Over-the-Counter (OTC) Derivative Primer 2: Counterparty Risk
Over-the-Counter (OTC) Derivative Primer: Counterparty Risk
By John Kiff
This post is the second in a series of over-the-counter (OTC) derivative primers. The first covered the product basics and this one will cover bilateral contract counterparty risk management. The next post will cover clearing and central counterparties (CCPs).
OTC derivatives expose counterparties to the default risk of others while those contracts have positive replacement values—the value or payment the nondefaulting party would receive if the contract were terminated early. The maximum loss from a defaulting counterparty is equal to the sum of the positive replacement values (“derivative receivables”). Plus, in the absence of legally enforceable offset rights (the contracts are not under a single master netting agreement) the performing counterparty would be on the hook to the non-performing counterparty for the sum of the negative replacement values (“derivatives payables”).
Valuations are generally derived from the value of arbitrage-free replicating portfolios of other financial instruments. Swaps are valued according to the market market prices of the implicit fixed and variable payment flows (e.g., LIBOR for most interest rate swaps). Options are valued on the market value of a dynamically hedged and funded portfolio of underlying assets. In the wake of the recent financial crisis, internal valuation models have become increasingly complex, factoring in counterparty risks, alternative discount rates and margin posting (see below).
However, most OTC derivatives are covered by bilateral master agreements that allow for close-out netting when one of the counterparties defaults. This is not the same as the payment netting that occurs throughout the life of a transaction in which all payment obligations in a single currency between counterparties are replaced with a single net amount on each payment date.
Close-out netting occurs at the end of a transaction when one party has defaulted. When default occurs, termination of the contract is typically triggered by the nondefaulting party, a single net amount due between the parties becomes payable, and the nondefaulting party is given access to its collateral if the defaulting party owes anything to the nondefaulting party. These permit the derivative payables to be used to offset derivative receivables within netting sets. (A netting set envelopes all trades with a given counterparty that can be legally netted together upon one of the counterparty’s default under the master agreement’s terms and conditions.)
Interestingly, the methodologies for calculating replacement values are typically not specified in master agreements. Instead counterparties use their own, often proprietary, valuation models so that valuation disputes are not uncommon. Murphy (2013) provides a worrisome example of the disputes between Goldman Sachs and the American Insurance Group (AIG) in 2007. (See the Goldman AIG Collateral Call Timeline on the Financial Crisis Inquiry Commission website.)
Counterparty risk can be further reduced by requiring collateral posting against outstanding exposures and residual risks. Residual risks include delays between when the new collateral requirements are calculated, called, and settled, the impact of minimum transfer amounts, and the potential for replacement value fluctuations from when default occurs and the contracts are closed out. In futures markets, the upfront amount is called “initial margin” and is viewed as a performance bond or guarantee that a counterparty will honor its contractual agreements.
Current exposures are covered by variation margin (VM) posting against day-to-day changes in net replacement costs. To cover potential future exposure, dealers usually require end users to post an “independent amount” or initial margin (IM). End-users include investment funds, hedge funds, and other nondealers. However, many sovereign and quasi-sovereign entities, and some corporate clients are often exempt from margin posting.
IM requirements for a portfolio of derivatives are typically calculated using a value-at-risk (VaR) based risk model (Murphy, 2014). Parameters are usually estimated with historical data over one- to five-year horizons, based on 99% confidence levels over 10-day liquidation periods. Add-on factors include collateral liquidity measures and stress test results. A typical stress test would be based on historical worst case moves over 10 business days (CGFS, 2010).
Margin is usually posted in the form of either cash or marketable government securities. Of the estimated $5 trillion of collateral posted against bilateral OTC derivative transactions at end-2014, about 78% was cash and 18% government securities. About 75% of VM margin was posted as cash, versus about 60% for IM. For centrally cleared transactions nearly 100% of VM was posted as cash versus about 35% for IM (ISDA, 2015b).
As with the underlying exposures, collateral is usually revalued on a daily basis. However, “haircuts” are often applied so that collateral requirements reflect the potential for values to decline between the time when the counterparty defaults and when the collateral is liquidated. A haircut is a discount applied to the posted collateral’s market value to reflect its credit, liquidity, and market risk. For example, a 5% margin requirement, meaning that the collateral to be posted has to be 105% of the collateral’s market value, would be equivalent to a haircut of 4.75%.
Multilateral compression can be used to reduce counterparty risk by eliminating redundant contracts that result from multiple bilateral transactions. Redundant trades are a byproduct of the practice of entering into offsetting trades instead of terminating trades. For example, suppose “A” and “B” enter into a 10-year swap, but a year later “A” wants to close the position. If “A” and “B” cannot agree on termination terms, “A” might enter into an offsetting 9-year swap with “C”. As a result “A” now has two swap contracts and double the counterparty risk. Multilateral compression can knock out the contract between “A” and “B” so for example, if “A” owes “B” $10, and “B” owes “C” $10, then “B” can be eliminated and “A” will owe “C” the $10. From 2003 through to month-end March 2015 compression eliminated $448 trillion notional of interest rate derivatives (ISDA, 2015a).
This post just scratches the surface of this very deep and complex topic. For more detailed coverage see Jon Gregory’s Counterparty Credit Risk and Credit Value Adjustment, and the Counterparty Credit Risk, Collateral and Funding book by Damiano Brigo , Massimo Morini, and Andrea Pallavinci. See John Hull’s Options, Futures and Other Derivatives for a broad coverage of derivative valuation theory and practice. For a deep dive into interest rate derivative pricing see Howard Corb’s Interest Rate Swaps and Other Derivatives, Columbia Business School, and Sadr, Amir, 2009, Interest Rate Swaps and Their Derivatives: A Practitioner's Guide, and for credit derivatives see Dominic O’Kane’s Modelling Single-name and Multi-Name Credit Derivatives. Also some important legal points were glossed over in the discussion of close-out netting, some of which are covered in the annex below.
Annex: Derivatives in Bankruptcy
U.S. and European derivative counterparties benefit from a variety of “safe harbors” if their counterparties enter into bankruptcy. They can offset winning against losing trades and if the amount is positive pay only a net amount. Normally the bankruptcy trustee can cherry pick by not performing on losing contracts and insisting on performance on winning ones (Pirrong, 2014). Amendments to the Bankruptcy Code in 2005 gave safe harbor to most U.S. derivatives in 2005. The European Union Directive on Financial Collateral Arrangements, which effectively harmonized bankruptcy safe harbor laws with U.S. laws, was passed on June 6, 2002 and was implemented in most member states by end-2005.
Safe harbor also allows derivatives counterparties to immediately access bankrupt counterparty posted collateral. Most other secured creditors are “stayed” from enforcing their rights. In addition they can immediately terminate (“closeout”) their trades, collect collateral to cover the bankrupt’s obligations, and become an unsecured creditor on the remainder. In addition, derivative contract counterparties are not subject to preference or fraudulent conveyance rules: the bankruptcy trustee can claw back cash taken out of a firm up to 90 days prior to its bankruptcy, except in the case of cash taken by derivatives (and repo) counterparties.
Derivatives were afforded this special treatment for financial stability reasons but this rationale is still being vigorously debated (Duffie and Skeel, 2012, and Schwarcz, 2014). And under pressure from the Financial Stability Board the International Swaps and Derivatives Association (ISDA) agreed to a new protocol that would introduce temporary stays to its Master Agreement. (See Major Banks Agree to Sign ISDA Resolution Stay Protocol and Backgrounder.) Under the protocol an initial set of 18 global systemically important banks and other large dealer-banks will be prevented from terminating swap trades with a failing bank for up to 48 hours. Buy-side firms are not included in the first phase because they may be unable to voluntarily adopt the protocol due to fiduciary responsibilities to their clients.