Counterparty risk is the risk associated with the other party to a financial contract not meeting its obligations. Every derivative trade needs to have a party to take the opposite side.
Credit default swaps, a common derivative with counterparty risk, are often traded directly with another party, as opposed to trading on a centralized exchange. Since the contract is directly connected with the other party, there is a greater risk of counterparty default since both parties may not have full knowledge of the financial health of the other (and their ability to cover obligations). This differs from products listed on an exchange. In this case, the exchange is the counterparty, not the single entity on the other side of the trade.
Counterparty risk gained visibility in the wake of the global financial crisis. AIG famously leveraged its AAA credit rating to sell (write) credit default swaps (CDS) to counterparties who wanted default protection (in many cases, on CDO tranches). When AIG could not post additional collateral and was required to provide funds to counterparties in the face of deteriorating reference obligations, the U.S. government bailed them out.
Regulators were concerned that defaults by AIG would ripple through the counterparty chains and create a systemic crisis. The issue was not only individual firm exposures but the risk that interconnected linkages via derivative contracts would jeopardize the whole system.
A Credit Derivative Has Counterparty Risk
While a loan has default risk, a derivative has counterparty risk. Counterparty risk is a type (or sub-class) of credit risk and is the risk of default by the counterparty in many forms of derivative contracts. Let’s contrast counterparty risk to loan default risk. If Bank A loans $10 million to Customer C, Bank A charges a yield that includes compensation for default risk. But the exposure is easy to ascertain; it’s roughly the invested (funded) $10 million.
A credit derivative, however, is an unfunded bilateral contract. Aside from the posted collateral, a derivative is a contractual promise that might be broken, thus exposing the parties to risk. Consider an over-the-counter (OTC) option sold (written) by Bank A to Customer C. Market risk refers to the fluctuating value of the option; if it is daily-mark-to-market, its value will be a function largely of the underlying asset price but also several other risk factors. If the option expires in-the-money, Bank A owes the intrinsic value to Customer C. Counterparty risk is the credit risk that Bank A will default on this obligation to Bank C (for example, Bank A might go bankrupt).
Understanding Counterparty Risk with an Interest Rate Swap Example
Let’s assume two banks enter into a vanilla (non-exotic) interest rate swap. Bank A is the floating-rate payer, and Bank B is the fixed-rate payer. The swap has a notional value of $100 million and a life (tenor) of five years. It is better to call the $100 million notional instead of principal because the notional is not exchanged. It is merely referenced to compute the payments.
To keep the example simple, assume the LIBOR/swap rate curve is flat at 4%. In other words, when the banks begin the swap, spot interest rates are 4% per annum for all maturities.
The banks will exchange payments at six-month intervals for the swap’s tenor. Bank A, the floating-rate payer, will pay a six-month LIBOR. In exchange, Bank B will pay the fixed rate of 4% per annum. Most importantly, the payments will be netted. Bank A cannot predict its future obligations, but Bank B has no such uncertainty. At each interval, Bank B knows it will owe $2 million: $100 million notional * 4% / 2 = $2 million.
Let’s consider counterparty exposure definitions at two points in time – at swap inception (T = 0) and six months later (T = + 0.5 years).
At the Start of the Swap (Time Zero = T0)
Unless a swap is off-market, it will have an initial market value of zero to both counterparties. The swap rate will be calibrated to ensure a zero market value at swap inception.
- The market value (at T = 0) is zero for both counterparties. The flat spot rate curve implies 4.0% forward rates, so the floating-rate payer (Bank A) expects to pay 4.0% and knows it will receive 4.0%. These payments will net to zero, and zero is the expectation for future netted payments if interest rates do not change.
Credit exposure (CE)
This is the immediate loss if the counterparty defaults. If Bank B defaults, the resulting loss to Bank A is Bank A’s credit exposure. Therefore, Bank A only has credit exposure if Bank A is in-the-money. Think of it like a stock option. If an option holder is out-of-the-money at expiration, default by the option writer is inconsequential. The option holder only has credit exposure to default if they are in-the-money. At swap inception, as the market value is zero to both, neither bank has credit exposure to the other. For example, if Bank B immediately defaults, Bank A loses nothing.
