Strategic Trading, Liquidity and Information Acquisition

We study endogenous liquidity trading in a market with long-lived asymmetric information. We distinguish between public information, tractable information that can be acquired and intractable information that cannot be acquired. Besides information asymmetry and noise, the adverse-selection spread depends on the diffusion of intractable information and on the interest rate. With endogenous liquidity trading, efficiency is lower than that implied by noise-trading models. Liquidity traders benefit from the information released through the insider’s trades in spite of their monetary losses. We study factors that affect the insider’s information acquisition decision, including the amount of intractable information, observability and information acquisition costs. ∗Graduate School of Business, Stanford University, Stanford, CA 94305-5105. e-mails: [email protected] and tunca [email protected]. We thank the Editor, Maureen O’Hara, and two anonymous referees, as well as Anat Admati, Bill Beaver, Douglas Foster, Craig Holden, Lasse Pedersen, Bob Wilson, seminar participants at the 2000 NASDAQNotre Dame Conference on Market Microstructure, 2001 annual AFA meeting at New Orleans, Boston University, Ohio State University, Penn State University and Stanford University. The usual disclaimer applies. Introduction Providing liquidity to investors who want to exchange financial assets for cash or vice versa without possessing superior information is a key role of financial markets [Amihud and Mendelson (1986), Grossman and Miller (1988)]. The choice problem faced by such liquidity traders and its effects on market performance are often overlooked in dynamic models of asymmetric information in financial markets. In this paper, liquidity traders endogenously determine the quantities they trade, taking information asymmetry and asset riskiness into account. We study the effect of endogenous liquidity trading on market performance and on the informed trader’s incentives to acquire information and to release it to the market. We propose a tractable model of endogenous liquidity trading and derive its equilibrium in closed form. The results enable us to make qualitative and quantitative predictions on the effects of different types of information on liquidity, efficiency and welfare. We also formulate testable implications regarding the determinants of the spread, the nature of the price-adjustment process and the effects of changes in the investor base and in accounting rules on market liquidity, efficiency, information acquisition and research. Our starting point is a multiperiod discrete time model based on Kyle (1985), with an informed trader who obtains a noisy signal about the value of a risky asset. Information about the security’s value also diffuses to the market as an independent process. We model liquidity (uninformed) traders as risk-averse agents with idiosyncratic liquidity preferences. Each liquidity trader chooses her order size, and competitive market-makers price the security at its expected value given all publicly-available information. In equilibrium, the risk-neutral informed trader maximizes his expected profit, and risk-averse liquidity traders maximize their expected utilities. We find that endogenous liquidity trading leads to qualitatively different equilibrium characteristics compared to dynamic models with pure noise trading. Our analysis shows that models with exogenous liquidity trading are likely to overestimate the speed of information dissemination. We also find that both liquidity and efficiency are increasing functions of agents’ discount rates, and decreasing functions of the liquidity traders’ risk aversion. We distinguish between intractable, tractable and public information. Intractable information cannot be acquired: its release is entirely exogenous, and it is unavailable to the informed trader, to the marketmakers and to the liquidity traders. In contrast, tractable information can be acquired by the informed trader, creating informational asymmetry. We find that intractable information increases the adverse selection spread although it does not involve informational asymmetry. Further, intractable information increases the sensitivity of the spread to asymmetric information and to the liquidity trading interest. Making information asymmetry endogenous reduces market efficiency. When the informed trader chooses how much information to acquire, he takes into account the effect of his choice on liquidity traders’ behavior, which creates an incentive to limit the information asymmetry. We find that even at zero information acquisition cost, he may still choose not to acquire all the tractable information he could. This is in contrast to traditional noise trading models, where he would acquire all available information when it