## _wp14203 - References ## Source details **Canonical URL:** [_wp14203 - References](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14203.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14203.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14203.pdf.json) --- ### Introduction: bank–nonbank nexus and systemic risk - Banking and non-banking activities carried on by banks are often closely intertwined and legally difficult to disentangle, especially for systemically important banks (SIBs). - A significant portion of SIB exposures may be non-deposit related—in some instances over 90 percent (example firms cited). - Nonbanks (hedge funds, asset managers, insurers, pension funds and central counterparties (CCPs)) do not overlap with SIBs in Figure 1, implying limited ex ante economic justification for taxpayer support of nonbanks if support is justified solely as protection of deposits. - CCPs have been labeled systemically important and treated as financial market infrastructures (FMIs) or financial market utilities (FMUs) in some jurisdictions. - The surge in products subject to mandatory clearing under the G20 agenda increases the likelihood that a CCP failure would have severe systemic consequences. - The recovery and resolution procedure for CCPs must be relatively swift, simple, and “arithmetically supportable” to preserve critical services and safeguard financial stability. ### CCPs: structural features, risks, and ownership questions - CCPs pre-fund liquidation of clearing member positions via initial margin (IM) and periodic variation margin (VM); residual extreme losses are covered through default fund and equity provision. - CCPs typically hold less “conventional” capital than other financial institutions while mandatory clearing exposes CCPs to less liquid instruments and less predictable price volatility, increasing residual failure risk. - Historical CCP failures (Caisse de Liquidation 1974, Kuala Lumpur Commodities Clearing House 1983, Hong Kong Futures Exchange 1987) offer limited guidance because prior CCP balance sheets lacked many derivatives. - Margin calls can stress market and credit liquidity; imposing losses on all participants can impact asset managers, pension funds, and insurers—some argue this impact justifies public rescues, but the paper considers that argument weak. - Ownership and governance considerations: - Greater alignment of mutualized risk and reward may require participant (user) ownership or different governance structures. - Some CCPs are user-owned or partly user-owned; many are owned by exchanges or independent shareholders and run for profit. - Treating CCPs as public-interest utilities is problematic unless they cover the full spectrum of “economic rents”; utility status may inhibit investment unless a statutory or de facto monopoly exists. - Regulatory intent is to avoid taxpayer bailouts by ensuring participants bear cost ex ante (via waterfall) or ex post (cash calls, haircuts). ### Loss-allocation mechanisms and legal constraints - Default arises when a clearing member (CM) fails to meet a variation margin call and the CCP assumes unmatched risk until hedging/auctioning restores a matched book. - Primary CCP protection is IM; prefunded default waterfall (default/guarantee fund) mutualizes residual loss among surviving CMs. - Waterfalls are not transparent; example: Hanmag Securities (Korea) default in December 2013 where KRX capital came after non-defaulting members’ default fund contributions. - Consistency with CPSS–IOSCO recommends end-of-the-waterfall losses be written into CCP rules ex ante; mandatory end-of-the-waterfall loss-allocation rules have support in the U.K. - Legal limits and constraints: - Haircutting VM needs to respect U.S. and EU legal frameworks. - EMIR prohibits using IM posted by non-defaulting clients/members to cover losses from another CM’s default; IM is bankruptcy-remote/segregated under U.S. law. - EMIR establishes a principle of limited CM liability that appears to preclude uncapped cash calls. - Under proposed Basel III, bank capital requirements are complicated by uncapped liabilities. - Targeted VM haircuts (haircutting variation margin gains) reduce recourse to government by taking profits from both CMs and indirect participants. ### VMGH: scale, mechanics, and policy role - Key statistics and estimates: - About US$3.8 trillion of collateral was used as margins (ISDA 2013). - Adjusting for double counting, about US$1.9 trillion of collateral is in-the-money. - Adjusting for a re-use rate of between 2 and 3, collateral in in-the-money positions may be in the order of US$700–900 billion (Singh 2013). - Under-collateralization in the OTC derivatives market since Lehman remains in the US$3–5 trillion range (BIS, 2014). - About 70 percent of OTC derivatives market is interest rate swaps. - Example scenario: two large players asymmetrically “in the money” and “out of the money” by US$50 billion; hedge fund AUM example US$10–15 billion; shortfall x satisfies 0 ≤ x ≤ US$50 billion. - Mechanics and implications: - VMGH (variation margin gains haircut) funds CCP losses by taking profits from CMs and indirect participants with in-the-money positions. - CCPs generally face CMs and cannot directly haircut end-users; options include: - Agreed pass-through of haircut from CM to end-user in line with end-user VM gain positions. - CM assumes all haircuts (possible due to netting and collateral reuse). - CMs charge end-users an ongoing fee for assuming VMGH exposure. - VMGH separates liquidity needs from solvency problems by providing a concrete measure of unfunded losses and a temporary recoupment mechanism from non-defaulting in-the-money VM payments. - VMGH is embedded in covenants of several CCPs (examples listed in the box and later section). - VMGH incentivizes participants to increase interest in CCP governance and risk management, reducing moral hazard and the likelihood of recourse to central banks or treasuries. ### Central bank liquidity and jurisdictional variation (Box summary) - CCPs must maintain sufficient liquid resources and may include liquidity-generating measures (calls on members, moratoriums, funding lines). - At least one historical episode (CME 1987) involved central bank backstop routed through member banks. - Jurisdictional differences in official backstops: - Under Title VIII of Dodd–Frank, a designated systemic FMU CCP can access Federal Reserve liquidity in “unusual or exigent circumstances,” subject to Treasury approval. - Title II orderly liquidation could allow FDIC to finance costs by borrowing from the Treasury up to a specified maximum. - Some central banks explicitly authorize liquidity to nonbank financial institutions (e.g., Sweden, Switzerland); Eurosystem practice varies and ECB requires banking licenses for credit operations eligibility. - VMGH offers an approach to reduce ambiguous labeling of liquidity versus solvency and diminishes incentives to seek public backstops. ### Policy recommendations and conclusions - CCPs have become “too important to fail” by regulatory fiat; the paper argues the case for public rescue is weak and proposes alternatives to reduce that risk. - Recommended measures: - Greater loss-sharing and robust waterfall structures in CCP rulebooks; mandatory end-of-the-waterfall allocation provisions are recommended. - VMGH as an additional recovery tool that: - Covers unfunded losses in CM defaults, - Encourages better governance (CMs and end-users will monitor CCP risk management), - Reduces need for central bank liquidity and government support, - Provides time to resolve legal complexities in CCP recovery/resolution. - To avoid uncapped VMGH moral hazard and regulatory arbitrage: - All users (or via CMs) should be willing to increase IM and default fund contributions at the waterfall level. - Users should be part of governance and risk-management committees. - Jurisdictions may diverge; some may offer low IM and no VMGH to attract market share with implied state guarantees. - If VMGH and robust waterfalls are insufficient, statutory powers for resolution may be necessary (Wendt, 2014), but preferable to secure recovery outcomes ex ante via CCP rule requirements. *Source: Excerpt from _wp14203 - References (IMF PDF chapter/section).* ### References ............................................................................................................. ### _wp14203 - References ### Figures - 1. The Bank–Nonbank Nexus ....................................................................................................3 - 2. A Hypothetical "Waterfall" Situtation ...................................................................................9 - 3. LOLR Funds vs. VMGH......................................................................................................10 ### Box - Central Bank Liquidity (when there is no uncapped VMGH) .................................................12 *Source: _wp14203 - References ......................................................................................................14* ### References ............................................................................................................. ### _wp14203 - References ### Introduction: bank–nonbank nexus and systemic risk - Banking and non-banking activities carried on by banks are often closely intertwined and legally difficult to disentangle, especially for systemically important banks (SIBs). - A significant portion of SIB exposures may be non-deposit related—in some instances over 90 percent (example firms cited). - Nonbanks (hedge funds, asset managers, insurers, pension funds and central counterparties (CCPs)) do not overlap with SIBs in Figure 1, implying limited ex ante economic justification for taxpayer support of nonbanks if support is justified solely as protection of deposits. - CCPs have been labeled systemically important and treated as financial market infrastructures (FMIs) or financial market utilities (FMUs) in some jurisdictions. - The surge in products subject to mandatory clearing under the G20 agenda increases the likelihood