Something to Consider
We have had a run of extremely close presidential elections, which is a break from the past, which saw the sort of large landslides which are unimaginable now. The notion of any candidate winning by 18 percentage points and winning all but one states, as in 1984, seems impossible. An easy explanation to jump to is a sort of efficient market hypothesis for political parties. The two entities are competing with each other, and they’re both equally capable. We might reasonably expect them to win exactly half the time, and for the vote margin to be close to each other. I don’t think this is actually true, though. There is no reason to expect the efficient market hypothesis to apply.
The efficient market hypothesis, or EMH, in its simplest statement is the idea that the future path of a security’s price cannot be predicted from its past behavior. It is a martingale, in other words. If it could be predicted, then profit driven traders would find it and move the price to where it will be immediately. The price of a security reflects all known information about the world. You cannot, then, expect to consistently beat the market. The assumptions needed for it to perfectly reflected are no transaction costs, costless information, homogeneous traders, and rational investors (with the last being that investors won’t enter into something they expect to lose money in). As you break these assumptions, EMH becomes less strong, but still directionally true. Acquiring information being costly, a la Grossman Stiglitz, simply gives a bound on how efficient markets can be — the last bits of information won’t be found, but this is itself efficient. Heterogenous traders, some of whom are irrational, is necessary to avoid the Milgrom-Stokey theorem, where no transactions ever occur, and so on.
Something to Consider is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