Expected exposure (EE)
This is the expected (average) credit exposure on a future target date conditional on positive market values. Bank A and Bank B both have expected exposure at several target future dates. Bank A’s 18-month expected exposure is the average positive market value of the swap to Bank A, 18 months forward, excluding negative values (because default won’t hurt Bank A under those scenarios). Similarly, Bank B has a positive 18 month expected exposure, which is the market value of the swap to Bank B but conditional on positive values to Bank B. It helps to keep in mind that counterparty exposure exists only for the winning (in-the-money) position in the derivative contract, not for the out-of-money position! Only a gain exposes the bank to counterparty default.
Potential future exposure (PFE)
PFE is the credit exposure on a future date modeled with a specified confidence interval. For example, Bank A may have a 95% confident, 18-month PFE of $6.5 million. A way of saying this is, “18 months into the future, we are 95% confident that our gain in the swap will be $6.5 million or less, such that a default by our counterparty at the time will expose us to a credit loss of $6.5 million or less.” (Note: by definition, the 18-month 95% PFE must be greater than the 18-month expected exposure (EE) because EE is only a mean.) How is the $6.5 million figured? In this case, the Monte Carlo simulation showed that $6.5 million is the upper fifth percentile of simulated gains to Bank A. Of all simulated gains (losses excluded from the results because they do not expose Bank A to credit risk), 95% are lower than $6.5 million, and 5% are higher. So, there is a 5% chance that, in 18 months, Bank A’s credit exposure will be greater than $6.5 million.
Does potential future exposure (PFE) remind you of value at risk (VaR)? Indeed, PFE is analogous to VaR, with two exceptions. First, while VaR is an exposure due to a market loss, PFE is a credit exposure due to a gain. Second, while VaR typically refers to a short-term horizon (for example, one or 10 days), PFE often looks years into the future.
There are different methods for calculating VaR. VaR is a quantile-based measure of risk. For a certain portfolio and time horizon, VaR provides the probability of a certain amount of loss. For example, a portfolio of assets with a one-month 5% VaR of $1 million has a 5% probability of losing more than $1 million. Thus, the VaR can at least provide a hypothetical measure of the risk of counterparty default on a credit default swap.
The most common method to calculate VaR is historical simulation. This method determines the historical distribution of profit and losses for the portfolio or asset being measured over a previous period. Then, the VaR is determined by taking a quantile measurement of that distribution. Although the historical method is commonly used, it has significant drawbacks. The main problem is this method assumes that the future return distribution for a portfolio will be similar to that of the past. This may not be the case, especially during periods of high volatility and uncertainty.
Go Forward Six Months in Time (T = + 0.5 years)
Let’s assume the swap rate curve shifts down from 4.0% to 3.0%, but remains flat for all maturities so it is a parallel shift. At this time, the swap’s first payment exchange is due. Each bank will owe the other $2 million. The floating payment is based on the 4% LIBOR at the beginning of the six month period. In this way, the terms of the first exchange are known at swap inception, so they perfectly offset or net to zero. No payment is made, as planned, at the first exchange. But, as interest rates changed, the future now looks different…better to Bank A and worse to Bank B (who is now paying 4.0% when interest rates are only 3.0%).
Current exposure (CE) at time T + 0.5 years
Bank B will continue to pay 4.0% per annum but now expects to receive only 3.0% per annum. Since interest rates have dropped, this benefits the floating-rate payer, Bank A. Bank A will be in-the-money, and Bank B will be out-of-the-money.
Under this scenario, Bank B will have zero current (credit) exposure; Bank A will have positive current exposure.
Estimating the current exposure at six months: We can simulate the future current exposure by pricing the swap as two bonds. The floating-rate bond will always be worth approximately par; its coupons are equal to the discount rate. The fixed-rate bond, at six months, will have a price of about $104.2 million. To get this price, we assume a 3.0% yield, nine semi-annual periods remaining, and a $2 million coupon. In MS Excel the price = PV(rate = 3%/2, nper = 9, pmt = 2, fv = 100); with a TI BA II+ calculator, we input N = 9, I/Y = 1.5. PMT = 2, FV = 100, and CPT PV to get 104.18. So if the swap rate curve shifts in parallel from 4.0% to 3.0%, the market value of the swap will shift from zero to +/- $4.2 million ($104.2 – $100). The market value will be +$4.2 million to in-the-money Bank A and -$4.2 million to out-of-the-money Bank B. But only Bank A will have a current exposure of $4.2 million (Bank B loses nothing if Bank A defaults). In regard to expected exposure (EE) and potential future exposure (PFE), both will be re-calculated (actually, re-simulated) based on the freshly observed, shifted swap rate curve. However, as both are conditional on positive values (each bank includes only the simulated gains where credit risk can exist), they will both be positive by definition. As interest rates shifted to the benefit of Bank A, Bank A’s EE, and PFE are likely to go up.
Summary of the Three Basic Counterparty Metrics
- Credit exposure (CE) = MAXIMUM (Market Value, 0)
- Expected exposure (EE): AVERAGE market value on future target date, but conditional only on positive values
- Potential future exposure (PFE): Market value at specified quantile (for example, the 95th percentile) on future target date, but conditional only on positive values
How Are EE and PFE Calculated?
Because derivative contracts are bilateral and reference notional amounts, which are insufficient proxies for economic exposure (unlike a loan where the principal is real exposure), in general, we must use Monte Carlo simulation (MCS) to produce a distribution of market values on a future date. The details are beyond our scope, but the concept is not as difficult as it sounds. If we use the interest rate swap, four basic steps are involved:
1. Specify a random (stochastic) interest rate model. This is a model that can randomize underlying risk factor(s). This is the engine of the Monte Carlo Simulation. For example, if we were modeling a stock price, a popular model is geometric Brownian motion. In the example of the interest rate swap, we might model a single interest rate to characterize an entire flat rate curve. We could call this a yield.
2. Run several trials. Each trial is a single path (sequence) into the future; in this case, a simulated interest rate years into the future. Then, we run thousands more trials. The chart below is a simplified example: each trial is a single simulated path of an interest rate plotted ten years forward. Then, the random trial is repeated ten times.
3. The future interest rates are used to value the swap. So, just as the chart above displays 10 simulated trials of future interest rate paths, each interest rate path implies an associated swap value at that point in time.
4. At each future date, this creates a distribution of possible future swap values. That’s the key. See the chart below. The swap is priced based on the future random interest rate. At any given future target date, the average of the positive simulated values is the expected exposure (EE). The relevant quantile of the positive values is the potential future exposure (PFE). In this way, EE and PFE are determined from the upper half (the positive values) only.
Dodd-Frank Act
Defaults on swap agreements were one of the main causes of the 2008 financial crisis. The Dodd-Frank Act enacted regulations for the swaps market. It included provisions for public disclosures of swap trades, as well as authorizing the creation of centralized swap execution facilities. Trading swaps on centralized exchanges reduces counterparty risk. Swaps traded on exchanges have the exchange as the counterparty. The exchange then offsets the risk with another party. Since the exchange is the counterparty to the contract, the exchange or its clearing firm will step in to meet the obligations of the swap agreement. This dramatically reduces the likelihood of counterparty default risk.
The Bottom Line
Unlike a funded loan, the exposure incurred in a credit derivative is complicated by the issue that value can swing negative or positive for either party to the bilateral contract. Counterparty risk measures assess current and future exposure, but Monte Carlo simulation is typically required. In counterparty risk, exposure is created with a winning in-the-money position. Just as value at risk (VaR) is used to estimate market risk of a potential loss, potential future exposure (PFE) is used to estimate the analogous credit exposure in a credit derivative.
Read the original article on Investopedia.