Market Efficiency and March Madness: Empirical Tests of Point Spread Betting
47 Pages Posted: 7 Feb 2003
Date Written: December 23, 2003
This paper presents the first empirical documentation of the entertainment value of gambling in national point spread gambling markets and the first axiomatically sound, efficient, and rational model of point spread gambling markets. I use asset classes to test National Collegiate Athletic Association Basketball Tournament point spread betting market efficiency via latent class analysis and Ordinary Least Squares (OLS) regressions. In sample, I document efficiency except for lower seeded favorites, which may suggest additional pricing factors like entertainment value. The point spread's explanatory power has increased from the early years, but diminishes as the tournament progresses. The point spread has more explanatory power than seeding dummy variables over both margin of victory and win incidence. I find a structural break in the OLS coefficients. Higher seeded teams' ability is more homogeneous.
Note: Previously titled: "Does March Madness Exist?: The Favorite-Longshot Bias versus Market Efficiency Using Point Spread Betting"
Keywords: Asset Pricing, Gambling, Betting, Sports, College Basketball, Market Efficiency, Favorite-Longshot Bias, Point Spreads
JEL Classification: G12, G14
Suggested Citation: Suggested Citation