Stress Testing Banks

Capital and liquidity can be categorized into 3 forms, namely (1) the capital or liquidity you have, (2) the capital or liquidity you need, and (3) the capital or liquidity regulators think you need. The right side of the balance sheet is addressed by capital adequacy while liquidity addresses the left side. In this chapter, the framework banks apply for stress testing is laid out. Capital and liquidity adequacy also applies the framework. A macro-prudential stress test of crises arising in the financial sector is the Supervisory Capital Assessment Program (SCAP) from the Federal Reserve, which has enabled banks to move to applying broad macro-scenarios with market-wide stresses from a single factor. The following are differences between pre- and post-SCAP stress testing.

  • Pre-SCAP is mostly single shock while post-SCAP involves a broad macro scenario and market stress;
  • Pre-SCAP applies to product or business unit level while post-SCAP is more comprehensively applied, firm-wide;
  • Pre-SCAP is static while post-SCAP mostly path dependent and dynamic;
  • While pre-SCAP is not usually tied to capital adequacy, pre-SCAP is an explicit post-stress common equity threshold; and
  • Pre-SCAP applies to losses only whereas post-SCAP is for losses, revenues, and costs.

Stress Testing in the Literature

Found in the relatively data-rich trading room environment, stress testing is a risk management discipline and is closely related to treasury function conducting interest rate scenarios and shocks. Stresses will typically take scenarios or sensitivities. In a data-rich environment, these stresses lend themselves to naturally understand financial risks. Nonfinancial risks heavily rely on scenario analysis and are harder to quantify. Given that quantitative credit risk modeling is a newer discipline, the banking books’ formal stress testing is more recent and is dominated by credit risk.

Designing the Stress Scenario

Coherence is the main limitation supervisors and firms face as they design stress scenarios. Risk factors are dynamic and all move simultaneously. However, specifying a joint outcome that is coherent for all risk factors that are relevant is challenging. A set of assets considered safe haven must be possessed by every market shock scenario that causes flights from risky assets. When considering stress scenarios for market risk, the challenge of coherence becomes more acute as compared to scenario design where it is generic. The challenge of finding a jointly coherent real and financial factor scenario makes this challenge even more compounded. GDP growth, unemployment and house price index (HPI) are the three dimensions of the state space as market risk scenario was historical experience-based. However, historical scenarios never test for something new. Depending on business mix and geographic footprint, some scenarios, imposed by supervisory stress tests on all banks, may be more severe to some banks as compared to others. As a result, the 2011 and 2012 Comprehensive Capital Analysis and Review (CCAR) asked banks to submit results based on their own scenarios in addition to the common supervisory stress scenario. This enabled supervisors to learn what the banks considered to be high-risk scenarios and undertake a direct comparison of results across banks from common scenarios without sacrificing risk-discovery.

Executing on the Stress Scenario: Losses and Revenues

Supervisory stress tests are in the U.S. and Europe contributed to by accompanying methodology documents disclosed by supervisors, which are focused heavily on banking books.

Modeling Losses

For particular products by geography, mapping into many immediate risk factors driving losses from few factors is the first task of a market active firm. Spanning 21 economically different sovereign nations was a way to directly confront the geographic heterogeneity problem by the European Banking Authority (EBA). Mapping from macro to a more intermediate risk factors problem is not limited to geography. Moreover, the problem of loose coupling of business cycle loss severity is not limited to auto loans. The question of whether the industry of the defaulted firm is distressed at the time of the default is an important LGD determinant for corporate credit. With respect to the position that the risk factor mapping problem has been solved, by using scenario risk factors, existing positions are simply re-priced. Receiving less attention is the counterparty credit risk stressing problem associated with derivatives activities.

Modeling Revenues

Banks’ total income can be divided according to interest and non-interest income. Due to interest rate hedging strategies, the net impact of rising or falling rates on profitability are ambiguous. On non-interest rate income, interest rate scenarios impact is not harder to assess.

Modeling the Balance Sheet

To define capital adequacy, the capital ratio is computed by capital over asset where the relevant capital is common equity and the denominator is the risk-weighted assets (RWA). To determine the risk-weights, we often use the Basel frameworks. For a two-year period, the modeling of both the income statement and the balance sheet are modeled to determine capital adequacy on a post-stress basis, regardless of the risk weight regime. The balance sheet is the point at which the required capital ratio is met by the bank. The challenge when modeling a balance sheet is to use the static raw form, risk-weighted form, or dynamic balance sheet assumption.

Stressing Testing Disclosure

There are large disclosure differences across distinct supervisory stress tests. Regulatory disclosures used to report realized losses only and not protected losses, thereby allowing markets to not only check the stress tests’ sovereignty in terms of scenario but also the resulting outcome at the bank level. The design and disclosure regimes in normal times diverge between Europe and the U.S. The benefits of stress tests disclosure may not outweigh the costs as banks sometimes make poor portfolio choices for maximizing its chances of passing the test, hence giving up longer-term value. Still, some disclosures are preferable compared to non-disclosure. Due to the averaging away of idiosyncratic errors while estimating bank conditions, aggregation is merited for being right.

The supervisors’ ability to correctly assess individual firms’ health in times of crises, and the markets’ inability to distinguish a good bank from a bad one, make the benefit of detailed bank-specific stress test disclosures very significant. It is easy to see the importance of a pop quiz in the form of bespoke stress testing between balance sheet opacity, regulatory arbitrage, and asset substitution.

The purpose of regulatory capital models, stress testing models, and economic capital models has always been to determine the amount of needed capital to support the ongoing activities of the bank. Due to a slow evolution of regulatory and economic capital models, the difficulty for banks and countries is in adapting to financial innovations and macro-conditions that are rapidly changing. By construction, stress tests are adapted to bank portfolios and the current environment.

Conclusion

In an entire banking system, a broad-based supervisory stress test is a new aspect being used. Result disclosures happen to be a critical component in the exercise. Due to the lack of information by publicly disclosed existing approaches, stress testing became an imperative. Consequently, supervisory authorities decided that disclosure should only be limited to allow math-checking by the market.

Practice Questions

1) The Supervisory Capital Assessment Program (SCAP) took place in the spring of 2009 during the financial crisis of 2008-2009 and was intended to measure the financial strength of the nation’s 19 largest financial institutions going forward. SCAP being the first macro-prudential stress test has some features to distinguish. Which of the differences listed below is NOT a feature that correctly describes the differences between pre-SCAP and post-SCAP?

  1. Post-SCAP applies to a broad macro scenario and market stress while pre-SCAP is mostly single shock.
  2. Pre-SCAP is dynamic while post-SCAP is static and path prudent.
  3. Post-SCAP applies to losses, revenues, and costs whereas pre-SCAP applies to losses only.
  4. Whereas pre-SCAP is not usually tied to capital adequacy, post-SCAP is explicit post-stress common equity threshold.

The correct answer is B.

To the contrary, post-SCAP is actually dynamic and path dependent as compared to pre-SCAP which is quite static.

The following are the correct differences between pre- and post-SCAP stress testing.

  • Pre-SCAP is mostly single shock while post-SCAP involves a broad macro scenario and market stress;
  • Pre-SCAP applies to product or business unit level while post-SCAP is more comprehensively applied, firm-wide;
  • Pre-SCAP is static while post-SCAP mostly path dependent and dynamic;
  • While pre-SCAP is not usually tied to capital adequacy, pre-SCAP is an explicit post-stress common equity threshold; and
  • Pre-SCAP applies to losses only whereas post-SCAP is for losses, revenues, and costs.

Leave a Comment

X