US banks’ internal stress tests vary

Choice of stress period affects market risk capital requirements

The stress periods used by the largest US banks to determine a portion of their market risk capital requirements vary markedly and change frequently. 

Stressed value-at-risk-based capital requirements are calculated by running a bank’s trading portfolio through a regulatory VAR model with inputs calibrated to historical data from a 12-month period of financial stress. 

Market risk capital rules state that a firm may select a stress period that is is appropriate to the “composition and directional bias of its current portfolio”.The six largest US banks all use data from various stages of the global financial crisis spanning December 1, 2007 to September 1, 2009.

Citi’s stress period begins the earliest of the group, on December 1, 2007. The 12-month period captures the Vix peak of 80.6 reached on November 20, 2008 as well as the US 10-year Treasury yield high point of 4.27% on December 26, 2007. It excludes the nadir of the S&P 500, which sunk to 677 on March 9, 2009 and the US Treasury yield slump, which bottomed out at 2% on December 30, 2008.  

Morgan Stanley’s stress period runs from September 1, 2008, capturing the VIX spike and both the S&P 500 and US Treasury yield collapse, but not the earlier Treasury yield lurch. 

The dealers also change their stress periods with varying degrees of frequency. Goldman Sachs has shifted its stress period the most, by 113 times since 2015, followed by Bank of America, with 61 changes. 

In contrast, Morgan Stanley has changed its stress period six times, Citi four, JP Morgan three and Wells Fargo just twice.   

What is it?

Banks subject to US federal agencies’ market risk rule must disclose their capital requirements in quarterly FFIEC 102 reports.  

Each bank’s market risk capital requirement includes both a VAR and a stressed VAR component: the former based on a VAR model calibrated to a 99% confidence level over a 10-day holding period and a 12-month observation period; and the latter using the same model fed with historical data from a 12-month period of significant financial stress. The stressed VAR-based measure must be calculated at least weekly.

The rules also state that a bank must have policies and procedures that explain why a chosen stress period is appropriate and describe the process for “selecting, reviewing and updating the period of significant financial stress”. 

Why it matters

Changes to the stress period, even by a few days, can have large knock-on effects on a bank’s stressed VAR-based measure. 

For example, when Wells Fargo shifted its stress period in the second quarter, for the first time since end-2015, its stressed VAR capital jumped $303 million (25%), out of kilter with its VAR-based measure, which decreased by $2 million (–1%).

On average, the stressed VAR-based measure made up 27% of the six dealers’ total standardised market risk capital requirements as of end-June.

The stress periods currently used by the major US banks suggest that JP Morgan and Citi’s market risk capital requirements are less sensitive to equity risk than their peers, and that Citi’s is less exposed to rock-bottom interest rates.    

As the choice of stress period and frequency of changes are a function of internal decision-making, it may be the case that Goldman Sachs and Bank of America – the two banks that have switched periods the most – do so in response to shifts in the make-up of their trading portfolios day-to-day or week-to-week. 

Bank of America declined to comment, but Goldman Sachs said its model “dynamically chooses the worst period corresponding to our positions per regulation”, lending credence to the above theory. 

Get in touch 

What's better: a stressed VAR model that switches stress periods dynamically in response to portfolio shifts, or one that changes only at management's discretion? Share your guiding philosophy by emailing or tweeting @LouieWoodall or @RiskQuantum.

Tell me more

VAR switch may explain $500 million capital hike at Wells Fargo

US banks curb market risk

View all bank stories

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact or view our subscription options here:

You are currently unable to copy this content. Please contact to find out more.

You need to sign in to use this feature. If you don’t have a account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here