The aggregate impacts of tournament incentives in experimental asset markets

Existing studies of the aggregate impacts of tournament incentives find that asset price bubbles in experimental markets are larger and do not dissipate with experience when participants trade under tournament incentives. However, these results potentially overstate the real-world impacts of tournament incentives for two reasons. First, they examine tournaments in a restrictive single-asset market setting, which constrains the risk-taking options available to traders. Second, by purely conferring additional rewards for good relative performance, the tournament contracts used ignore the risk-moderating role played by penalties that are also written into or implicit in real-world counterparts. We address these gaps by examining how prices behave under tournament incentives in experimental markets where participants can trade two assets with differing risk-levels. In addition, we compare price behaviour under tournament incentives with and without an embedded penalty for poor performance. Our findings suggest that the results in the existing literature are driven by the single-asset nature of their markets – we do not find any compelling evidence that prices in two-asset markets are more distorted under tournament incentives than normal incentives. Moreover bubbles in these markets do diminish with experience under tournament incentives. Also, while penalties embedded into tournament contracts are associated with less trading activity in markets compared to reward-only tournament contracts, they are also associated with longer periods of overvaluation and higher prices, albeit only with inexperienced traders. These results are particularly significant in light of the recent debate attributing financial market instability to convex incentive structures such as tournament incentives.