Essays on Intermediation in Financial Markets

I study broker-dealers’ trading activity in the US corporate bond market. I find evidence of market making and of proprietary trading exploiting possible mispricings. Market making occurs when customers both buy and sell a bond in a day, which happens half of the time: as predicted by market making theories with adverse selection or inventory costs, prices go down (up) as customers sell (buy). Otherwise, evidence is in favor of broker-dealer initiated trades, i.e. proprietary trading: prices go up (down) when customers sell (buy). I test one aspect of proprietary trading predicted by theories of limits of arbitrage: dealers buy (sell) bonds that are relatively cheap (expensive) with respect to bonds of similar maturity, or with respect to Treasury bonds. These proprietary trading strategies are reduced after the crisis. Relatedly I show that before the 2007-2009 crisis, large broker-dealers borrowed and sold Treasury bonds in amounts similar to their corporate bond holding, but not after. I study why financial institutions trade forward and future contracts on assets they could buy or sell directly. I provide a dynamic trading model in which this occurs because of 1) imperfect competition and 2) uncertainty about future customer trades. Under imperfect competition, risk averse traders realize gains from trading inventory imbalances slowly, leaving them differentially exposed to a supply shock: sellers fear customers sells that depress prices, buyers fear customer buys that increase prices. Opposite exposure to the supply shock implies gains from trading the risk through forward contracts: in equilibrium sellers of the asset sell forwards to buyers, and risk sharing in the asset is slowed down. The cost of slower risk sharing is compensated by the benefit of more certain future trading surplus. Traders are more willing to create inventory imbalance with forward contracts, leading to tighter spreads and/or higher trading volume in fragmented markets. I study the opportunity for dealers, i.e. intermediaries in financial markets to open restricted markets parallel to a centralized, all-to-all market. In a dynamic trading model with imperfect competition, dealers have the opportunity to open a parallel market so that a restricted subset of them trades with customers. Dealers in the parallel market choose to have all customer trades in the parallel market, which makes both customers and dealers not trading in the parallel market worse off. Before dealers learn whether they will have an opportunity to trade with customers in the parallel market, they choose to open the parallel market, as long as the surplus from future transactions are sufficiently high compared with the cost of holding the asset until future transaction, highlighting the role of dynamic trading rent.Le résumé en français n'a pas été communiqué par l'auteur