Counterparty credit risk is a complex risk for investment banks to measure and quantify.

There are a number of reasons for this complexity.

Firstly, counterparty risk is bilateral and not unilateral.

This means unlike a loan where the lender is the only party with credit risk, counterparty risk can change over time.

At a certain point during a contract's life, one counterparty may be winning from the contracts, creating the risk that they do not get paid what they are owed in the future.

Or in other words, they face counterparty credit risk.

However, as market conditions change over time, the first counterparty may end up in a losing position from the contract, meaning they no longer face counterparty risk.

Since they are no longer owed any money under the contract and the counterparty risk exposure will have transferred to the other party, to the trade, since they are now owed money from the trade, you need to be winning from a trade in order to face counterparty risk, since that's the only circumstance where the counterparty owes you money.

This results in counterparty risk being a hybrid between credit and market risk, depending on both changes in the credit worthiness of the counterparty and movements in underlying market risk factors.

This dynamic relationship can make counterparty risk difficult to quantify. Like lending credit risk however, the quantification of counterparty risk is ultimately driven by the probability of default or PD and the size of the exposure at default, or EAD.

The probability of default can be determined using the same methodology as would be used to assess standard credit risk, such as using a bank's own internal credit assessment tools or external credit rating agency ratings.

Also, the probability of default can be inferred from credit spreads on the company's debt or from credit derivatives, such as credit default swaps.

However, the size of the exposure at default is much harder to quantify since the exposure at the time a counterparty may default in the future is unknown today since it depends on market conditions at the time of default.

Although quantifying exposure at default is complex, requiring sophisticated modeling, the factors which need to be known to estimate the exposure at default include the current exposure details of any netting arrangements that might be in place, allowing for positive and negative exposures with the same counterparties to be netted off should default occur, and whether any collateral is currently in place or may be in place in the event of default in the future.