Strategic Trading in Multiple Assets and the Effects on Market Volatility

by Chenghuan Sean Chu, Andreas Lehnert, and Wayne Passmore
Federal Reserve Board

Abstract

We study government policies designed to increase liquidity by extending government guarantees to fundamentally illiquid assets. We characterize the effects of such policies on equilibrium price dynamics, trading strategies, and welfare. We build on the strategic trading framework of Brunnermeier and Pedersen (2005) and Carlin, Lobo, and Viswanathan (2007) by adding multiple assets and by modeling all agents’ utility functions. The assets in our model differ in their fundamental liquidity, i.e., the price reaction of the nonstrategic (or “retail”) traders to amounts sold by the strategic traders. Nonstrategic traders are willing to accept greater amounts of the more liquid asset with less short-term price reaction. These additions allow us to consider the welfare implications of, for example, shifting some assets from the illiquid category to the liquid category, a proxy for government guarantees on a risky asset. As in other models of this type, the strategic players “predate” on each other when one becomes distressed and is forced to liquidate its holdings. As others have shown, the more liquid the asset, the cheaper it is to predate on a distressed firm. Our model features an additional channel between liquidity and predation: because of the cross-elasiticies of demand across assets, firms can create liquidity in one asset by shorting a complementary asset. We find that when firms begin with larger endowments in highly liquid assets, forced liquidation of those assets tends to result in higher price volatility than would otherwise be the case. For plausible parameter ranges, such a policy also results in lower welfare for the nonstrategic traders. This finding suggests that market interventions designed to alleviate illiquidity in particular asset markets may instead unintentionally exacerbate price volatility.

JEL Codes: G0, G01, G12.

 
Full article (PDF, 30 pages 541 kb)

Discussion by Bruce Carlin