High Prices and Large Spreads: The Effect of Transaction Costs and Excess Choices in Prediction Markets

Looking at the Academy Award for Best Actor we can see that the current offers add up to 218.2, very overpriced. Since the prices of the contracts are approximately the percent chance of the contract paying off we can tell the market is overvalued. If everything was priced accurately the percentages should add up to around 100 (it will always be a little more due to the spread). Think of an election with two candidates; there will only be one winner. The chances of success for both candidates should add up to 100%, e.g. Candidate A has a 30% chance of winning so Candidate B should be at 70%. The highest chance of someone winning is 100%, so if the market is factoring in more than that it is over exuberant.

Since the market is factoring in over 200% chances of winning the Best Actor Academy Award there should be several opportunities to short the market. Unfortunately when we look at the sell side, the spread is so wide that the opportunities disappear. The Bids total up to only 92%. Not only are the spreads wide but in some cases the other side of the trade doesn’t exist. For instance on Crowe as Best Actor the Offer is 4.9 with 0 on the bid. What is happening here is that the chance of Crowe winning is so small that there is no incentive to take the other side of a guaranteed losing trade. In theory the contract can trade low enough to generate interest, but because of transactions costs it isn’t going to happen. No one can make money selling at the real odds so the market ceases to exist.

That leads to the next reason why the huge disparity between bids and offers exists; too many betting choices. With so many options, each with their own bids and offers, the market has trouble closing the gap. For instance, the GOP Nomination has 26 different candidates that you can take a position on. With volume of over 80,000 contracts the market is fairly deep but the spread between the bid and offer has stubbornly refused to close and remains at 21.3 (Bid 96.6 – Offer 117.9). Similar to our trader who will sell Crowe at 4.9 but wouldn’t buy at any price, the long shots skew the market. The more choices, the wider the gap between the bids and offers, even when sufficient liquidity exists. Looking at the Canadian Elections, a trade with only three choices, we can see more rational pricing. The spread is only 7.25 (Bid 98.2 – Offer 105.5) even though liquidity per contract is substantially similar to the GOP Nomination (3,076 GOP vs. 3,259 CAD contracts traded).

Not only is contract volume a factor in closing the spread between bid and offers, the number of choices are as well. However we expect the market to rationalize as the expiration approaches and the smart money begins to pay more attention.