Essays On Market Microstructure And Return Predictability Of Mutual Funds

This thesis contains three papers. Each paper addresses a distinct research question and is implemented on a separate dataset. The first paper concludes that daytime auctions, together with market opening and closing intervals, contribute to the periodicity of the cross-section of stock returns. By applying the model of infrequent rebalancing, I show that model parameters fit the data for the after-auction intervals. I thus conclude that after-auction periods take over a large share of infrequent rebalancing and show that this effect is driven by the concentration of liquidity traders. Small, low-fragmented stocks heavily traded on the home market show the strongest evidence for infrequent rebalancing after the daytime auctions. The second paper sheds light on how traders allocate risk of stock portfolios in a trading day. Traders decrease risk before the market close. They do so by selling stocks with the highest marginal risk and buying stocks that decrease the risk of their portfolio the most. As our measure of portfolio risk relates to the one that clearing houses use for the margin requirements, we conclude that the risk-reduction behavior is driven by traders' reluctance to provide end-of-day margin contributions to the CCP. These trading flows in the direction of risk contraction distort closing stock prices. The third paper replicates and combines eight prominent predictors of mutual fund returns to obtain a composite, aggregate fund predictor. While only three of the eight individual variables are significant predictors of future fund performance in a multivariate setting, the composite predictor has strong forecasting power. A hypothetical quintilebased long-short strategy based on the composite predictor realizes a four-factor alpha of 6% per year. The performance spread is robust to different regression specifications, is similar for different size classes and investment styles, and persists over time. Our results p.