Comments on: Reporting prediction market prices

By: David Pennock

David Pennock — Mon, 28 Apr 2008 15:57:24 +0000

Byrne, while I agree “percent chance” glosses over many factors, I still think it’s the most succinct and understandable way to state the “prediction” part of a prediction market.

Mike, you may be right. Without the typo, the Reuters numbers would probably have been within 0.2% or so of adding up “correctly”.

By: Mike Linksvayer

Mike Linksvayer — Sat, 26 Apr 2008 21:01:41 +0000

The strictly factual description is great for educating people about prediction markets, but too long and requires too much parsing in nearly all “reporting” contexts I imagine.

Any market with a large bid/ask spread is probably too illiquid to bother reporting. If the market is liquid, the spread doesn’t add much information, again for typical “reporting” uses, e.g. everyday financial reporting uses last trade.

Better market technology is badly needed in any case.

By: Byrne

Byrne — Sat, 26 Apr 2008 15:28:20 +0000

The problem with this kind of arguing is that there are factors other than the odds, and those factors are hard to determine. The confidence people have in their ability to get their money if they win, the speed at which the exchange clears transactions, the legal status of gambling, the tax status of their gains, and the possible returns of other investment opportunities, all play a role in determining prices.

Instead of a footnote to this effect under every mention of the ‘odds’ versus the price, it would be better (in my opinion) to just state the outcomes. When the media talk about stock options, they say they give the right to buy the stock at X price on Y date, giving a particular profit if by that time the stock has risen to Z. They rarely take us through the Black-Scholes. Somehow, this is enough to get the idea across, but doesn’t stop professionals from using more powerful and precise tools.