Price Experimentation with Depth in a Specialist Market


Asli Ascioglu

Document Type



Request a copy of the paper from the author: Asli Ascioglu


We investigate how a market maker actively influences order flow and induces information from traders in a market that includes both traders with private information about the security and uninformed liquidity traders. Market maker?s learning from the trades, i.e., price experimentation, has been studied previously by considering the market maker?s use of only the bid and ask prices. We introduce a framework that takes into account the use of bid and ask depths, in addition to the bid and ask prices, in explaining the price discovery by a monopolistic market maker. We use a dynamic programming model to show that the market maker experiments with prices and depths in the earlier rounds of trading, then recoups any losses in later rounds. According to the model, the market maker only learns about the true value of the security when informed traders are trading. The market maker attempts to encourage informed trading in the ޲st round, when information asymmetry is high. Since informed traders prefer to trade in large sizes and liquidity traders prefer to trade when the spread is narrow, the specialist encourages informed trading by widening the spread and increasing the depth. In the second period, the market maker faces less information asymmetry, since prices already reflect the private infor- mation obtained in the ޲st round. Therefore, the specialist sets narrower spreads, making the market more attractive to liquidity traders. Thus, our model predicts that when information asymmetry declines, the market maker simultaneously reduces both spreads and depths. Further, as a consequence of the market maker?s actions, both depths and spreads exhibit large variations from one quote to the next. We test our model?s predictions with the NYSE data and find supporting empirical results for a specialist?s experimentation with depth.

This document is currently not available here.