According to physicists, thermodynamic theory makes it impossible to get useful work out of the random thermal motion of atoms. This is usually expalined in terms of a demon: suppose you have two containers, linked by a pipe, in the middle of the pipe is a shutter, controlled by a demon. When a fast atom (i.e. a “hot” atom moves into the pipe, the demon opens the shutter, and allows the fast atom to enter the other container. In this way, one container becomes hot, the other cold. This situation would be an example of getting something (heat) without expending work…a violation of the 2nd law of thermodynamics.
Now, the stock market. Prof. Burton Malkiel calims that the stock market is like a heated gas…the movement of invidual stocks (like the atoms of gas)behave in random ways. He also implies that it is impossible to construct a “filter” (akin to Maxwell’s demon) which would pick “hot” stocks and reject “cold” stocks.
Do the concepts of thermodynamics apply to a system like the stock market? Could a “demon” pick stocks according to simple rules, and realize a profit? 
If the assumptions fit the model, then they’re logically bound to follow the predictions of the model. One economist who attempted to undermine the mathematical foundations of economics in his book More Heat Than Light goes into great detail about the parallels between the assumptions of the neo-classical consumer and the physical system of something in balance (I forgot the word). His problem, though, is that just because the assumptions are such that they fit a convenient physical model, it does not imply that they aren’t useful approximations.
The stock market is random according to many, including Malkiel, and because of that one can’t make the sort of fine predictions that chart readers and fundamental analysts claim to make. But that doesn’t mean that there isn’t an upward drift, which is a point Malkiel may have made to you (in whatever form you got his insights). That’s why he advocates a long-term, extremely diverse investment strategy; he advocates using a broad, long-range strategy to capture that upward drift in the market. To that extent, the statistical model of gas isn’t (or is, for all I know) appropriate.
I think another law of physics is more applicable here: the vaguely-interpreted rules of quantum mechanics. For instance, if you do have a stock-picking paradigm, as soon as it is generally known, the market takes that into account, sort of like you can only be sure of a particle’s position OR velocity.
As to whether it is possible to beat the market: of course it is. You just have to know which industries are going to be doing well 5-10 years down the road. It didnt take a genius in the mid 90s to know that technology stocks would have outsized gains.
Now, is it possible to beat the market when you take into account trading fees and taxes? Considering that even industries that DO do better than expected by the market encounter unexpected falls in stock prices that can last for years even when they are poised to bounce back, it makes it exceedingly difficult.
I believe that often the market is indeed mistaken about actual chances of a business sector, and even of a given business.
However, if you are talking about a purely market-based, quantitative analysis to determine which stocks to pick, I’d agree that it is pretty much impossible to beat the market ONCE you take into account fees and taxes.
I’d say that, in stock parlance, the Demon is working with inside information about the atoms picked that the world at large does not have. Give someone that inside information about the stock world, and they will most assuredly outperform the market. I believe that many people in power do so on a regular basis, legit corporate insiders and politicians are two groups I’ve heard about who consistently beat the market.
The problem of Maxwell’s demon was resolved very quickly once information theory came on the scene. I’d need to have my books with me to post the answer, so I’ll do it tonight if no one else has. Meanwhile, shouldn’t this be in GQ?
I’ve poked around in my books a little, and I think the above is as good an explanation as you’ll get.
Something more like rational expectations may be a better explanation, in that if a pattern emerges, then it will be exploited and arbitrage will push extra-ordinary profits back to zero. Paulos posits an informal hypothesis in his book A Mathematician Plays the Stock Market that profit seeking behavior on behalf of many market players is what guarantees the random & unpredictable behavior of stock prices.