Lauretta wrote: ↑
Sat Oct 13, 2018 4:39 am
In a discussion yesterday where I was trying to articulate my thoughts on EMH, I wrote that since EMH cannot be tested because it has no predictive power (due to the joint hypothesis problem - meaning that you will never kown whether it's EMH that is false, or the asset pricing model used that is false), it has no scientific value as a hypothesis: as Karl Popper noted, scientific statements must be falsifiable to be meaningful.
Since the response of several posters was that I did not understand EMH, I researched it a bit further and came across this post, saying the same thing:
https://obliviousinvestor.com/testing-e ... s-problem/
In other words, one will never
be able to prove that EMH is wrong, because it is always possible to 'blame' the asset pricing model, and say that it's the latter that needs improving.
The author of the post is regularly invited to speak at Bogleheads meetings, so perhaps his opinion will give more authority to this statement.
To be sure his piece concludes that since EMH cannot be falsified, one should be wearly of those who say they proved it's wrong.
But my claim is that precisely because it cannot be tested and thus is not falisfiable, it has absolutely no scientific value. It is just a tautology and I am more and more puzzled why people (besides academics) take it seriously.
In fact even Fama thinks that the momentum effect makes EMH very dubious, since one would need a particular crazy asset pricing model to justify the quickly shifting risk in momentum stocks. I saw a video in which Fama joked that he hopes that the momentum effect goes away, so that EMH will become more believable (or less incredible). Perhaps this gives an indication of how EMH should be taken, light heartedly and more as a joke, rather than with the seriousness and dogmatic attitude that many (who aren't equipped to think of the grounds on which it is built) have?
EMH as I've said is one of the most misunderstood concepts on this website. But I think you are right. The way that EMH is often understood is "near tautological". It's hard to conceive of a situation where it would not be true, except of course the situation where a market is so inefficient that the masses are capable of making money from inefficiencies they see.
I don't entertain efficient market theory at all, at least not in any academic well understood form. At such a high level of generality it becomes a useless and meaningless philosophy. It is knowledge that does no work, provides no benefit. Tautologies are useless. They are ambiguously self-proving, allowing no cogent set of deductions to made from them.
The point at which the idea of an efficient market can work for you as a constructive piece of knowledge is to ask "how efficient is the efficient market?"
Consider the Stock market vs your local real estate market, the collectibles market, the market for books, the market for antique typewriters, the market for used cereal boxes. Each of these markets operate at a different level of efficiency. Each of them has a different set of barriers to entry. Each of these have major players and actors. Each of them has a daily volume, a finite number of transactions which occur daily.
I think the most misunderstood point about the stock market is that it is efficient, thereby abnormal profits cannot be made forever. First and foremost, in a strict sense that cannot be true. And those who follow other markets understand why. It's not that the stock market is efficient, it's that it is much more efficient than your local used cereal box market. There is no point where a market gets to be deemed "efficient." It's just an arbitrary point based on emotions, gut checks, perspectives, etc.
Is my local used cereal box market efficient? Let's say the reasonable answer is no. Well why not? There's so few people participating in something so odd, that if I make profit in it, it's cause there weren't enough eyes watching the closing seconds on that vintage 1979 box of cheerios on ebay.
Does that make it "inefficient?" Again the demarcation of inefficient is nothing but an arbitrary point.
If one gets used to the idea that all markets are really on a spectrum of inefficiency and efficiency, you can get some work done. You can say to yourself, I know my efforts are likely wasted in trying to predict the market. It's too efficient of a market for me to make proper headway. My stray thoughts about a hot stock tip I'm getting from a neighbor probably aren't enough to counter the legions of analysts doing this for a living.
On the other hand, it allows you to raise the question to yourself, "what markets are inefficient enough that with my level of insight, I can make abnormal profits?" That's where the idea of an efficient or inefficient market can start to help you out, do real work for you.
That's the right question to ask.