Credit Default Swap Data, Contagion and Financial Resilience
Lead Research Organisation:
University of Oxford
Department Name: Oxford Man Institute
Abstract
The resilience of the UK financial system is of vital importance to UK society, and more specifically, to the corporations and fund managers that invest on behalf of our pension funds. It is apparent that our understanding of these systems leaves many questions unanswered and that there are fundamental issues that need to be addressed. If the research base is to inform policy decisions in any sort of adequate way there needs to be a strong development of insights from many disciplinary directions, which then need to be drawn together to be effective. This application is a part of the Oxford-Man ambition to develop tools and high quality insights that can collectively assist the wider project of enhancing the resilience of the UK financial system. This application brings together two interrelated projects "Containing Financial Crises" and "CDS Markets and Financial Contagion", and will allow us to put together the talents of some outstanding researchers at different points in their careers. This application is very timely because, at a wider level, the Oxford-Man Institute is engaging with the Bank of England and other academic partners, in a collaborative study to develop understanding of the effects of macro-prudential capital tools on financial stability. As such, it will provide an excellent environment to nurture this research and ensure strong impact.
Morrison and Wilson's project will look at understanding interbank connectedness and its relation to financial stability and financial contagion.
Schweikhard and Tsesmelidakis propose to develop their novel structural model that is capable of estimating the standalone credit quality of a firm, and use it to look at designing a market-based systemic risk tax, bank capital regulation, what policy actions or announcements are most useful in arresting the financial crises, and the side effects associated with the "too big to fail" policy.
Morrison and Wilson's project will look at understanding interbank connectedness and its relation to financial stability and financial contagion.
Schweikhard and Tsesmelidakis propose to develop their novel structural model that is capable of estimating the standalone credit quality of a firm, and use it to look at designing a market-based systemic risk tax, bank capital regulation, what policy actions or announcements are most useful in arresting the financial crises, and the side effects associated with the "too big to fail" policy.
Planned Impact
The resilience of the UK financial system is of vital importance to the UK society and, more parochially, to the corporations and fund managers such as Man that invest on behalf of our pension funds. This application is a part of the Oxford-Man ambition to develop tools and high quality insights that can contribute to enhancing the resilience of the UK financial system. At a wider level OMI is engaging with the Bank of England, the KTN, and other academic partners in a collaborative study to develop novel tools for understanding the effects of macro-prudential capital tools on financial stability. As a result this research is well positioned to ensure strong impact for its outputs.
Publications
Hilscher J
(2015)
Are Credit Default Swaps a Sideshow? Evidence That Information Flows from Equity to CDS Markets
in Journal of Financial and Quantitative Analysis
Morrison A
(2017)
Identifying Contagion in a Banking Network
Schweikhard F
(2013)
The Internalization of Systemic Risk: An Analysis of Bank Levy Schemes
Schweikhard F
(2014)
Beyond Capital Regulation: An Underestimated Risk Source
Tsesmelidakis Z
(2012)
The Value of Implicit Guarantees
in SSRN Electronic Journal
| Description | This project has allowed access and analysis of the importance to the wider economy of the network of CDS issuance between bank. In our original application we noted that it would likely be extremely difficult to obtain access to the data, and so it proved. None the less we have succeeded and after many tribulations and negotiations, for the past six months, we have been co-operating with two researchers at the Bank of England, Filip Zikes and Michalis Vasos, who have access to part of the CDS issuance data through their employment at the Bank to analyse this data. The arrangement allows us to access (indirectly) an excellent quality dataset. The custodian of the CDS issuance data, the Depositary Trust and Clearing Corporation (DTCC), has made available those parts of the data to national regulators of the banking system which fall under their jurisdiction. For our purposes, this means that through the Bank, we can work with data on all CDS contracts, regardless of counterparty, written on UK-incorporated reference entities (UK CDS), and all non-UK CDS with at least one UK-regulated counterparty. We emphasize that we ourselves do not have direct access to this data, only Filip and Michalis. We can propose hypotheses and regression specifications, see and discuss results (at the Bank) and write papers, but we do not ourselves run the regressions. This kind of collaboration is both relatively novel, potentially very productive, and frustrating because both Filip and Michalis have full-time jobs at the Bank as well as being researchers. They do not have as much time as we would like to pursue this project, and so progress has been slower than designed. Nevertheless, we have designed the regression specification which we believe can identify (if present) an effect of one bank loss on another bank purely by virtue of their connection through a network. Our Bank colleagues have the data they need to conduct the baseline regression: (1) bank equity returns; (2) bank CDS exposures; (3) reference entity CDS spread changes; (4) bank distances from each other in the network of CDS issuance. The baseline specification looks for an impact of a neighbouring bank losses on a given bank's equity return while controlling for the given bank's direct exposures. In addition to the baseline regression, we also want to look at indirect exposures revealed through the announcement of fines and lawsuits. In our sample period, many banks have had to pay enormous sums to settle civil and criminal litigation and regulatory actions. In the UK, many of these actions were conducted by the Financial Conduct Authority (FCA), and have the virtue (from our point of view) of the existence of the action being announced to the market at the same time as the fine. (See Armour, Mayer and Polo, (2012): Regulatory Sanctions and Reputational Damage in financial Markets, working paper, Oxford University.) We have the complete list of these fines and the dates they were announced, and can use these to conduct an event study looking at the impact of these announcements on bank equity returns, both directly and indirectly through their connection to a network of counterparties. For other data on bank fines and settlements, we have received an offer of assistance from Roger McCormick of the Cost of Conduct project. He and his research centre have compiled a comprehensive list of fines and lawsuits for the 12 largest global banks (all of which are central to the CDS network), including relevant announcement dates. So at this stage we do not have an outcome that decides the impact, It is of course still possible that we will not find evidence of contagion using this method. But we have an excellent outcome in terms of positioning of work in progress. We have two further ideas which we expect will be tested on the CDS data further adding value to the achievement so far. The first is to look for a reduction in the dispersion of prices following mandatory disclosure of CDS trade prices in early 2013 (this was mandated by the SEC for the benefit of all market participants). The second is to relate changes in the shape of the network to changes in prices to test the theories of Babus and Kondor (Trading and Information Diffusion in OTC Markets, R&R, Econometrica) as to how the network shape enables dealers to exploit their local market power. We fully intend to address both these research projects with our Bank colleagues next year, but given their time constraints, we are currently focussing on the intial, contagion-related, research. |
