Three structural empirical studies of games, advertising and market structure

The dissertation consists of three separate empirical papers. The first paper proposes a structural empirical approach for characterizing observed behavior from a dynamic game. The empirical approach leverages a continuous parameterization of the "cooperativeness" of player preferences to facilitate inference about how a dynamic game is being played. The parameterized empathetic preferences generate "as if" equilibrium paths to the dynamic game. Through this structural model of empathetic preferences I propose a measure of the cooperativeness of observed play in a dynamic game that can be intuitively interpreted. The paper includes an empirical application that studies advertising and pricing competition between Coca-Cola and Pepsico from 1968 to 1986. I find that pricing competition is more aggressive than previously reported in 1992 study by Gasmi, Laffont and Vuong. The proposed modeling approach also indicates that advertising expenditures were only modestly collusive.; The second paper examines bid-ask spreads for S&P 500 Index futures contracts for customers and for market-insiders. Estimated realized bid-ask spreads are smaller for customers than for market insiders when they act like customers and seek liquidity rather than make markets. The larger spread for market insiders reflects the timing of their trading with market movements and their desire to execute against customer order flow rather than against quotes from other market-insiders.; The third paper re-examines the Chicago Mercantile Exchange's 1997 decision to split its S&P 500 Index futures contract and Bolen, Smith and Whaley's 2003 analysis of this contract split. I find that differences in the bid-ask spread for customers around the contract split were driven primarily by changes in market volatility rather than by changes in market liquidity around the contract split. Overall relative volume continues to decline following the contract split. Volume declines reflect decreases in order flow from large, intra-day trading customers. Increases in the expense of trading based on fee increases and bid-ask spread increases as well as macro market shocks can explain the flight of large, intra-day traders from the S&P 500 Index futures market. Finally, although spreads expanded, volume declines led to total revenue declines for market-making in a higher volatility marketplace.