First off, let me state that I am a huge fan of markets. I think they are fairly semi-strong form efficient (not perfectly so), but can get caught up a bit with fads, etc. Inactive markets are particularly prone to such mispricing since there are not enough "arbitragers" (And here I am using Thaler's view that is not necessarily a risk free arbitrage) to "fix" items that are priced incorrectly.
That said, this just has to be wrong. Tradesports has a market for the likelihood of a Category 3 or above hurricane hitting various states in the US. Ok, so Florida is most likely. I will buy that. Then Texas, Louisiana, and the Carolinas. Ok, so far so good. But then it gets a tad wacky. MS and NY have about the same likelihood.
New York? Yes, NY has been hit before and no doubt will be again. But the odds of it happening are much less than that of Mississippi. For instance, it is estimated that there is a 26% chance of NY being hit with a category 4 or greater storm in the next 50 years). But given the odds backed out of the midpoint of Tradesports bid and ask, it seems almost a 10% chance THIS year which does not fit with "expert" predictions.
So is the market wrong? It does seem like it is. So why doesn't someone fix it? Probably not worth it. Depth of the market is small and transaction costs (spreads) are pretty high. Which is exactly the type of market where theory would predict inefficiencies to exist.
But on the other hand (can you tell it is late?), it could be that the market is made largely of NY residents who are net hedgers. Thus they are "betting" on a hurricane hit so that if it does happen, the payoff will act like insurance. These hedgers would be willing to settle for a lower return. (so maybe the market is efficient afterall ;) ) But it is too late (2:50 AM) for me to think that one through...lol...