Efficient Market Hypothesis
Some of the comments I received on my post about 401K discussions at my company eventually led to further and further research on the web, bottoming out in the Efficient Market Hypothesis (EMH) family of market valuation theories. Empirical analysis seems to suggest that at least some form of the medium EMH theory holds for the analyzed stock markets.
The theory is quite old, stemming from the 1960s, and the markets since then have developed with the knowledge of this theory, e.g. SPDR and other index funds are the direct result of this theory, as far as I understand it. I leave it to the economists to figure out how the introduction of this feedback loop changes or does not change the theory. (Though maybe one could claim that the information aspect of the theory would predict the emergence of such index funds as well ... :)).
The biggest problem with it seems to be that one has to by into the terminology of rational expectations, which while an improvement of requiring the assumption of rational agents. Unfortunately, the Wikipedia articles for EMH and rational expectations are written with each other in mind, making it impossible to decide whether one presupposes the other or whether they are mutually dependent on each other. Historically, it seems to be the case that rational expectations allowed the formulation of the EMH, but that's of course no argument against their mutual dependency.
Two things that I find compelling about the EMH/rational expectation tuple is that it allows random behavior on part of the participants (by positing that the overall behavior follows a uniform distribution) and it assumes that overall, all relevant information is taken into consideration. This allows the theory to predict super-star fund managers like Warren Buffett and might make it resilient against Keynes observation that people look at the behavior of others to determine their own behavior in the stock market (instead of economic fundamentals), because they are trying to predict whether the stock will rise or fall--which appears to be just more information that is taken into consideration by the market.
I just realized that I do not find the claim compelling that people cannot always act on the best information due to other constraints; the model only requires that the distribution acts on the best information and the claim merely is trying to explain why the behavior is not random, which the model does not require either, only that the overall distribution be as if the behavior was random.
The biggest puzzle in EMH/rational expectation seems to be the stock market bubbles and crashes, because they violate the claim that the market provides appropriate valuation for each asset. I wonder whether an additional assumption of uniform distribution, say, over the time of the market, would allow the theory to handle these spikes, just the same way the uniform distribution allows the theory to predict (someone like) Warren Buffett.
Suffice it to say that the aspects of the theory that appear compelling to me are reason enough to mistrust the claims of fund management and go for index funds and spiders spread around the whole of the global economy. If there is one thing the dot-com crash teaches us then it is that after the prior crash of the Tiger-states, the money had to go somewhere.
PS: Someone suggested that if American Social Security was not invested in US governmental bonds, but in Chinese governmental bonds, even a Republican US government would be more supportive of the baby boomers cashing in on the bonds.
The theory is quite old, stemming from the 1960s, and the markets since then have developed with the knowledge of this theory, e.g. SPDR and other index funds are the direct result of this theory, as far as I understand it. I leave it to the economists to figure out how the introduction of this feedback loop changes or does not change the theory. (Though maybe one could claim that the information aspect of the theory would predict the emergence of such index funds as well ... :)).
The biggest problem with it seems to be that one has to by into the terminology of rational expectations, which while an improvement of requiring the assumption of rational agents. Unfortunately, the Wikipedia articles for EMH and rational expectations are written with each other in mind, making it impossible to decide whether one presupposes the other or whether they are mutually dependent on each other. Historically, it seems to be the case that rational expectations allowed the formulation of the EMH, but that's of course no argument against their mutual dependency.
Two things that I find compelling about the EMH/rational expectation tuple is that it allows random behavior on part of the participants (by positing that the overall behavior follows a uniform distribution) and it assumes that overall, all relevant information is taken into consideration. This allows the theory to predict super-star fund managers like Warren Buffett and might make it resilient against Keynes observation that people look at the behavior of others to determine their own behavior in the stock market (instead of economic fundamentals), because they are trying to predict whether the stock will rise or fall--which appears to be just more information that is taken into consideration by the market.
I just realized that I do not find the claim compelling that people cannot always act on the best information due to other constraints; the model only requires that the distribution acts on the best information and the claim merely is trying to explain why the behavior is not random, which the model does not require either, only that the overall distribution be as if the behavior was random.
The biggest puzzle in EMH/rational expectation seems to be the stock market bubbles and crashes, because they violate the claim that the market provides appropriate valuation for each asset. I wonder whether an additional assumption of uniform distribution, say, over the time of the market, would allow the theory to handle these spikes, just the same way the uniform distribution allows the theory to predict (someone like) Warren Buffett.
Suffice it to say that the aspects of the theory that appear compelling to me are reason enough to mistrust the claims of fund management and go for index funds and spiders spread around the whole of the global economy. If there is one thing the dot-com crash teaches us then it is that after the prior crash of the Tiger-states, the money had to go somewhere.
PS: Someone suggested that if American Social Security was not invested in US governmental bonds, but in Chinese governmental bonds, even a Republican US government would be more supportive of the baby boomers cashing in on the bonds.