UNCOVERING RISK ON THE ROAD

Our consulting group spends a lot of time on the road, traveling from bank to bank and working with customers to improve risk management practices. As with many of the roads they drive, some are straight, some are winding and some are downright dangerous. One thing the group has found in their travels is that while bankers are working harder than ever to manage the risk inherent in the balance sheet, many still have room to improve when it comes to the asset-liability management ("ALM") process. In general, banks do a good job of measuring +/- 300bp interest rate shocks, but still run monthly processing in a perfunctory manner. Each month, bankers produce reams of analysis and reports for senior management and board discussion, much of which is worthless. To be effective, modeling practices must incorporate not only basic rate shocks, but also attempt to model reality. Let's face it, the reason so many ALM committees virtually ignore the reports they are provided, is because these reports are so far from becoming reality it is easy to do so. Effective banks understand this and as such also incorporate realistic scenarios. This gives management time to discuss the potential risks, prepare for the eventuality and have time to think about it before reacting. Banks with a firm grasp on risk management stand a better chance of stabilizing earnings streams and protecting franchise value. Let's run through an example to help drive this home. Currently, the market believes the FOMC is probably done raising interest rates. Historically, after a cycle where rates have been rising, the FOMC ends up having to cut them about 6M to 12M after the final rate hike (due to overshooting the mark and tightening too much). If 6/26/06 was the final rate increase, then somewhere between 2Q 2007 and the end of next year; the FOMC is likely to start cutting interest rates. Glancing at the short-end of the forward curve, that is exactly what we see. In 3Q, the market is predicting Fed Funds will drop to 5.00% and stay there through 2008. While many could argue this is so small it is irrelevant, recall that risk is also not linear. By the time the first 25bp cut happens, sentiment shifts and the economy rapidly softens. By modeling this event, bankers are forced to think dynamically. Further, consider that current market conditions are driving many banks to originate and hold longer term fixed rate loans, extend liabilities, use hedging tools, waive prepayment options, embrace yield maintenance and put floors into floating rate loans. Each of these strategies has risks and rewards that effective and realistic modeling will highlight. In this environment, banks should leverage modeling to find out if strategies are making them better, worse or the same. This will help narrow down the best options as one heads on down the road. In short, the market is telling us realistic scenarios to model should include a steepener where short rates drop, a steepener where long rates spike, an inversion where short rates spike, slower prepayments on loans with yield maintenance, more expensive funding costs and faster prepayments (due to heavy competition) on the largest and best credits. With rare exception, like driving between two points, the road is rarely straight. Like auto mishaps, many ALM accidents happen when drivers don't consider what might lie around the next curve.