| Exploitation Route | I am very pleased to confirm that ultimately, the process initiated by this grant made possible effective access to the data, which in turn facilitated a particularly interesting analysis performed collaboratively by university researchers and researchers from the Bank of England and the Federal Reserve Board. Oxford academic staff working on the project have now completed a substantial impact report which we would like to, and eventually will, put in the public domain. Due to the sensitivity of the data, the BOE and the FSB are currently performing an audit of the report before it becomes public (this is in agreement with the signed NDA terms and conditions). The BoE and the FRB have, however, approved the abstract of the core paper and achievement of the project to be made public which indicates the nature of the impact: "This paper studies the impact of trading profits and losses on bank counterparty borrowing costs using data from a derivatives trade depositary. We use the network of credit default swap (CDS) transactions between banks to identify bank CDS returns attributable to counterparty losses. We find that counterparty losses significantly affect bank CDS returns by increasing their spreads, while non-counterparty losses do not. We also find that the effect on bank CDS returns through this counterparty loss channel is large relative to the direct effect on a bank's CDS returns from its own trading losses." "The views expressed in this paper are those of the authors and not necessarily those of the Bank of England or any of its committees, or the Federal Reserve Board or any other person associated with the Federal Reserve System." |
| Sectors | Financial Services, and Management Consultancy,Government, Democracy and Justice |
| Description | This project has allowed access and analysis of the importance to the wider economy of the network of CDS issuance between bank. In our original application we noted that it would likely be extremely difficult to obtain access to the data, and so it proved. None the less we have succeeded and after many tribulations and negotiations, for the past six months, we have been co-operating with two researchers at the Bank of England, Filip Zikes and Michalis Vasos, who have access to part of the CDS issuance data through their employment at the Bank to analyse this data. The arrangement allows us to access (indirectly) an excellent quality dataset. The custodian of the CDS issuance data, the Depositary Trust and Clearing Corporation (DTCC), has made available those parts of the data to national regulators of the banking system which fall under their jurisdiction. For our purposes, this means that through the Bank, we can work with data on all CDS contracts, regardless of counterparty, written on UK-incorporated reference entities (UK CDS), and all non-UK CDS with at least one UK-regulated counterparty. Through a delicate collaboration that preserves confidentiality, the researchers have propose hypotheses and regression specifications, discussed results and have been able to draw significant conclusions. This kind of collaboration is both relatively novel, potentially very productive, and frustrating. But the outcome seems to have been worth it. Our focus was to identify (if present) any effect of one bank loss on another bank purely by virtue of their connection through a network. Our Bank colleagues have the data needed to conduct the baseline regression: (1) bank equity returns; (2) bank CDS exposures; (3) reference entity CDS spread changes; (4) bank distances from each other in the network of CDS issuance. The baseline specification aimed to look for an impact of a neighbouring bank losses on a given bank's equity return while controlling for the given bank's direct exposures. Ultimately, the process initiated by this grant made possible a particularly interesting analysis performed collaboratively by university researchers and researchers from the Bank of England and the Federal Reserve Board. The Oxford academic staff working on the project have now completed a substantial impact report which we would like to, and eventually will, put in the public domain. However, due to the sensitivity of the data, the BOE and the FSB are currently performing an audit of the report before it becomes public (this is in agreement with the signed NDA terms and conditions). The BoE and the FRB have, however, approved the abstract of the core paper and achievement of the project to be made public, and these indicate the nature of the impact: "This paper studies the impact of trading profits and losses on bank counterparty borrowing costs using data from a derivatives trade depositary. We use the network of credit default swap (CDS) transactions between banks to identify bank CDS returns attributable to counterparty losses. We find that counterparty losses significantly affect bank CDS returns by increasing their spreads, while non-counterparty losses do not. We also find that the effect on bank CDS returns through this counterparty loss channel is large relative to the direct effect on a bank's CDS returns from its own trading losses." "The views expressed in this paper are those of the authors and not necessarily those of the Bank of England or any of its committees, or the Federal Reserve Board or any other person associated with the Federal Reserve System." In addition, the award supported two early career scientists who have written a number of very interesting papers brining a a broader perspective to the contagion/too big to fail / cds / questions and their papers from this period are also uploaded. |
| First Year Of Impact | 2015 |
| Sector | Financial Services, and Management Consultancy,Government, Democracy and Justice |
| Impact Types | Societal,Economic,Policy & public services |
| Description | Oxford University and the Bank of England |
| Organisation | Bank of England |
| Country | United Kingdom |
| Sector | Private |
| PI Contribution | This project relied on building a relationship with the Bank of England to work jointly on very sensitive and mutually interesting data. The project was a success and a significant collaboration emerged. The most recent paper appeared as a BOE report in January 2017. |
| Collaborator Contribution | Research, analysis, and data. |
| Impact | E.g. BOE Staff Working Paper No. 642 |
| Start Year | 2012 |