What does it mean to “short” on prediction markets?
This is probably one of those questions that only geeks and PM aficionados are interested in … and fitting that description I took my own shot at answering the question.
My own view, fully expressed over at Usable Markets, is that no true shorting, as I understand the word, takes place in the world of prediction markets, although shorting at the conceptual level (as in, I want to bet against X), certainly does.
But maybe I’m wrong.
Also, if you’re going to Consensus Point’s upcoming NYC PM conference, in NYC, I’ll be there moderating a conversation on the experiences people have had implementing a prediction market at their company. Please stop by.
~alex
Most prediction markets are still binary: you can bet on whether or not X will happen. You could do real shorting once we have continuous markets (betting on the future magnitude of Y — not “Hillary will be President; $1.00 if so, $0, if not,” but “The economy’s growth will be higher under a Hillary Presidency than under a Bush Presidency; $1.00/percentage point increase in GDP growth.”)
Some stocks have a similar situation to what you’re describing. When an airline is involved in a labor dispute, for example, the stock stops behaving like it measures a stream of cash flows, and starts acting like it measures either a) the same stream, or b) nothing, because the company will go bankrupt. At that point, the only bets you can make are “Strike resolved” or “Strike not resolved,” unless you do something complicated with fixed income.