RDP 2016-07: The Efficiency of Central Clearing: A Segmented Markets Approach 1. Introduction

Following the global financial crisis, the G20 agreed on a number of reforms to financial markets. Member nations agreed to improve market transparency, reduce interconnectedness between market participants, and improve the management and collateralisation of bilateral credit risk. To this end, a key area of the reform agenda has been to mandate the use of central counterparties (CCPs) for certain over-the-counter (OTC) derivatives.

In line with these agreed reforms, Australian financial regulators have implemented mandatory central clearing requirements for certain interest rate derivatives (APRA, ASIC and RBA 2015; ASIC 2015; Australian Treasury 2015). A number of other jurisdictions – including the European Union, Japan and the United States – have implemented similar requirements.^[1]

This paper studies how mandatory central clearing affects market participants' incentives in OTC derivatives markets. The questions that we address are: How does a regulatory requirement for central clearing affect trade in derivatives? Under what conditions does a CCP improve welfare when managing counterparty credit risk? What does the optimal structure for mandatory central clearing look like?

To address these questions we use a micro-founded general equilibrium model with segmented and incomplete markets. The main advantage of this approach is that traders, allocations and prices endogenously respond to different central clearing arrangements; these responses can be material for the efficiency of any reform proposed. It also facilitates quantitative statements about the welfare implications of alternative market reforms. The existing literature has typically not considered endogenous allocations and prices or permitted quantitative welfare analysis. The disadvantage of our approach is that we cannot consider all market frictions relevant to the question of optimal central clearing, such as the role of size and networks in contractual relationships (Allen and Gale 2000; Jackson and Manning 2007; Duffie and Zhu 2011; Heath, Kelly and Manning 2013), or market opacity (Acharya and Bisin 2014).

The model is motivated by the market segmentation model of Edmond and Weill (2012). Traders face idiosyncratic risks that they can hedge by buying or selling OTC derivatives from other traders in their market. These contracts reference an underlying financial price and are zero sum.

There are two key market frictions. The first is that traders can only exchange contracts in the market in which they trade. Second, traders are exposed to the risk of counterparty default, which cannot be privately insured against. This environment creates a role for CCPs to provide default insurance. The CCP in our model does so by guaranteeing payment on all contracts that it clears. It covers its obligations by collecting two types of resources: per-contract initial margin requirements – resources that the CCP collects from traders in advance and uses if the trader defaults; and a default fund contribution that can be used to cover any losses remaining once defaulting traders' initial margin has been exhausted. We do not consider variation margin – cash exchanged daily in response to mark-to-market changes in participants' OTC derivatives portfolios – as the model is calibrated to match quarterly data (the coupon frequency of the majority of OTC derivatives).

Using this framework, we find that mandatory central clearing:

• can be welfare improving because it mutualises counterparty credit risk
• increases welfare if initial margin requirements are set optimally, but can be inefficient otherwise
• is more beneficial the higher the probability of default and the more effective are OTC derivative contracts as a hedge against idiosyncratic risk.

Our findings capture the trade-off between the default insurance that a CCP provides and the incentive for market participants to trade ‘too much’ when default losses are mutualised through the CCP's default fund. This is akin to a moral hazard problem in insurance where an agent undertakes too much of a risky activity when insurance is provided.

Default insurance is beneficial in our model because it reduces traders' consumption variance. Since, in the model, traders do not have access to private insurance against the risk of counterparty default, the core benefit of a CCP is that it provides insurance that would not otherwise exist.^[2] This default insurance function is a material benefit of central clearing that, to date, the literature has not explored in a general equilibrium framework.

The main cost of central clearing that we consider is that it alters market participants' incentives. With default insurance, traders have the incentive to write more financial contracts than they otherwise would. Because the cost of insurance provided by the CCP is mutualised among all CCP participants, it is not internalised at the individual level. This generates a ‘mutualisation externality’, which means that privately optimal trade does not coincide with socially optimal trade.

In terms of related literature, the role of CCPs as insurance providers and how CCPs affect trader decision-making have not been widely studied. However, our paper connects to an emerging research field on optimal CCP design including: Acharya and Bisin (2014) who study an externality in financial trade due to opacity in counterparty positions; Duffie and Zhu (2011) who study the role of a network and contract netting; Heath, Kelly and Manning (2015) who study the transmission of financial stress through a network; Heath et al (forthcoming) who use portfolio-level data from 41 large OTC derivatives market participants to inform analysis of the stability of the financial network and how losses might be transmitted in times of stress; and Koeppl, Monnet and Temzelides (2012) who study cross-subsidisation between centralised and non-centralised markets. Our paper is also related to Pirrong (2009), who discusses the importance of the moral hazard problems that central clearing can engender, and Haene and Sturm (2009) who use a stylised model to analyse the extent to which CCPs should rely on initial margin versus their default fund. The modelling approach we use draws from Edmond and Weill (2012).

The paper is structured as follows: Section 2 provides background on OTC derivatives markets and CCPs, and a brief overview of the existing literature. Section 3 outlines our model. Section 4 presents our results, including: the welfare benefits of OTC derivatives; the effect of central clearing on trade and welfare, and how appropriately set margin requirements are important in ensuring the welfare benefits of central clearing are realised. We then turn to the question of the optimal level of mandatory clearing and discuss some policy implications of our results. Section 5 concludes.

Footnotes

As of November 2015, 12 of 24 Financial Stability Board member jurisdictions have a legislative framework in force for mandatory central clearing requirements (FSB 2015). [1]

Although private market default insurance exists in the real world – in the form of credit default swaps (CDS) – single-name CDS exist only for fairly large institutions. Moreover, the protection buyer is still exposed to the credit risk of the counterparty on the CDS. Bilateral initial margin can also provide a form of default insurance, though it is only partial. We do not consider bilateral margining in this paper. [2]