Silicon Valley Bank Tackles a New Kind of Stress

As Marc Verissimo surveys the new stress testing landscape, he's taking stock of his institution's level of protection against whatever economic calamity may come down the pike.

"As a business bank, we have a pretty good loss history … [but] on the operational side, that's more of an art. We've been reaching out to other banks and working with consultants to work through the types of data we need to capture," says Verissimo, the chief risk officer for Silicon Valley Bank (SIVB), a $19.6 billion-asset institution based in Santa Clara, Calif.

Silicon Valley Bank, which specializes in loans to tech companies, venture capital companies and wine businesses globally, is prepping for mandated stress tests under regulations such as Basel III and Dodd Frank. "We've been requantifying all risk to get ahead of the [regulatory curve], and this will feed into how we perform stress tests in the future," Verissimo says. As it is for most banks, stress testing is becoming more stressful for SVB as regulators demand banks dig much deeper to determine the amount of capital to hold against not only credit risk, but also the impact of external factors and operational risks. "The Fed was direct in saying banks should look at quantifying all risks, not just credit risks," Verissimo says.

The Dodd-Frank law requires annual stress tests for bank holding companies, state member banks and savings and loan companies with between $10 billion and $50 billion in assets to determine risk-based capital levels. The deadline has been a moving target, but the Federal Reserve, the Office of the Comptroller of the Currency and the Federal Deposit Insurance Corp. have said the implementation timeframe will likely be the fall of 2013.

SVB has been using a Basel-driven formula to determine capital reserves as a placeholder, but is now expanding that to calculate the broader economic risks that may occur within and across the varied business categories that the bank targets for lending. To help build out the model, SVB purchased a Moody's product called RiskFrontier, which identifies risk concentrations by industry, geography or asset type. It then computes expected and unexpected losses and calculates distribution of portfolio values, losses and capital. It's built on GCorr, or the Global Correlation Model, which uses a global dataset of publically traded corporations, retail credits and commercial real estate loans spanning several business cycles to measure the relationships between the risks in different asset types and factors such as industry and location.

SVB will use RiskFrontier to help calculate default probabilities for different corporate niches and measure how credit will perform over a period of time based on different economic cycles. The challenge is to diversify the corporate loan portfolio to mitigate complex risk exposures within and across industry segments while remaining true to the bank's traditional mission.

"[SVB] has experience in lending in the Silicon Valley area. You want to recognize what the cost of doing business is and prepare yourself accordingly. It's a fine line in terms of economic diversification and not forgetting what the business is actually good at," says Amnon Levy, a managing director at Moody's Analytics, whose clients also include Bank of America (BAC) and HSBC (HBC).

The new stress tests have produced a large industry of new modeling and analytic tools, as companies such as Invictus Consulting Group and Oracle (ORCL) target banks with software that marries traditional credit risk with more complex measurements of the chain reaction of economic performance on a series of companies and a bank's portfolio.

Levy says that for corporate loan originators, analysis of the pricing or vetting of any potential corporate loan can now be viewed within the broader context of the bank's portfolio. He says different sectors respond differently based on type of economic shock, and those differences must be taken into consideration. "That originator can [be required to account for greater risk] if that loan creates exposure to an industry that's risky. That kind of analysis is pretty encouraging, because it recognizes, at the point of origination, the issues that we went through during the crisis," Levy says.

For reprint and licensing requests for this article, click here.
Bank technology
MORE FROM AMERICAN BANKER