that a CCP failure would have severe systemic consequences. - The recovery and resolution procedure for CCPs must be relatively swift, simple, and “arithmetically supportable” to preserve critical services and safeguard financial stability. ### CCPs: structural features, risks, and ownership questions - CCPs are designed to pre-fund liquidation of clearing member positions via initial margin (IM) and periodic variation margin (VM); residual extreme losses are covered through default fund and equity provision. - CCPs typically hold less “conventional” capital than other financial institutions, while mandatory clearing exposes CCPs to less liquid instruments and less predictable price volatility, increasing residual failure risk. - Historical CCP failures (Caisse de Liquidation 1974, Kuala Lumpur Commodities Clearing House 1983, Hong Kong Futures Exchange 1987) offer limited guidance because prior CCP balance sheets lacked many derivatives. - Margin calls can stress market and credit liquidity; imposing losses on all participants can impact asset managers, pension funds, and insurers—some argue this impact justifies public rescues, but the paper considers that argument weak. - Ownership and governance questions: - Greater alignment of mutualized risk and reward may require participant (user) ownership or different governance structures. - Some CCPs are user-owned or partly user-owned; many are owned by exchanges or independent shareholders and run for profit. - Treating CCPs as public-interest utilities is problematic unless they cover the full spectrum of “economic rents”; utility status may inhibit investment unless a statutory or de facto monopoly exists. - Regulatory intent is to avoid taxpayer bailouts by ensuring participants bear cost ex ante (via waterfall) or ex post (cash calls, haircuts). ### Loss-allocation mechanisms and legal constraints - Default arises when a clearing member (CM) fails to meet a variation margin call and the CCP assumes unmatched risk until hedging/auctioning restores a matched book. - Primary CCP protection is IM; prefunded default waterfall (default/guarantee fund) mutualizes residual loss among surviving CMs. - Waterfalls are not transparent; example: Hanmag Securities (Korea) default in December 2013 where KRX capital came after non-defaulting members’ default fund contributions. - Consistency with CPSS–IOSCO recommends end-of-the-waterfall losses be written into CCP rules ex ante; mandatory end-of-the-waterfall loss-allocation rules have support in the U.K. - Legal limits: - Haircutting VM needs to respect U.S. and EU legal frameworks. - EMIR prohibits using IM posted by non-defaulting clients/members to cover losses from another CM’s default; IM is bankruptcy-remote/segregated under U.S. law. - EMIR establishes a principle of limited CM liability that appears to preclude uncapped cash calls. - Under proposed Basel III, bank capital requirements are complicated by uncapped liabilities. - Targeted VM haircuts (haircutting variation margin gains) reduce recourse to government by taking profits from both CMs and indirect participants. ### VMGH: scale, mechanics, and policy role Key statistics and estimates preserved from the source: - About US$3.8 trillion of collateral was used as margins (ISDA 2013). - Adjusting for double counting, about US$1.9 trillion of collateral is in-the-money. - Adjusting for a re-use rate of between 2 and 3, collateral in in-the-money positions may be in the order of US$700–900 billion (Singh 2013). - Under-collateralization in the OTC derivatives market since Lehman remains in the US$3–5 trillion range (BIS, 2014). - About 70 percent of OTC derivatives market is interest rate swaps. - Example scenario: two large players asymmetrically “in the money” and “out of the money” by US$50 billion; hedge fund AUM example $10–15 billion; shortfall x satisfies 0 ≤ x ≤ US$50 billion. Mechanics and implications: - VMGH (variation margin gains haircut) funds CCP losses by taking profits from CMs and indirect participants with in-the-money positions. - CCPs generally face CMs and cannot directly haircut end-users; options: - Agreed pass-through of haircut from CM to end-user in line with end-user VM gain positions. - CM assumes all haircuts (possible due to netting and collateral reuse). - CMs charge end-users an ongoing fee for assuming VMGH exposure. - VMGH separates liquidity needs from solvency problems by providing a concrete measure of unfunded losses and a temporary recoupment mechanism from non-defaulting in-the-money VM payments. - VMGH is embedded in covenants of several CCPs (examples listed in the box and later section). - VMGH incentivizes participants to increase interest in CCP governance and risk management, reducing moral hazard and the likelihood of recourse to central banks or treasuries. ### Central bank liquidity and jurisdictional variation (Box summary) - CCPs must maintain sufficient liquid resources and may include liquidity-generating measures (calls on members, moratoriums, funding lines). - At least one historical