What it means
A stress test is a forward-looking simulation, not a forecast. The regulator specifies a hypothetical path for unemployment, growth, share prices and property values over roughly two to three years, and each bank projects its losses, revenues and capital along that path.
The headline output is a capital ratio under stress. If a bank starts with a common equity tier 1 ratio of 12.5% and ends the scenario at 8%, the test has consumed 4.5 percentage points of capital, and the question is whether what remains still clears the regulatory minimum.
The consequences are real and immediate. In several jurisdictions the stress result sets a capital buffer that directly constrains dividends and share buybacks, so a poor showing means shareholders receive less that year.
Banks therefore manage their portfolios with the scenario in mind, sometimes shrinking exposures that score badly. There is a business-customer angle too.
Loan books that perform badly under stress, such as commercial property, leveraged lending and unsecured consumer credit, become more expensive for the bank to hold, and pricing or appetite in those sectors tightens accordingly. The main criticism is that everyone models the same scenario.
Common assumptions can push banks towards similar portfolios and similar blind spots, and a shock the scenario never imagined remains just as dangerous as before.
In practice
Real-world examples.
Example
A large bank publishes a stressed ratio of 9.2% against a 4.5% minimum and announces a $3,000,000,000 buyback the same week. A rival lands at 5.1%, cancels its buyback and holds its dividend flat for the year.
Example
A mid-sized lender runs an internal stress test showing that a 30% fall in office values would cost it $210,000,000, more than a year's profit. It cuts new office lending, and a developer client finds its refinancing offer reduced from $40,000,000 to $28,000,000.
Example
A corporate treasurer reading published stress results notices that her main bank's projected credit losses are concentrated in the exact sector her company operates in. She adds a second banking relationship before her facility comes up for renewal.
Formula
Calculation
Stressed capital ratio = (starting capital + stressed pre-provision net revenue - stressed losses - distributions) / stressed risk-weighted assets
Take a bank starting with $12,000,000,000 of common equity tier 1 capital against $96,000,000,000 of risk-weighted assets, a starting ratio of $12,000,000,000 / $96,000,000,000 = 12.5%.
Under the severe scenario it projects $4,500,000,000 of pre-provision net revenue over nine quarters, $7,200,000,000 of credit losses, $800,000,000 of trading and other losses, and $500,000,000 of dividends it cannot cancel. Capital at the end is $12,000,000,000 + $4,500,000,000 - $7,200,000,000 - $800,000,000 - $500,000,000 = $8,000,000,000.
Risk-weighted assets grow to $100,000,000,000 as credit quality deteriorates, so the stressed ratio is $8,000,000,000 / $100,000,000,000 = 8.0%. Against a 4.5% minimum the bank passes with 3.5 percentage points to spare, having burned 12.5% - 8.0% = 4.5 percentage points of capital along the way.Case study
Seen in the real world.
Cordell Union Bank is a fictional mid-sized lender used here to illustrate the mechanics of a stress test. Its internal exercise showed capital falling from 13.0% to 7.4% under a severe scenario, comfortably above the 4.5% regulatory minimum but below the 8.0% floor the board had set for itself.
The detail mattered more than the headline. Two-thirds of the projected losses came from a $1,800,000,000 book of loans to leisure and hospitality operators, roughly 14% of the bank's $13,000,000,000 of total lending, and much of it written in a single two-year burst.
Over the following year the illustrative bank sold $600,000,000 of that portfolio, tightened new lending criteria and re-ran the test, which produced a stressed ratio of 8.6%. The exercise changed nothing about the current year's reported profit and almost everything about what management was willing to underwrite.
Watch out
Common mistakes.
- Reading the stressed scenario as a forecast. It is a deliberately severe hypothetical designed to be uncomfortable, not a prediction of what supervisors expect to happen.
- Judging a bank purely on whether it passed. The margin above the minimum, and where the losses came from, tell you far more than a pass or fail label.
- Assuming stress tests only concern very large banks. Smaller lenders run scaled internal versions, and the results quietly shape their appetite for particular sectors.
Questions
People also ask.
How severe are the scenarios?
Typically far worse than an average recession, with unemployment roughly doubling and sharp falls in commercial property and equity values over two to three years.
Does passing mean a bank cannot fail?
No, it means the bank survives that specific scenario on those assumptions; a different shock, or a run on funding, sits outside what the test measures.
Can a business see the results?
For the largest banks yes, supervisors publish summary results and the banks disclose their own figures, though internal tests at smaller institutions stay private.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%