With post-crisis recapitalisations hopefully in the past, stress-testing has become a means to push banks to enhance their risk and data governance.

As regulators prepare a new round of stress-tests, there is growing focus on enhancing risk management and information about systemic risks rather than just capital adequacy. The Bank of England unveiled a new strategy for stress-testing alongside the 2015 test, the results of which were published on December 1, 2015. The European Banking Authority (EBA) has also given more insight into how its 2016 stress-test will be conducted, with scenarios due to be published in February 2016, and results released in the third quarter.

Speaking at a conference at the London School of Economics in October 2015, the US Federal Reserve’s deputy director for financial stability policy, Andreas Lehnert, emphasised the difference between what he called “war-time and peace-time” stress-tests. The first type is designed to identify capital shortfalls and re-establish the credibility of the financial system and regulators during a crisis.

“Now we have moved beyond that stage and are well into an economic recovery, stress-testing has become part of routine process for regulators and for authorities that are accountable for financial stability. Our thinking is that we will go into the next downturn with a much better capitalised banking system and with a set of quantitative tools with forward-looking, imaginative hypotheses about what the risks are in the system,” said Mr Lehnert.

The 2016 stress-test of 53 EU banks provides an unambiguous example of this change. The EBA has decided not to set a pass/fail threshold for how much capital banks must be left with at the end of the adverse scenario.

“However, competent authorities will apply stress-test results as an input to the supervisory review and evaluation process,” the EBA said in an explanatory note.

Test of governance

On a similar note, all seven banks assessed passed the UK stress-test in December 2015 in terms of their capital ratios. But the Bank of England’s report included lengthy observations on the shortcomings of data and process management at the banks.

“For some banks, the coverage, scope and adequacy of model management standards were found to have improved. But others needed to make considerable improvements, including implementing and embedding model management policies more fully. Some banks lacked formal processes to approve stress-testing models and had weak model governance,” the prudential regulator noted.

In addition, the report flagged deteriorating data quality in a number of areas, including net interest income, traded risk and structured finance.

“While the quantitative impact of stress-testing remains important for capital and dividends planning, and the setting of regulatory buffers, the focus of regulators is increasingly on the provision of reconciled granular data to describe the detailed risks banks are exposed to, and to test that banks are able to effectively understand, quantify and manage these risks,” says Gerald Chappell, a partner in the financial services advisory division at EY.

This also takes the Bank of England down a similar road to the US Federal Reserve, which has in the past objected to bank capital plans in the comprehensive capital analysis and review (CCAR) due to perceived quantitative failings. Adam Girling, a principal in EY’s financial services office in the US, says future CCAR exercises could pose a higher capital hurdle if the surcharge for global systemically important banks (GSIBs) is included. Nonetheless, the distinguishing feature of the CCAR in general is the level of sophistication required for internal scenario development.

“The CCAR is not just a risk or loss-based stress-test, but rather a comprehensive financial statement and regulatory capital projection exercise. Banks need to look at how their balance sheet, income statement, risk-weighted assets and related liquidity needs and funding costs dynamically evolve under a stress scenario, and the leading US banks have incorporated all of those aspects and their interdependencies,” says Mr Girling.

Management tool?

Mr Girling believes firms still need to develop more robust infrastructure to extend the stress-testing process to the business-as-usual capital allocation of the bank. In the meantime, banks are already using stress-tests to identify particular sensitivities in their risk-weighted assets models or potential sources of loss.

“The integration of stress-testing with risk appetite, business planning and return-on-equity management is still an aspiration, rather than being fully embedded at most banks. That is partly because banks have had to respond rapidly to regulatory stress-test demands that impose reporting views, assumptions and methodology requirements which are very different from their business-as-usual approaches to managing risks. So they need to focus on being able to comply with the exercise first, and then look to extract value by feeding into management decision making,” says Mr Chappell.

The UK Prudential Regulatory Authority is pushing hard to oblige banks to integrate stress-testing into management and risk appetite decisions. Alex Brazier, Bank of England executive director for financial stability strategy and risk, announced in October 2015 that UK tests would now be used as a countercyclical tool. Stress scenarios defined by the regulator will be more severe at the top of the cycle, to oblige banks to build capital buffers during the good times that can be depleted during downturns. Perhaps the most radical step is the requirement announced by Mr Brazier for banks to assume continued lending growth of 10% per year for each of the five years in the stress scenario.

“We’re not just interested in whether banks stay afloat. We want the system to be strong enough to continue to serve the real economy, even in the storm,” said Mr Brazier.

In other respects, the Bank of England stress-test is dynamic, with banks allowed to set out how they would respond to the shock. Those management responses themselves are carefully examined by supervisors, but may sit awkwardly with the loan growth requirement.

“It is true that you do not want a self-reinforcing cycle of deleveraging, but one has to think about what would be happening in the real economy in the stress scenario. Loan-to-value ratios would be deteriorating significantly, so it may not be realistic to expect banks to preserve their risk appetite in those conditions,” says Cecilia Gejke, head of stress-testing at Japan’s Mizuho International in London.

Expected credit losses

Stress-testing is just one aspect of the pressure on banks to strengthen internal risk data governance. Another major change is the adoption of the IFRS 9 international financial reporting standard for credit loss accounting, which is set for adoption from January 2018. This will switch financial reporting from an incurred loss to an expected loss protocol. Assets must be moved from one-year expected loss accounting to lifetime expected loss in the event of a “significant deterioration” in credit quality.

In December 2015, the enhanced disclosure task force (EDTF), which is convened by the Financial Stability Board, published a detailed 32-page report on the impact of expected credit loss approaches on bank risk disclosures. The report concluded that: “For many banks, significant changes to systems and processes may be required, which will require substantial time and resources to deliver. Some banks will need to develop and enhance governance over the recognition and measurement of credit losses, particularly to develop capability to make informed judgements about the use of forward-looking information.”

Tara Kengla, an assurance partner in the financial services team at EY, says there are estimates already in bank financial statements, including the current loan loss provisioning numbers. But expected credit loss will require more datapoints, models, processes and controls to audit, potentially leading to further guidance on auditing this complex estimate.

“With forward-looking lifetime calculations there are additional modelling needs and inputs, and there will also be a need for additional credit risk modelling specialists and economists to think about how to incorporate the macroeconomic factors and forecasting,” says Ms Kengla.

To some extent, deciding the trigger point between one-year and lifetime losses will place greater emphasis on the governance elements that are also under scrutiny in stress testing. The Bank of England noted in its December 2015 report that certain aspects of the stress-test required expert judgement that went beyond pure financial models.

“There were features of the scenario, such as falling corporate profits in the UK, that [we] had expected banks to consider separately through the use of judgement and quantitative analysis. Such analysis was not evident in banks’ submissions,” the regulator concluded.

The EDTF is recommending that GSIBs disclose the quantitative impact of IFRS 9 on their financial statement from 2017. Ms Kengla emphasises, however, that this 2017 disclosure will be based on the portfolio at that point in time, so investors will not yet be able to assess the effects of IFRS 9 on actual reported numbers from 2018 onward.


All fields are mandatory

The Banker is a service from the Financial Times. The Financial Times Ltd takes your privacy seriously.

Choose how you want us to contact you.

Invites and Offers from The Banker

Receive exclusive personalised event invitations, carefully curated offers and promotions from The Banker

For more information about how we use your data, please refer to our privacy and cookie policies.

Terms and conditions

Top 1000 2023

Request a demonstration to The Banker Database

Join our community

The Banker on Twitter