Stress tests at risk of becoming too complex, say bank heads

Regulator demands could lead to "tick-the-box" exercise, hear delegates at Quant Summit Europe

shutterstock-297573704

The increasing number and complexity of supervisory stress tests is likely to render them less effective, and risks turning them into a “tick-the-box” exercise, according to bank stress-testing heads.

Speaking as part of a panel discussion at Quant Summit Europe on April 13, Duško Dincov, global head of international banking stress-testing at Barclays, suggested complex stress-testing methodologies would make it more difficult to use the outputs as a tool to understand and manage the business through different cycles.

"There is a danger where, as we move to ever more complex statistical methods, as with loss modelling, that banks may reach a point where everybody is starting to model their income predictions in the same way with the same statistical tools, and it almost becomes disentangled from senior management and business line validation, and becomes just a statistical exercise," he said.

Since the 2008 financial crisis, the Bank of England (BoE), European Banking Authority (EBA) and US Federal Reserve Board have all begun conducting regular supervisory stress tests to evaluate banks’ resilience. The exercises, which include both quantitative and qualitative assessments, require banks to estimate losses under adverse economic conditions.

Speaking on the same panel, Anant Saxena, global head of stress-testing methodology and scenario design at Credit Suisse, said there was a danger of the tests turning into a “tick-the-box” exercise, with little practical value. "What's the value in that? Apart from obviously not getting in the bad books," he said.

My personal view is the internal debates that go on are sometimes a bit misplaced, because honestly speaking, no-one knows what is going to happen
Anant Saxena, Credit Suisse

Saxena said banks spent too much time and effort coming up with scenarios to use in stress testing.

"There is a lot of emphasis on getting the scenario specification right, and my personal view is the internal debates that go on are sometimes a bit misplaced, because honestly speaking, no-one knows what is going to happen,” he said. "It's almost a bit pointless to say that under this scenario we think the S&P 500 should be down 30%, not 25%, because it's not going to be 25% and it's not going to be 30%. It's going to be a third case that nobody knows.”

Saxena added: “Instead of focusing on getting the scenario right, which we know no-one is going to get right, it's more important to ask, 'Will the event hit our risk profile in a way that is relevant for us at senior management and board level?'"

The panel noted that while the EBA's stress tests were relatively formulaic and based on historical data, both the BoE’s stress tests and the Fed's Comprehensive Capital Analysis and Review (CCAR) require models that must comply with specific parameters of the given scenarios. Business divisions such as risk, treasury and finance are closely scrutinised, and a bank's loss, treasury and liquidity positions are stressed on a dynamic basis.

Banks say the demands associated with these supervisory stress tests have risen exponentially each year, and model-validation teams have grown dramatically since the introduction of the CCAR in 2011.

Supervisory stress tests are just one of an increasing number of areas where information on adverse scenarios was neccessary, said Barclays’s Dincov. He pointed to a range of other requirements, such as International Financial Reporting Standard 9 (IFRS 9), along with the Basel Committee's Fundamental review of the trading book (FRTB) and its rules on risk data aggregation, also known as BCBS 239, as places where banks could redeploy the same scenarios and data.

"Fundamental challenges" are similar

"I would say any bank in the UK is faced with a number of different challenges, and the fundamental challenges are pretty similar – whether it's IFRS 9, FRTB or BCBS 239 compliance. All of these initiatives use the same data, the same systems, very similar scenarios and scenario expansion tools in order to comply with whatever is prescribed as part of the new initiative," he said.

However, most banks are not there just yet. "I still see it as a few years down the line until we reach some sort of automation and stability that is required," Dincov added.

Additionally, it was worth bearing in mind that stress tests were not best placed to capture every adverse scenario, he said. One example was Brexit – the UK withdrawing from the European Union in the wake of a planned referendum on June 23.

“[The Brexit scenario] is really judgemental, and it would be a conversation between risk and finance and the business to assess what is likely to happen, given the composition of the portfolio," Dincov explained. "I wouldn't go down the route of the [more formulaic] scenario, because it's a thankless task in a way and it will be very difficult to validate."

  • LinkedIn  
  • Save this article
  • Print this page  

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact [email protected] or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact [email protected] to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: