pull down to refresh
Yes, it requires insiders who are not actually insiders.
But I believe this is a real and, more importantly, arrogant category. There are loads of people who think they have the inside scoop, but really they are just deluding themselves. These people get washed out of the market quickly, but luckily there seems to be an endless supply of them.
The main difference is that people without inside information can still participate profitably in the stock market, since it's not a zero sum game. (Stocks represent shares in business ventures which generate future income).
In a prediction market, it's a zero sum game, so on average there isn't any point participating unless you have inside info. But if the only people participating are the ones with inside info, it lessens the incentive for insiders to participate.
I think your earlier point about there always being overconfident people to participate is probably true. Another related issue is insiders who have conflicting info. But if the market is populated primarily by overconfident people or conflicted insiders, then it's not 100% obvious to me that the market price shakes out to the correct probability.
lemons! It's a lemons market, omg!
I'm still unconvinced. If the insiders are known to trade in the markets, then there wouldn't be much incentive for non-insiders to participate. If non-insiders don't participate, then there wouldn't be much incentive for insiders to participate either (no one to take advantage of for gains.) Thus, there won't be enough liquidity to really get a good signal about event probabilities.
The "prediction markets reveal inside information" hypothesis seems to require the long run presence of uninformed traders who consistently lose money to insiders, which doesn't seem plausible at a large scale.