On June 23, 2016, the US Federal Reserve Board (FRB) announced the results of the Dodd-Frank Act Stress Tests (DFAST), which was the first of the two stages in bank stress tests in the US. On June 29, 2016, the FRB released the results of the second stage, the Comprehensive Capital Analysis and Review (CCAR).

The specifics of DFAST and CCAR 2016

DFAST 2016 covered bank holding companies (BHCs) supervised by the FRB that have $10 billion or more in total consolidated assets. For BHCs that have $50 billion or more in total consolidated assets, the FRB included both severely adverse and adverse scenarios covering the variables that capture economic activity, asset prices, and interest rates of various countries or country blocks.

Using these scenarios, along with the restrictive assumptions on dividend payment and stock repurchase, the FRB projected balance sheet values, risk weighted assets, pre-provision net revenue, net income, and capital ratios over a nine-quarter period beginning from the first quarter of 2016. BHCs with less than $50 billion in total consolidated assets will separately release their company-run stress test results on or before July 8, 2016.

As part of the capital plan under CCAR, BHCs with $50 billion or more in total consolidated assets were required to carry out their own stress tests with at least one internally developed scenario, in addition to choosing relevant FRB-designed scenarios.

What do the results say?

The results of DFAST and CCAR 2016 highlight the differences in business focus, asset composition, sources of revenue and expenses, and characteristics of portfolio risks across banks.

DFAST:

• In the case of severely adverse scenarios, aggregate losses were projected to be $526 billion, with a significant amount ($385 billion) being accounted for by accrual loan losses. The other two sources of losses were trading and counterparty default. Together, the top three losses accounted for 95% of the projected losses.
• In the adverse scenario, aggregate losses were pegged at $324 billion, with $252 billion being accounted for by accrual loan losses.

CCAR:

• Large, non-complex BHCs fared better than the complex ones, both in terms of improvements over time, and in meeting supervisory expectations. Though many large complex BHCs have made progress since CCAR 2015, they continue to fall short of FRBs expectations, especially in the areas of risk identification, assessment of stress testing assumptions and results, and internal controls for capital planning.
• The FRB did not object to the capital plan of any BHC on quantitative grounds, just that M&T Bank Corporation had to adjust its capital actions to maintain post-stress regulatory capital above minimum requirements.
• The capital plans of Deutsche Bank Trust Corporation and Santander Holdings USA were objected to on qualitative grounds, due to material unresolved supervisory issues. These BHCs had faced the same fate in CCAR 2015. It is important to note here that both these BHCs displayed a strong position in terms of quantitative criteria, that is, the fulfillment of minimum capital ratios.
• A large, complex BHC Morgan Stanley received a conditional non-objection to its capital plan due to deficiencies in scenario design, modeling practices, and governance and control around scenario design and modeling practices. This BHC too had a strong position in terms of quantitative criteria.

Implications for banks

Based on these results, we have identified the following areas that banks need to focus on:

Stress testing as a strategic choice: Many banks continue to view supervisory and bank-specific stress testing as regulatory mandates, which makes them an operational task that does not get the requisite attention from the board of directors and the top management. This should change. Financial institutions must view stress tests as strategic tools to effectively manage financial, business, and operational risks.

Integrated stress testing: Despite advances in enterprise-wide risk management, risk analytics, and modeling, stress testing in many banks is carried out in silos, at the business unit level. For example, credit losses are derived at the product level and simply summed up to arrive at the portfolio level losses. This defeats the larger purpose since it omits portfolio-level interrelationships, often resulting in inadequate identification of risks. Centralizing stress testing with a judicious balance between top-down and bottom-up considerations is the best way forward.

Risk and finance integration: Modern stress testing such as DFAST and CCAR involve simultaneous modeling of risk and finance factors, unlike in the past when stress testing was centered primarily on risk factors. This would require due diligence on the part of banks to choose the most suitable approach for risk and finance integration.

Model governance and validation: Stress testing is all about relying on a set of assumptions, scenarios, and models. Since these can go wrong or may not perform as expected, the need of the hour is a board-level involvement and independent validation of assumptions, models, and so on, at regular intervals.

With increasing asset quality issues, trading losses, counterparty defaults, and operational losses across the globe further threatening the profitability and stability of banks, stress testing would continue to be the most important tool for regulators and banks to ensure stability of the financial ecosystem.