Credit risk

Counterparty risk management: estimating extreme quantiles for a bank

Counterparty credit risk (CCR) is a complex risk to assess and banks did not have scientifically sound methods to calculate their level of potential exposure. Qiwei Yao, with his collaborators, has developed an innovative methodology for estimating counterparty credit risk, which can help banks meet regulatory requirements and calculate appropriate capital reserves.

Impact Case Series — Research Excellence Framework (REF)

The Basel III framework is a set of internationally agreed measures aimed at strengthening the regulation, supervision and risk management of banks, in response to the weaknesses revealed by the global financial crisis of 2007 to 2009. One of its requirements is better management of counterparty credit risk (CCR) – the risk of incurring a loss because another party to a contract fails to fulfill its part of the agreement. Under Basel III, investment banks such as Barclays are required to apply backtesting procedures to estimate their levels of counterparty credit risk and to ensure that they hold sufficient capital and liquidity to cover worst-case scenarios of potential losses. This strong emphasis on the proper management of the CCR by banks is important for the stability not only of individual banks but also, in light of their interconnections, of the financial system as a whole.

However, CCR is a complex risk to assess; as a hybrid of credit and market risk, it depends on both the evolution of the creditworthiness of the counterparty and the evolution of the underlying market risk factors. Banks such as Barclays previously lacked scientifically sound methods to calculate their level of potential exposure and instead adopted ad hoc approaches to calculate the level of financial cushion they need.

What have we done?

Between 2012 and 2014, I worked with the Quantitative Exposure Manager at Barclays to develop a more reliable and robust backtesting methodology for the bank to estimate potential future exposure to counterparty credit risk. I was invited to join the project because of my expertise in statistical analysis, especially in time series and dependent data. Backtesting is an analytical tool used by banks and their regulators to monitor the performance of their risk factor assessment methods. It uses historical price data to test the effectiveness of existing risk factor models. In particular, it tests whether the models’ extreme quantiles of potential future exposure – that is, the maximum credit exposure expected over the lifetime according to predetermined probabilities – are correctly quantified.

The first step is usually to simulate various future market risk factors such as interest rates, stocks and currency exchange rates. Then, all of the bank’s derivative positions are calculated at each time horizon of each of these simulated market scenarios, to determine the bank’s potential future exposure to counterparty default. The amount of capital required to cover counterparty credit risk is then calculated, in accordance with applicable regulations. For example, a backtesting setup can create a price pattern for an asset that can trade with different time frames at different prices. Two complicating factors are: (i) the interdependence of prices at different time horizons; and (ii) the explicit unavailability of prize distributions.

Although distributions are not available, banks store 1,000 simulated price trajectories as an approximation of these, allowing backtesting based on them. (Various constraints mean that most banks, including Barclays, can only store 1,000 simulated price paths.) Yao’s work addressed the challenge of estimating the extreme potential future exposure that would occur with odds between 1:5,000 and 1:10,000 based on the small samples available of 1,000 simulated price trajectories. His method takes advantage of the fact that the required extreme quantiles are determined by multiple random variables. The key idea here is that it is not necessary to go to extremes along a component variable in order to observe joint extreme events. This seemingly counterintuitive observation is critical to the success of the new approach, which, although easily demonstrable, had never been explored before in the literature or in practice. The resulting method developed by Yao and colleagues provides a satisfactory solution for quantifying extreme quantiles of potential future exposure with precision and reliability. The method is conceptually simple, theoretically sound and easy to implement – ​​and it provides robust performance in practice.

What happened?

Barclays holds counterparty credit risk-weighted assets worth tens of billions of US dollars. The methodology was applied to this entire portfolio, allowing the bank to calculate an appropriate capital buffer to protect both its own interests and those of its clients. For the bank, underestimating potential losses leads to exposure to potential uncovered financial losses. An overly conservative estimate creates unnecessary additional overhead and, therefore, increases in borrowing costs and decreases in investments. Since its first use by Barclays in late 2013, the new method has withstood rigorous backtesting under Basel III. This ensured that Barclays and its clients avoided exposing themselves to very risky positions. If its backtesting had failed, the bank would have been forced to increase its estimates of potential future exposure and therefore hold additional capital, which would be extremely costly. By avoiding this, the direct saving to the Bank from the use of this new scientifically calculated buffer is substantial. This, in turn, reduces the cost of borrowing and potentially increases investment and economic growth, with substantial indirect benefits for society.

The introduction of this new methodology improves the overall stability of Barclays. The interests of the bank and its customers are protected by reducing exposure to unhedged high-risk positions with a low probability (such as 0.05% or 0.01%). This in turn contributes to the stability of the global financial system by mitigating the potential impact of one bank’s failure on others. As such, research has indirectly contributed to providing greater security to the financial system at lower cost, with wider economic benefits.


  • This blog post appeared first on the LSE impact case studies page.
  • The post expresses the views of its author, not those of LSE Business Review or the London School of Economics and Political Science.
  • Image selected by Adam Nowakowski on Unsplash
  • When you leave a review, you agree to our Feedback Policy.