Trading the Tradesports NFL Market: An Analysis of Liquidity and Pricing Efficiency
25 Pages Posted: 5 Jul 2007
Date Written: June 2007
We investigate 1104 contracts traded on Tradesports' 2005 NFL prediction market. We find that liquidity as measured by bid-ask spread is around 2-4% of contract nominal value, and that lower bid-ask spreads are associated with higher volume. Price-takers' returns when trading 100 contracts ($1000 value) are only about 1% lower than trading a single contract ($10 value), indicating that the depth of the market is adequate in these volume ranges. Trading volume during the games is greater than trading volume leading up to the game, with most of this concentrated in the first half.
We find a favorite bias but one that has not been identified without the additional price spreads offered by Tradesports. Trading favorites is a profitable strategy, particularly for contracts in the middle price range of each point-spread contracts' typical price. Combining all contracts and examining trading on price, there are higher returns to buying low-priced/high odds favorite contracts and that returns decline as the price increases. Trading favorite team contracts with high price/low odds loses money. In the traditional definition, this is a reverse favorite-longshot bias as high odds trades have higher returns have low odds trades have low returns. Therefore, there is a favorite team bias combined with a reverse favorite-longshot bias.
Keywords: Liquidity, Prediction Market, Favorite-Longshot Bias, Efficiency
JEL Classification: G10, G12, G13
Suggested Citation: Suggested Citation