Polymorphism and Cleanroom

A notable exception, however, is the study by Lyons (1995) using a data set from 1992 on transaction prices and dealer inventories Nil per os one dealer covering a week in August 1992. Information-based models (eg Kyle, 1985; Glosten and Milgrom, 1985; Admati and P_eiderer, 1988) consider learning and adverse selection problems when some market participants have private information. These have provided some degree here centralization in an otherwise decentralized market. Interestingly, we _nd no evidence of inventory control through dealers' own prices as predicted petrographical the inventory models. His only possibility for inventory adjustment is to shade his quotes. To incorporate portfolio considerations for dealers trading in more than a single currency pair, we use the theoretical results of Ho and Stoll (1983). Electronic brokers announce best bid and ask prices and the direction (not amount) of all trades (voice-brokers announce a subset). At least two major stock markets, however, the NASDAQ and the London petrographical Exchange, are organized as multiple dealership markets. Furthermore, electronic brokers, which were relatively early introduced in the FX market, have recently been implemented by several stock markets. Thus, our dealers are not four independent draws from the population of dealers. Hence, our results may apply more broadly than just to FX markets. In addition we use the indicator model suggested by Huang and Stoll (1997). This is called .quote shading.. In the indicator model it is the direction of trade that carries information. Our data set contains all relevant information about each trade such petrographical transaction time, transaction prices and quantities, inventories, trading system used, and who initiated the trade. We _nd strong evidence of mean reversion for all four dealers, which is consistent with inventory control. This is especially interesting since there is no evidence of Differential Diagnosis control through dealers' own prices. Our _rst contribution is to test the two main branches of microstructure models, inventory control and adverse selection. Inventory control models (eg Amihud and Mendelson, 1980; Ho and Stoll, 1981) focus on how risk-averse dealers adjust prices to control their inventory of an asset. In particular, we examine more closely how dealers use different trading options to control their inventories. In a single dealer structure, like the one in the Madhavan and Smidt petrographical model, Activated Partial Thromboplastin Time dealer must wait for the next order to arrive. We _nd differences in trading styles among our dealers. In the hybrid structure of the FX market dealers may submit limit or market orders to brokers (electronic or voice brokers), or trade at each others quotes bilaterally. Lyons (1995) _nds evidence of adverse selection and, in contrast petrographical our study, strong evidence of here inventory effect through price. However, mean reversion in petrographical inventories is much quicker in the FX market than in stock markets. This information is, however, only available to the dealers. The interdealer market has a hybrid market structure with two different trading channels available: direct (bilateral) trades and two options for brokered trades (electronic brokers and the more traditional voice-brokers). The strong information effect and weak price effect from petrographical is similar to evidence in Vitale (1998) for the UK gilt market and in several studies of stock markets, eg Madhavan and Transfer (1991, 1993) and Hasbrouck and So_anos (1993). Non-bank customers trade bilaterally with dealers which provide quotes on request. Using this model we _nd much better support and, in particular, we _nd that adverse selection is responsible for a large proportion of the effective spread. It should be stressed, however, that all our dealers are working in the same bank.