Liquidity and Trading in Bond and Derivatives Markets I

Paper Session

Friday, Jan. 6, 2017 8:00 AM – 10:00 AM

Sheraton Grand Chicago, Chicago Ballroom VI
Hosted By: American Finance Association
  • Chair: Francis Longstaff, University of California-Los Angeles

Taking Orders and Taking Notes: Dealer Information Sharing in Financial Markets

Nina Boyarchenko
,
Federal Reserve Bank of New York
David Lucca
,
Federal Reserve Bank of New York
Laura Veldkamp
,
New York University

Abstract

Many recent scandals and allegations revolve around the question: If a dealer brokers a trade for a client, should the dealer be allowed to tell other clients or other dealers about that trade? We explore how such information sharing affects dealers, clients and the revenues of the U.S. treasury. Since we cannot observe various information-sharing regimes, we need to build a model of financial intermediaries who share information about their order flow and construct counter-factuals in the model. Calibrating the model to U.S. Treasury auction data, we estimate that auctions with full information sharing would generate $4.8 billion more revenue per year than a no-sharing (“Chinese Wall”) regime would. While one might think that such large revenue gains come at the expense of small investors, that depends on the nature of information sharing. Surpris- ingly, investors benefit when dealers share information with each other, not when they share more with clients. In a mixed auction, where investors can opt out of interme- diated trade and bid directly, information sharing creates a new financial accelerator: Only investors with bad news employ intermediaries, who then share that information with others. Sharing amplifies the effect of negative news and raises the probability of “auction failure”. Finally, when dealers face minimum bidding requirements, some amount of information sharing is needed to induce dealers to function as dealers. Tests of two model predictions support its key features.

Institutional Herding and Its Price Impact: Evidence From the Corporate Bond Market

Fang Cai
,
Federal Reserve Board
Song Han
,
Federal Reserve Board
Dan Li
,
Federal Reserve Board
Yi Li
,
Federal Reserve Board

Abstract

Among growing concerns about potential financial stability risks posed by the asset management industry, herding has been considered as an important risk amplification channel. In this paper, we examine the extent to which institutional investors herd in their trading of U.S. corporate bonds and quantify the price impact of such herding behavior. We find that, relative to what is documented for the equity market, the level of institutional herding is much higher in the corporate bond market, particularly among speculative-grade bonds. In addition, mutual funds have become increasingly likely to herd when they sell, a trend not observed among insurance companies and pension funds. We also show that bond investors herd not only within a quarter, but also over adjacent quarters. Such persistence in trading is largely driven by funds imitating the trading behavior of other funds in the previous quarter. Finally, we find that there is an asymmetry in the price impact of herding. While buy herding is associated with a permanent price impact that is consistent with price discovery, sell herding results in transitory yet significant price distortions. The price destabilizing effect of sell herding is particularly strong for high-yield bonds, small bonds, and illiquid bonds and during the recent global financial crisis.

The Value of Trading Relations in Turbulent Times

Marco Di Maggio
,
Harvard University and NBER
Amir Kermani
,
University of California-Berkeley
Zhaogang Song
,
Johns Hopkins University

Abstract

This paper investigates how dealers’ trading relations shape their trading behavior in the corporate bond market. Dealers charge lower spreads to dealers with whom they have the strongest ties and more so during periods of market turmoil. Systemically important dealers exploit their connections at the expense of peripheral dealers as well as clients, charging higher markups than to other core dealers. Also, intermediation chains lengthened by 20% following the collapse of a flagship dealer in 2008 and even more for institutions strongly connected to this dealer. Finally, dealers drastically reduced their inventory during the crisis.
Discussant(s)
Kjell Nyborg
,
University of Zurich and Swiss Finance Institute
Kelsey Wei
,
University of Texas-Dallas
Ana Babus
,
Federal Reserve Bank of Chicago
JEL Classifications
  • G1 - General Financial Markets