A Theory of Participation in OTC and Centralized Markets

Abstract

Should regulators encourage the migration of trade from over-the-counter (OTC) to centralized markets? To address this question, we study a model in which banks make costly decisions to participate in an OTC market, a centralized market, or both markets at the same time. Banks differ in their ability to take large positions, what we call their trading capacity. In equilibrium, intermediate-capacity banks find it optimal to participate in the centralized market. In contrast, low-and high-capacity banks find it optimal to participate in the OTC market, due to an endogenous complementarity. Namely, low capacity banks receive worse terms of trade than in the centralized market but better risk sharing, thanks to the intermediation services offered by high-capacity banks. High-capacity banks receive worse risk sharing than in the centralized market, but profit from the provision of intermediation services to low-capacity banks. While the social optimum has qualitatively similar participation patterns, it prescribes that more customers migrate to the centralized market, and that more dealers enter the OTC market.

OTC Discount

“…There is a strong e↵ort to improve OTC market transparency both in Europe, where the MiFID II regulation was recently rolled out with the intention of improving market conditions in and beyond European markets and in the U.S., where FINRA has started to collect data similar to that in the TRACE database also for sovereign bonds. As pointed out by Dugast, Üslü and Weill (2019), even if a centralized exchange market is socially optimal, agents' decisions for trading protocols are driven by their private incentives. While our empirical analysis cannot shed light on the socially optimal market structure, our results suggest that OTC, exchange, and broker-intermediated trading play complementary roles in serving the di↵erent needs of dealers in a way that a single venue might not be able to achieve.…”

Bargaining under liquidity constraints: Unified strategic foundations of the Nash and Kalai solutions

“…A second advantage of the Kalai solution is its tractability: it preserves the concavity of the surpluses with respect to z and it generates simple closed-form expressions. 7 Relative to Nash, however, the Kalai solution is not scale invariant and lacks solid strategic foundations.…”

“…Another view is that the rules of the extensive-form game that determine the allocations in pairwise meetings are not manipulable by the players and are a primitive of the economy (which is the standard assumption in the literature on decentralized markets). 7 One example that showcases the tractability provided by the proportional solution is Lester et al (2012) on costly information acquisition. Another example is Geromichalos et al (2016) on term structure in OTC markets.…”

“…In each round, the amount of goods up for negotiation is y =N , and the players bargain according to a Rubinstein (1982) alternating-o¤er game with exogenous risk of termination (or, equivalently, a stochastic horizon). We show that for all N this game admits a (essentially) unique subgame-perfect equilibrium (SPE) 1 Examples of these di¤erent applications include: Lehmann (2012) or Gu et al (2019) for a labor search model where wages are negotiated subject to the cash constraint of the …rm; Lagos and Zhang (2019) for a model of an OTC market where traders' purchases are constrained by their cash holdings; Dugast et al (2019) where OTC traders are constrained by a market trading capacity; Rocheteau et al (2018) for a corporate …nance model where loan contracts are negotiated bilaterally between a bank and an entrepreneur; and all the New Monetarist literature surveyed in Rocheteau and Nosal (2017) and Lagos et al (2017). 2 The Kalai solution should not be confused with the Kalai-Smorodinsky (1975) solution that is scale invariant but not strongly monotone.…”

Centralizing Over-the-Counter Markets?

“…Following the literature, we assume exclusive participation per market segment Dugast et al (2019). only recently relaxed this assumption.…”