episode (CME 1987) involved central bank backstop routed through member banks. - Jurisdictional differences: - Under Title VIII of Dodd–Frank, a designated systemic FMU CCP can access Federal Reserve liquidity in “unusual or exigent circumstances,” subject to Treasury approval. - Title II orderly liquidation could allow FDIC to finance costs by borrowing from the Treasury up to a specified maximum. - Some central banks explicitly authorize liquidity to nonbank financial institutions (e.g., Sweden, Switzerland); Eurosystem practice varies and ECB requires banking licenses for credit operations eligibility. - VMGH offers an approach to reduce ambiguous labeling of liquidity versus solvency and diminishes incentives to seek public backstops. ### Policy recommendations and conclusions - CCPs have become “too important to fail” by regulatory fiat; the paper argues the case for public rescue is weak and proposes alternatives to reduce that risk. - Advocate greater loss-sharing and robust waterfall structures in CCP rulebooks; mandatory end-of-the-waterfall allocation provisions are recommended. - VMGH is proposed as an additional recovery tool that: - Covers unfunded losses in CM defaults, - Encourages better governance (CMs and end-users will monitor CCP risk management), - Reduces need for central bank liquidity and government support, - Provides time to resolve legal complexities in CCP recovery/resolution. - To avoid uncapped VMGH moral hazard and regulatory arbitrage: - All users (or via CMs) should be willing to increase IM and default fund contributions at the waterfall level. - Users should be part of governance and risk-management committees. - Jurisdictions may diverge; some may offer low IM and no VMGH to attract market share with implied state guarantees. - If VMGH and robust waterfalls are insufficient, statutory powers for resolution may be necessary (Wendt, 2014), but preferable to secure recovery outcomes ex ante via CCP rule requirements. *Italic: Source: Excerpt from _wp14203 - References (IMF PDF chapter/section).* ### REFERENCES ### REFERENCES ### Regulatory reports and standards - Bank for International Settlements, 2014, Semi-Annual OTC Derivative Statistics (various issues). - _____, 2014, “Recovery of Financial Market Infrastructures,” Committee on Payments and Market Infrastructures/ISOCO, October. - Bank of England, 2010, “Financial Stability Report,” December. - CPSS–IOSCO, 2013, “Principles for Financial Market Infrastructures,” July. - International Monetary Fund, 2010, “Global Financial Stability Report,” Chapter 3, April. - CPSS–IOSCO (see ISDA, IIF, The Clearing House letter), 2013 correspondence referenced: ISDA, IIF, The Clearing House, 2013, letter to CPSS-IOSCO, re: Recovery of Financial Market Infrastructures, October 11. ### Central counterparty (CCP) governance, recovery, and loss allocation - Elliot, David, 2013, “Central Counterparty Loss Allocation Rules,” Bank of England, Financial Stability Paper No. 20. www.bankofengland.co.uk/research/Documents/fspapers/fs_paper20.pdf - Gibson, Matt, 2013, “Recovery and Resolution of Central Counterparties,” Reserve Bank of Australia, December quarter bulletin. www.rba.gov.au/publications/bulletin/2013/dec/5.html - ISDA, 2013, CCP Loss Allocation at the End of the Waterfall. - Wendt, Froukelien, 2014, “Central Counterparties: Addressing Their Too Important to Fail Nature,” IMF Working Paper. ### Collateral, margin, and OTC derivatives market structure - Heller, Daniel and Nicholas Vause, 2013, “Collateral Requirements for Mandatory Central Clearing of Over-the-Counter Derivatives,” BIS Working Paper 373 http://www.bis.org/publ/work373.htm - _____, 2014, Margin Surveys (various issues). - Singh, 2010, “Collateral, Netting and Systemic Risk in the OTC Derivatives Market,” IMF Working Paper 10/99. www.imf.org/external/pubs/ft/wp/2010/wp1099.pdf - _____, 2012, Puts in the Shadows,” IMF Working Paper 12/229. www.imf.org/external/pubs/ft/wp/2012/wp12229.pdf - _____, 2013, “New Regulations and Collateral Requirements—Implications for the OTC Derivatives Market, October for Swift Institute,” SWIFT-Institute-Working-Paper-No-2012-004-New-Regulations-and-Collateral-OTC-Derivatives-Singh_v7-FINAL.pdf 1.13MB. ### Industry commentary, case studies, and market reports - ISDA, IIF, The Clearing House, 2013, letter to CPSS-IOSCO, re: Recovery of Financial Market Infrastructures, October 11. - Securities Finance Monitor, 2014, “The Korea Exchange: A Cautionary Tale on CCP Waterfalls and Non-Defaulting Members Taking the Loss,” March 18. - Summe, Kimberly, 2012, “An Examination of Lehman Brothers’ Derivatives Portfolio Post-Bankruptcy and Whether Dodd-Frank Would Have Made Any Difference,” Hoover Institution, April. http://media.hoover.org/documents/Kimberly-Summe-Dodd-Frank-20110421.pdf - Tucker, Paul, 2013, speech at Institute of International Finance, 2013 Annual Membership Meeting, Washington, DC, October 12, 2013. www.bis.org/review/r131015a.pdf *Source: _wp14203 - REFERENCES* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14203.pdf_