Enterprise Stress Testing 2.0: Scenario-based analysis

  • By Paul Franks
  • SAS

SAS Director of Financial services Paul Franks assesses the latest developments in stress testing and the crucial role this approach still plays in ensuring banks continue to address competing risks. 

Regulators, locally and globally, continue to impose more regular and stringent stress tests involving a limited set of prescribed downturn scenarios to financial institutions to evaluate their capital adequacy and resilience.  

Whilst many are passing the fundamental tests, there is a significant concern that many utilise sub-optimal processes and infrastructure for performance mandated enterprise stress tests.  Banks not meeting expectations are being called upon to remediate and improve and move beyond a pass mark for compliance.

Regulatory stress testing requires a significant investment by financial institutions in technology, skilled people and time, and the process itself continues to become even more complex as programs mature and regulatory expectations increase.

Savvy financial institutions are looking for and realising a significant return on investment by reaching beyond compliance. They are seeing more effective, consistent analytical processes. 

Also, an ability to address complex questions from Board and executive management as to the sensitivity of financial performance to changes in macroeconomic factors. 

Rethinking your stress test framework

The regulatory stress test framework in place today performs capital adequacy-oriented stress testing over a multiperiod forecast horizon. 

At each period, a scenario exerts its impact on the net profit or loss based on the as-of-date business, including portfolio balances, exposures, and operational income and costs. 

The net profit or loss, after being adjusted by other financial obligations and management actions, will deter¬mine the available capital that is available for the next period on the scenario path. 

A natural evolution of the portfolio and business under a given scenario leads to a state of the business at the next horizon, which then starts a new evaluation of the available capital. 

The risk profile of this business evaluation also determines the capital requirement under the same scenario. The capital adequacy assessment can be performed through this dynamic analysis of capital supply and demand. 

It can become a daunting exercise on a multiperiod scenario because of the forward-looking and path-dependent nature of the analysis.

Stress testing has greatly improved banking sector resilience. In regions where stress testing capabilities are more mature, banks have built up adequate capital and have performed well in recent years.

As these programs mature, many jurisdictions are raising their requirements, both quantitively and qualitatively. 

Stress testing and scenario analysis are part of Pillar 2 in the Internal Capital Adequacy Assessment Process (ICAAP). 

Institutions are expected to use stress tests and scenario analyses to improve their understanding of the vulnerabilities that they face under a wide range of adverse conditions. 

These demands will require the adoption of new technologies and best practices. Modernisation of stress test capabilities will be essential.

New institutional capabilities

As stress testing capabilities have matured, people are exploring how to use these capabilities for strategic business purposes. 

There are many high-impact strategic use cases for stress testing across the enterprise including:

(1) financial planning

(2) risk appetite management

(3) what-if and sensitivity analysis

(4) emerging risk identification, and 

(5) reverse stress testing.

Stress testing is one form of scenario-based analysis. Scenario-based analysis is also useful for forward-looking financial planning exercises such as analysing the impact of changing environmental factors and portfolio shifts on overall financial performance and balance sheet.

Banks can leverage the technologies and processes used for regulatory stress testing. However, infrastructure and program processes must be developed so that business-as-usual scenarios and alternatives can be easily managed, and the models and assumptions can be adjusted.

A closely related topic to stress testing and capital planning is risk appetite. A dynamic relationship exists between stress testing, risk appetite and capital planning:  

1) Risk appetite is defined by the institution to reflect its capital strategy, return targets and its tolerance for risk 

2) Capital planning is conducted in alignment with the stated risk appetite and risk policy and 

3) Scenario-based analyses is carried out to ensure a bank can operate within the risk appetite under a range of scenarios, typically, planning, baseline and stressed

Scenario analysis to the fore

COVID-19 has presented a sudden shock to the financial plans of lending institutions. Leading banks seeking a timely understanding of the potential financial impacts have increasingly turned to scenario analysis. For scenario analyses to be meaningful, the process must:

(1) scale to an increasing array of input scenarios as the situation continues to develop

(2)  provide a controlled process to perform and summarise numerous iterations of analysis

(3) provide understandable and explainable results in a timely fashion

(4) provide process transparency and control for qualitative and quantitative assumptions and 

(5) maintain detailed data to support ad hoc reporting and concentration analysis.

Faster, richer what-if analysis is perhaps the most powerful and demanding way to extend a bank’s stress testing utility.  

Sensitivity analysis usually supplements stress testing and differs from other scenario- based analyses in that it’s usually defined parametrically to answer questions about scenario, assumption and model deviations.

For modelling purposes, sensitivity tests can be viewed as an expanded set of scenario analyses. If banks perform sensitivity tests, they must be able to scale their infrastructure to complete large numbers of tests within a reasonable time frame and be able to easily compare the results.

Econometric-based stress testing of portfolio-level credit, market, interest rate and liquidity risks is now a relatively established practice but its application to reputational and strategic risks is less mature.  Scenario-based analysis provides a viable solution, though it requires proper translation from the scenarios involving these risks into a scenario that can be modelled.

We are in unprecedented times and stress testing capability has never been more important as multiple stressor events are occurring simultaneously. How does your institution fare in its operational maturity and capability for enterprise stress testing?  Will you exceed regulatory expectations and requirements and earn a distinction grade, or will you just receive a pass mark?