexing works. They didn’t seem to see my point of view so let’s try this again and see if we can finally get this industry on a path towards understanding how incredibly silly the EMH actually is.
The original EMH stated that markets priced information efficiently. That is, if some news was released the markets would price this so that it was incredibly difficult to take advantage of this new information. This was basically a way of saying that markets work better than people who intervene in a discretionary manner. It’s not surprising that this idea was largely developed by the Chicago School of Economics during the period of Milton Friedman’s reign of terror because this established a theory of finance that was consistent with a theory of economics that was inherently anti government (interventionist).
Since then, the theory of economics has basically collapsed as it’s been proven largely false. Monetarism, as it was originally constructed, is no longer taken very seriously. But the theory of finance has been heralded as a great success. The banner has been carried triumphantly by Eugene Fama and others. The problem is that this theory is also collapsing (though much, much more slowly). When Fama realized that there was some unexplained outperformance in certain types of assets he had to change the EMH. EMH went from being a strong form to a semi strong to a weak form. And his factor models evolved from one factor to three to five and now there are even hundred factor versions. The original EMH is a mere shadow of its former self and no one finds it all that useful. For some reason it is still attributed as a great success in describing why active portfolio managers don’t perform well and this still gives it credibility in many circles.
The flaw in the EMH was that it was always constructed on an ideological misunderstanding. Fama implemented hundreds of studies showing that “active” portfolio management didn’t outperform indexing. This seemed to prove that discretionary intervention was unwise and Fama claimed that it showed that markets are smarter than individuals who try to intervene in markets. Except it didn’t prove that at all. After all, all of the benchmarks he was comparing these active managers to were also actively comprised components of broader aggregates (like the GFAP which I have explained in the past). That’s right. The S&P 500 is a slice of US stocks inside a broader aggregate of global stocks with a specific set of actively chosen criteria. The same is true of any other index. Fama was not comparing discretionary intervention to non-discretionary intervention. He was simply comparing active funds with economies of scale versus active funds without economies of scale. He was comparing low cost funds to high cost funds. This said NOTHING about how “efficient” the market is. It simply showed that a high cost manager will, on average, underperform a low cost manager.
Of course, this should be patently obvious to anyone who understands the arithmetic of the markets. As I’ve explained before, the financial markets are like a basketball game where the teams score an average of 100 points and are all docked 10 points by their coaches and then 20 points by the league at the end of the game. When all is said and done the average score is 70. Isn’t that crazy? The best basketball players in the world only score 70 points relative to a 100 point benchmark! They must be terrible at basketball, right? Of course not. This is the same for investment managers who earn 10% on average, charge 1%, incur 2% in taxes and are then compared to a pre-tax and pre-free benchmark. It’s not surprising that they underperform on average. It’s just math. And it says nothing about their skill level or whether their active decisions are smart or not. It just shows that the less active low friction manager will, on average, do better than the highly active high friction manager. The key point being that all portfolio construction involves active and discretionary decisions and when everyone starts using low fee and tax efficient approaches then SOMEONE must, by definition, be making smarter discretionary decisions than others which lead to better performance.
And that’s the big kicker here. The EMH doesn’t tell us anything about the skill of the S&P 500 Index Committee relative to the Fidelity Magellan fund manager. In fact, the EMH doesn’t really tell us anything useful at all. The only thing an indexer needs to know about why they should use index funds is that costs and taxes matter. A LOT. More importantly, all asset allocators ultimately end up being forced to actively choose how their portfolios will be comprised even if they pick the index funds that comprise that portfolio. Using low cost index funds are very smart ways to do this, but using low cost index funds will not help you circumvent the inevitable reality that you will need to make discretionary decisions at times and you will need to use funds that were constructed by people who chose, with discretion, how those funds would be constructed. And you will, by definition, deviate from global cap weighting which will force you into being an “active” asset picker.
When you understand that the EMH doesn’t explain why indexing works then it becomes little more than a meaningless idea about why prices may or may not move when they respond to new information. It doesn’t prove if the response to that information is “efficient”, rational, intelligent or ignorant. In essence, it simply says that prices react to new information, which, to be kind, is not very useful for any sort of financial or economic analysis.
While the Chicago School made many great contributions to modern finance over the last 50 years the EMH is one idea that cannot die fast enough. It is a tired old political idea with no real use in modern finance. My hope is that, while it dies, it takes the myth of “passive indexing” to its grave with it. Unfortunately, given how deeply ingrained this idea is in Modern Finance, I doubt it’s going anywhere fast.