All of bjterry's Comments + Replies

Define binary, because my imagination does not come up with many examples that are not more like some linear events, e.g. amount of rain as opposed to "it rains". Also, in almost all of the cases I come up with, there is a benefactor. With rain people stay indoors and use more electricity, consume more television and hang out more on the internet. With a negative scientific result a product becomes unviable which ensures the market position of another market participant. And so on. Keep in mind that for a prediction market to work in an unbiased manner, there only needs to be a benefactor relative to someone that loses, that is total wealth can decrease, but not uniformly. Lastly, the rise of the second and third world will ensure that there are more than enough market participants for most questions, at least I think so. I have no model to predict this, but base this on a gut feeling. Maybe we can construct some examples to get a feeling on what point we disagree on?

I don't know how to reply on your blog, so I write my thoughts on a couple of points here. The bet on the event itself is already the insurance. Or viewed differently, the insurance itself is a bet in some kind of prediction market on the event. Which goes to show that it does not need two economic actors with opposite exposition to some event, as in your example of the baker and the farmer w.r.t wheat prices, but it is enough that two actors have different exposition to some event. Though of course in the case of non-opposite exposition there will have to be some kind of premium for one of the sides, meaning that the market price can not be interpreted simply as a probability of the event happening. Edit: The more I think about it, the more I have to disagree with the basic point about liquidity. For a proper market there needs to be unbiased participant and the person being affected by the event. And there always is the second person or the outcome of the event is without any meaningful economic consequence. Weather affects groth of plants, avenues visited by customers and traffic for transport of goods. Sports outcomes affect a populations happiness and their choices of activity afterwards. Scientific results make specific products plausible or impossible. And so on. If I am affected by a specific outcome of an event in a positive way, I can reduce my exposition by betting on the opposite outcome and increase my exposition by betting on the exact outcome. If there is a person being affected in opposite ways than I am, we can take the opposite sides of the bet. Else I need someone willing to merely take on some more risk than he is already exposed to, though the person will want to be compensated for that additional risk in some way. I wonder, is this distortion measurable so to recover the "underlying" probabilites?