Many in the global financial community are focusing on stress tests to assess the fiscal health of the global financial system, and some believe such stress testing of banks is a significant step forward in protecting bank shareholders, taxpayers and the global economy.
The stress tests do indicate the relative stability among the banks tested, for the particular economic risk scenarios included. However, the tests misdiagnose the ability, or inability, of a bank's immune system to resist some of its biggest threats. Looking at this from an ERM perspective, the tests have several shortcomings.
The stress tests only include one-at-a-time risk scenarios. This is not how the world works. Many things can go wrong at once, and usually do.
This is especially true for some of the most damaging situations. An industry study said more than 80% of the largest losses in shareholder value during the study's 10 years resulted from two or more risk events occurring simultaneously. Companies can absorb one strong blow, but a combination can knock them out. Doing stress tests without multiple-risk events can miss a potential combination that could knock out a bank.
Bank stress tests have been restricted to a small subset of a bank's key risks. The tests focus exclusively on financial sources of risk. This is typical of where banks usually focus most, if not all, their risk management attention. Bank risk models are largely occupied with market, credit and liquidity risks.
However, every industry study I have seen shows that financial issues are the smallest source of risk, even for financial institutions. The largest sources of risk relate to strategic issues: incorrect strategy, poor execution, competitor risk, regulatory risk, supplier risk, etc. The next most important sources of risk stem from operational issues: human resources, technology, litigation, disasters etc. Strategic and operational risks account for most of a company's volatility.
Why, then, do bankers continue to focus virtually all their risk modeling efforts, including stress testing, on financial sources of risk? One answer is that this comes naturally. Bankers' education, training, skills and experience, even the names of their departments, relate to financial sources of risk. A more comprehensive approach is offered by those certified as chartered enterprise risk analysts, who routinely deal with a range of risk sources.
Another answer is that industry standards, such as the Basel Accords, do not include robust requirements for measuring strategic and operational risks. Finally, banks also seem to lack a practical methodology for robustly quantifying strategic and operational risks.
Banks typically approach stress testing by measuring in terms of potential impact on current capital. The problem with this common approach is that it provides a snapshot of the current or near-term position, as opposed to a longer-term view of damage that can continue, or worsen. This is near-sighted and significantly understates the impact of risk events. This is particularly true for strategic and operational risks, which tend to have a larger impact on future revenues and expenses than on current capital positions. To properly measure the full impact of all key risks, banks should expand their capital-based risk models into value-based models that fully capture the effect of risks on projected free cash flows.
Stress tests are defined by stakeholders external to the bank. This presumes that those doing the defining can actually determine what the "worst-case" scenarios are for all banks. This is not so. Each risk scenario can affect each bank differently. Only bank executives know which risk scenarios would be truly disastrous based on their specific products and markets.
Many banks are undoubtedly relieved that the stress tests omitted risk scenarios that might have revealed far more ruinous outcomes for them. So in addition to the standardized stress tests, it would be very revealing if banks were asked to identify, quantify, certify and disclose what they believe to be their worst-case scenarios.
Predicting a healthier global economic system based on bank stress testing is premature; testing's many shortcomings expose its ineffectiveness. Until these issues are addressed, stress tests will remain a placebo.
