Efficient Market for Information

Here's a question that has plagued me recently. Do we want our collective market for information to be efficient? The answer to this is not really clear. First of all, this post analyzes the current landscape through the efficient market hypothesis (EMH) lens. The EMH says that all the available information in the economy is reflected in the prices of assets. As a result, it is impossible to "beat the market." After all, no one individual can predict returns because no one individual has all the available information of the entire economy. Here's a concrete and simple example of what "beating the market" means. Suppose the returns for one stock of Amazon is 10% and the market returns are 8%. In that case, an individual with an Amazon stock has "beat the market" by 2%. This can get more technical than this, but this definition should suffice. Using this lens in our current landscape of "fake news" and "conspiracy theories," we ask the question again: do we want our collective market for information to be efficient? Most individuals may answer in the affirmative for game-theoretical reasons, i.e., I want to have access to all the best information to make the best possible decisions. However, for many other reasons, power is one; some individuals will respond in the negative. Let us take both aspects of the debate and think through their implications.

An answer in the affirmative implies a normative statement that the returns to a truly efficient market for information should follow a random walk. In other words, if efficient, the returns to the information derived from the market of information should yield returns that essentially follow a random pattern. Thus, no one person should predict the returns and capitalize on them. An excellent example of an inefficient information market happened when elected officials sold their stocks before the crash in the financial market brought on by COVID19. The most significant question mark hanging over this side of the debate is the question of what an efficient market of information even looks like.

I believe that an efficient information market must take on some aspects of a free market, i.e., small firms that compete for many buyers. This still leaves the puzzle of whether we want our market for information to be efficient. From my vantage point, this is the game-theoretically wise point of view. It removes many moral hazard problems from an inefficient market for information. In other words, if information, broadly speaking, is as accurate, easily accessible, and evenly distributed as possible, no one individual should perform better than the aggregate.

An answer in the negative implies a descriptive statement that the returns to a truly efficient market do not follow a random walk and are thus predictable, especially in recent times - see covid. In fact, we can almost certainly say that there is a predictable pattern to the returns from the information market. Much of these can be seen in how both the right and the left decries the other and the vitriolic political discussions (although I am not convinced that this is a uniquely 21st-century phenomenon). The biggest question mark hanging over this side is the question of extremes. In other words, is the market for information more inefficient at the tail end of the distribution? Or is it a mean issue? A good test for this and the efficient market hypothesis, in general, is to check the variance. At the tails, the variance in opinion between both sides is definitely not 1.

Both sides are correct in approaching this issue; the one difference is time frame. Usually, descriptive statements differ from normative ones in their time frames and certainty. Descriptive are certain short-term statements about the present, while normative ones take on a more long-term hopeful approach. In political discussions today, we should be much less eager, on both sides, to claim that what we want is an efficient market for information when we do not act in accordance with this principle. Those in the affirmative would usually agree to proposals about increasing privacy, mainly for security reasons. Those in the negative would also not hesitate to use things like Google and the computer programs that keep the system running. These systems are built on a complex web of programmers and an efficient information market. We should do more to examine our views on what we really want in this domain.

Certainty Rating: 82%


Recent Posts

See All