Seven steps may help banks develop better stress tests.
In the beginning, stress tests were ad hoc initiatives launched by national regulators to assess the current state of their domestic banking industry in the aftermath of severe external shocks. Increasingly, however, they are becoming regular features of the ongoing regulatory process. While initial stress-testing exercises mainly focused on capital adequacy, looking in particular at capital ratios, more recent tests have placed greater emphasis on qualitative and quantitative modules to evaluate broader risks within banks. Stress testing has now become one of the regulators’ main tools for regularly assessing the health of supervised banks. Banks operating internationally (for example, non-US-owned banks with major US operations) will face multiple and more comprehensive stress-testing requirements to satisfy both home country and host country requirements.
Stress tests are cost-intensive endeavors. They demand a large number of dedicated employees and the active involvement across virtually the entire organization—risk, finance, treasury, business units, audit, senior management, and the board of directors. Additionally, a failure to comply with a regulator’s requirements could result in severe penalties, for example, a mandatory recapitalization, forced sale of operating units and portfolios, and management changes. It has therefore become a key priority for banks to develop an effective and cost-efficient approach to stress testing.
For the original round of tests, there was a necessary focus on data collection and mechanisms to allow an unfiltered transfer of information to relevant evaluators. However, the latest rounds of testing focus much more significantly on the quality of banks’ internal risk management and capital planning processes—it has become increasingly important that banks are able to demonstrate to supervisors the quality and robustness of their internal capabilities in capital planning and stress-testing analysis. Banks can benefit from taking a more proactive approach, for instance, by developing extensive commentary on their analyses and frequently communicating with regulators. Like other regulatory requirements, a stress test is essentially a process, and it can be made more efficient with regard to time and costs, particularly if banks realize the need for a clearly articulated and internally aligned story line. This working paper sets out a seven-module process, starting with thorough data preparation and extending to a clear communication strategy and comprehensive modeling of macroscenarios (exhibit).
Module 1: Data preparation
The starting point of each stress-testing exercise is the preparation of data tapes. For this step, it is essential to provide the evaluator with a comprehensive and easy-to-navigate data set augmented by an authoritative commentary that seeks to explain and interpret the data. Also, it is vital to ensure integrity between reporting and accounting data sets. For these purposes, banks can benefit from the creation of a dedicated data team responsible for data aggregation as well as for responding in a timely fashion, often under severe pressure, to evaluators’ requests.
Module 2: Strategic project-management office (PMO)
In some banks, stress testing is seen more as an operational than a strategic task, and consequently it receives relatively little—or even no—top-management attention and buy-in. This can be overcome if a bank establishes a strategic PMO. The PMO serves as single point of contact for the evaluator. It ensures the consistency and alignment of all data submissions and communication. In addition, it is responsible for the active steering of the entire stress-testing process and can significantly raise the likelihood of effective internal governance. Ideally, the PMO consists of top managers from every relevant business function and is headed by a leader one level below the CEO in the organization. Top-management involvement serves to highlight the importance of the exercise to both internal and external stakeholders.
Module 3: Robust methodologies
Data models are at the core of each stress test. The creation of viable scenarios and the correct assessment of their implications for the bank are essential. Often, the applied models neglect second-order effects, external shocks, and other significant risk factors. In addition, many models are not robust with respect to the scenarios’ effect on business volumes, revenues, and costs. Banks therefore need to initiate a forward-looking projection of financial statements and their underlying drivers. They should also develop comprehensive stress-testing manuals, including guidelines, standard processes, and best practices.
Module 4: Superior portfolio analytics
Regulators are increasingly focusing their evaluation on selected asset classes and portfolio types. In order to forecast the behavior of specific portfolios under stress, banks need to establish granular bottom-up stress-testing models on a portfolio level, incorporating all risk types (for instance, credit, market, operational, and liquidity risk). Only then can they assess the impact of various scenarios on their financial statements and ratios. To make this convincing and compelling to the evaluators, banks need a deep understanding of their applied methodologies so that their explanatory materials meet the necessary standard.
Module 5: Persuasive communication
Both internal and external communications are essential for an efficient stress-testing exercise. Banks therefore need a clear communication strategy. An overarching story line not only helps to set the frame for future analyses internally but also provides context for the regulator’s evaluation. Follow-up communication with regulators should be precise, consistent, and fact based, which can help avoid uncertainties as well as misunderstandings. It is vital that the bank representatives who present findings to regulators are as well prepared as possible because an overly conservative assessment of the bank would have serious disadvantages.
Module 6: Macroscenarios and correlations
Banks often apply a “black hole” approach to developing stress-testing models. Their econometric models fail to link macroeconomic factors and core banking drivers and usually are not based on historical data. In many cases, banks also fail to provide any explanation as to how core drivers of their performance might be counterintuitively sensitive to macroscenarios. This could raise concerns on the part of the evaluator and lead to a conservative assessment. It is therefore advisable to build strong in-house capabilities for scenario development and analysis. Additionally, banks should place high emphasis on developing sophisticated stress-testing models that consider a wide range of macrotrends.
Module 7: Stress-testing model
It is important that stress tests are not seen as simply necessary and costly elements of a bank’s compliance with regulatory requirements. Provided they are of sufficient quality, the data and analyses required to fulfill a test can serve as a strong base for strategic decisions. Banks should create central risk-aggregator models designed to combine individual portfolio results from business lines and incorporate forward-planning assumptions. Banks also need to develop capabilities for reverse stress testing, allowing them to assess the impact of potential strategic actions on key ratios and financial statements. Finally, they need to ensure that stress tests are followed by concrete action plans with full account taken of governance, processes, and responsibilities.
Stress testing can be a painful and resource-consuming process for banks. Because the potential penalties from a poor assessment are so severe, there is little room for inefficiencies or, worse, actual mistakes. Given the recent announcement of multiple stress-testing initiatives, we can be confident that this will remain the case for most banks that operate internationally. Due to tight deadlines and increasingly detailed regulatory requests, the bare submission of unexplained data is no longer feasible. An efficient approach to stress testing must involve a comprehensive and annotated data tape, outstanding modeling capabilities, stringent process management, and clear-cut internal and external communications. The seven-step process introduced in this paper suggests that the real aspiration for senior managers should be to use a mandatory and apparently long-lasting regulatory requirement to trigger improvements in how they make strategic decisions.