Investigators will spend years probing the debacles last year involving Orange County and Barings, when two seemingly solid players in the market collapsed under the weight of unforeseen trading and investment risks.

But the problems that brought down Barings and buffeted Orange County could have been avoided if management had followed two essential practices: performing value-at-risk analysis across the enterprise and separating the people responsible for measuring and analyzing value-at-risk from those charged with running the trading operation.

Value-at-risk, associated mainly with the trading room operation, consists of a complex series of detailed calculations that quantify investment risks over a specified period of time. Advanced approaches to value-at-risk incorporate explicit risk/return trade-offs and feature Monte Carlo simulations, the results of which reveal the magnitude and nature of the risk.

Value-at-risk addresses the problem of interest rate fluctuation, a key wild card in the investment analysis. A shift in interest of just a few basis points can trigger a string of actions that directly affect the current and future value of an investment and an entire portfolio.

But banks increasingly are recognizing that value-at-risk analysis should not be limited to the trading room. Rather, it provides valuable insight when applied across the entire institution. These managers are finding, often to their surprise, that in areas not traditionally associated with value-at-risk analysis, they indeed have significant interest rate risk.

Almost every department, from revolving credit to deposits, has portfolios that are affected by interest rate shifts. In this period of intense competition, when banks must introduce new products quickly to meet market demands or to counter competitive activities, a bank can find itself in a situation where the very success of a marketing campaign for a particular product may be overloading the bank with unacceptably high levels of risk. Unaware of the value at risk, managers fail to take the necessary hedging and risk management measures that prudence requires.

The problem is most acute where there is extensive interest rate optionality risk, everything from adjustable-rate mortgages to credit card teaser promotions. For example, the "double-up" certificate of deposit has emerged as a popular new product in certain areas of the country. With a double-up CD, the investor has the option to increase the investment at the existing rate during the life of the CD. This puts the bank at significant interest rate risk. Should interest rates drop during the life of the program, the bank will likely find itself committed to twice the amount at the higher initial rate. Without value-at-risk analysis, managers may initiate a successful marketing campaign that serves to compound dangers they haven't fully assessed. Before they know it, the bank is holding millions of dollars in paper and scrambling to put together a hedging strategy.

This does not mean that banks should avoid such products, rather that managers must fully assess this risk through enterprisewide value-at-risk analysis as they design and introduce products. To see that this happens, the treasury staff should be involved in product development.

Unfortunately, conventional value-at-risk analysis is a time-consuming, data-intensive process. It can require detailed historical data that are not readily available for banks. While there are historical data for investment and trading-related activities, such as foreign exchange, there are few useful historical data for mortgage or corporate loan markets - areas where banks need it most.

It is problematic even where third-party historic data are available. Based on arbitrary metrics for time and weight, the historical correlations may not be relevant to a bank's situation today. The past is a notoriously poor predictor of the future. Despite the problems, the historical approach using the available data is better than nothing.

The forward-looking approach comes at the problem from a completely different angle. In this approach, you analyze the behavior of the bank by capturing a snapshot of the current model and then calculating how that will change under different interest rate scenarios. By employing submodels (the settings of which may be based on historical information or management's views of the future), analysts can examine such activities as prepayment actions based on different indexes.

Banks can also conduct a Monte Carlo simulation, in which they calculate changes in value for numerous interest rate scenarios rather than a single scenario. The result is a distribution of potential outcomes representing a range of risk levels and probabilities. Such simulations, however, are impractical without the aid of automation.

Armed with the results of the Monte Carlo simulation, bank managers can make informed decisions based on potential of the opportunity, the level of the risk, and the likelihood of the occurrence. For example, adjustable- rate mortgages are among the most complex products in a bank's portfolio. The presence and combination of teaser periods, life caps, life floors, periodic caps, periodic floors, composite rate indexes, negative amortization, and more makes these both attractive to the customer and potentially dangerous to the lender. However, by modeling the various options using the forward-looking approach and the Monte Carlo simulation, managers can tweak the various attributes to create a product that will appeal to the customer but will be more readily hedged and/or managed by the institution.

Banks must also ensure that the people responsible for measuring and monitoring risk are not the same as those who manage the product or investments that are the source of the risk. If those who measure, monitor, and report risk do not have a financial stake in the source of the risk, the temptation and the ability to obscure and manipulate value-at-risk assessment is effectively removed.

Banks need to take their cue from the trading room and apply value-at- risk analysis across the entire institution. Dave LaCross is chief executive of Risk Management Technologies in Berkeley, Calif.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.