Using Time Series Methods to Assess Information And Inventory Effects in a Dealer Market in Il-liquid Stocks

Publication Date
Financial Markets Group Discussion Papers DP 242
Publication Authors

The purpose of this paper is to test for the existence of inventory control and asymmetric information in stock market price quotes, and then quantify these microstructure effects. The paper extends the time series work of Hasbrouck (1988, 1991) to the institutional setting of the London Stock Exchange. In contrast to the NYSE work our model and institutional framework enables us to deduce exact restrictions on the effects of public and liquidity-plus-private information shocks, within a simple bivariate VAR for price quotes and inventories. We show that the existence of asymmetric information or inventory control rests on the significance of precise functions of parameters in a single estimating system. We decompose price changes into a component due to the arrival of public and private news about fundamentals, and another component due to dealers desire to exploit noise trades and control inventories. We are able to assess the relative importance of public information and private information revealed through trades, on the change in prices. We test the model on trade-by-trade observations for fifteen relatively illiquid stocks on the LSE. Our findings are that both asymmetric information and inventory control are a robust feature of our sample of less-liquid stocks. This result accords with previous findings concerning NYSE stocks, particularly with regard to the speed of adjustment of inventories, the existence of a shift in their desired levels and the pervasive influence of trades on the long run level of prices through their role in revealing information on the stock's fundamental value. Further, those stocks in which in which we find microstructure effects are important, also display high quoted spreads, which is consistent with theoretical predictions.

Download