Abstract: Concern for the stability and resilience of financial markets has recently revived, in the wake of the sizeable number of "flash events'' that have occurred in recent years. A unifying characteristic of these episodes seems to be the jamming of the "rationing'' function of market illiquidity. In "normal'' conditions, traders perceive a lack of liquidity as a cost, while arbitrageurs and liquidity suppliers regard it as an opportunity. Thus, an illiquidity hike leads the former to limit their demand for immediacy, and the latter to increase their supply of liquidity, stabilizing the market. However, on occasions, a bout of illiquidity has a destabilizing impact, and fosters a disorderly "run for the exit.'' In these cases, traders attempt to place orders despite the liquidity shortage, and arbitrageurs flee the market, foregoing profitable opportunities. In such conditions, liquidity is fragile. What can account for such a dualistic feature of market illiquidity? We show that, consistent with empirical evidence, access to order flow information allows traders to supply liquidity via contrarian marketable orders. Lack of market transparency can make liquidity demand upward sloping, inducing strategic complementarity and multiple equilibria. Then an initial dearth of liquidity may degenerate into a liquidity rout (as in a “flash crash”) and traders faced with the largest cost of trading are those consuming more liquidity at equilibrium. An increase in order flow transparency and/or in the mass of dealers who are in the market at all times has a positive impact on total welfare.
Discussant: Michael Sockin, University of Texas-Austin
Abstract: During times of stress when demand for liquidity surges, dealers' willingness to provide liquidity is essential to the proper functioning of the U.S. corporate bond market. The existing literature on bond market liquidity emphasizes the roles played by funding costs and the regulatory reforms passed in the aftermath of the 2008 financial crisis. However, what determines dealers’ willingness to intermediate trades when the market becomes one-sided is still to be fully understood. We show that dealers' prime brokerage relations with certain hedge funds help improve their liquidity provision in a one-sided market. During the March 2020 liquidity crisis, hedge funds increased their corporate bond positions when the bond market faced excessive selling pressures. Dealers with stronger connections to hedge funds that are natural buyers of corporate bonds charged lower transaction costs on heavily sold bonds. Dealers' leverage and funding constraints do not explain our results, nor do connections with hedge funds that are natural buyers of other asset classes. We find that hedge funds that were larger and better able to absorb risk provided more liquidity during the crisis. Our findings reveal that dealers' willingness to provide liquidity in a one-sided market depends on their connections with natural buyers of corporate bonds.
Discussant: George Aragon, Arizona State University
Abstract: We develop a model of market making that incorporates the cost of dealer inventory and the level of adverse selection. With a high cost and low adverse selection, which we argue describes the current corporate bond market, dealers engage in both principal and agent trading. Trade transparency reduces volumes by shifting more trades into the (uncertain) agent protocol, and increases bid-offer of principal trades, particularly for bonds that are hard to “match” in agent trades. To test these predictions, we construct a novel database of euro corporate bond transactions. We exploit exogenous variation in transparency generated by Brexit, and show that transparency decreases transaction costs for small trades but increases transaction costs by 23% for larger and more difficult to match trades. Our results can be used to inform policy makers in light of recent proposals to shorten reporting delays for corporate bond transactions in Europe.
Discussant: Kumar Venkataraman, Southern Methodist University
Abstract: We propose a new way to quantify portfolio investment frictions. Representing any given friction as a constraint, it's emph{Constraint-Induced Factor} is the tracking error from replicating the tangency portfolio subject to the constraint. Constraint-Induced Factors carry a structural interpretation in a standard asset pricing framework, where their price of risk corresponds to the mass of affected capital and the marginal cost of the associated friction in risk-sharing terms. In empirical applications to global bond markets, we show that the magnitude of pricing effects associated with U.S. Treasury convenience and capital controls for Chinese government bonds is consistent with 20-25% of global bond market capital being constrained to invest in each of those markets. We also discuss applications to stock markets, including ESG investing and indexing.
