Alternative Trading Systems' impacts on depth and informed trading

This study investigates the impact of Alternative Trading Systems (ATS) on depth and the probability of informed trading (PIN). The probability of informed trading and VNET, a metric of depth that determines the net directional volume necessary to induce a price change or return, are estimated for volume-matched samples of New York Stock Exchange (NYSE) and Nasdaq listed firms. Trade and quote data are obtained for two time periods before, and two time periods following, the introduction of trading in decimals on January 29,2001 and April 19, 2001 for the NYSE and Nasdaq, respectively. The data are used to estimate VNET and the probability of informed trading through time for the NYSE and Nasdaq matched portfolios, separating the impact of decimalization on VNET and the probability of informed trading. Estimates of the immediate effect of decimalization on market depth show VNET decreased on the NYSE from predecimalization 2001 to post-decimalization 2001. However, decimalization caused no immediate impact on depth on the Nasdaq. Over the year following decimalization, VNET increased on the Nasdaq, but decreased on the NYSE. The impact of A TS on depth for Nasdaq listed companies has been significant. Higher proportions of quote activity through ECNs are related to increased depth. On the other hand, higher proportions of trading through SelectNet are related to decreased depth· and increased spreads. I find a positive relationship exists between the probability of informed trading and the proportion of trades from SOES and SelectNet, and a negative relationship with ECN quotation activity. Therefore, informed traders' decisions of order execution venue are partially determined by differences in trading protocol between ECNs, SOES, and SelectNet. Estimates of the probability of informed trading for matched portfolios of NYSE and Nasdaq listed companies suggest that adverse selection costs are lower on the NYSE.