Earnings at risk is only a short-term view of interest rate risk
Originally published 2/12/2007 © 2021 Olson Research Associates, Inc.
A favorite part of my job is seeing the “ah-ha!” moment of understanding about interest rate risk (no fair making comments about how I need to get a life!). Seriously, it’s a great feeling, when you’re dealing with such an obscure and focused topic, to help somebody else see the importance of it all. Don’t get me wrong, I don’t think that someday interest rate risk will be a glamorous subject, or that this niche we’ve carved out to help smaller community bank’s measure interest rate risk will someday make me a millionaire, but teaching (no matter what subject) can really be satisfying.
Over the years the most common “ah-ha!” moment happens when I’m talking with a Board of Directors about interest rate risk. To keep the conversation simple we always start out by talking about earnings-at-risk. Almost every board member I’ve ever met is first and foremost focused on earnings (no big surprise there). Earnings-at-risk is a measurement of how much the bank’s margin could change given a change in interest rates. For example could the bank lose as much as 15 or 20% if rates change? Or does the potential loss appear to be very small, like 2 or 3%?
Many smaller bank’s, for better or worse, still tend to look at their bank’s interest rate risk by referring to a gap report (doh!). Often times their gap report will show a one-year-gap that’s pretty well matched, say somewhere between 90 and 110%. Not surprisingly, after running an income simulation, many of these institutions will also show a low earnings at risk, say below 5%. Measurements like these tend to leave senior management thinking that the bank has little or no interest rate risk.
What they usually don’t consider is that earnings-at-risk (and gap for that matter) are tools that really only measure short-term interest rate risk. Earnings-at-risk simulations usually only project the change in interest income one-year into the future. And the typical gap measurement is the “one-year cumulative gap”. But what about potential risk beyond one-year? When I emphasize the one-year forecast part of earnings at risk, I usually get a few people in the room thinking. But my goal is to make sure everybody understands why looking at earnings at risk just isn’t enough. It’s only a short-term view of interest risk risk.
To reach the “ah-ha!” moment I like to give the board this example. Suppose you were Simple Bank & Trust and you had no earnings at risk at all. Here’s your balance sheet (remember I said simple, so forget about portfolio turnover; assume no sales, and no early withdrawals):
Assets: $100,000 in 10–year U.S Treasuries
Funding: $90,000 in 3–year Small CDs
Equity capital: $10,000
The question is what’s the earnings at risk if rates rise or fall? The simple answer is none! If rates rise or fall there is no impact on the level of income from the bonds for 10 years, there’s certainly no change in income for the first year. The same is true for the CDs. For three years the level of expense is fixed. So we have a fixed yield and a fixed cost for one-year, nothing reprices, therefore no earnings-at-risk. But does this bank have interest rate risk? Yes. After three years the bank will have to refinance its investment in treasury bonds. If rates have risen over that time frame the bank’s margin will narrow; if rates have fallen the new CDs will be less expensive and there will be a favorable change in margin.
Today, more than ever, bank’s are adding to their balance sheets instruments that may not impact this year’s performance: 3–1 or 5–1 ARMs, Callable Bonds, Convertible Debt, etc. It’s important to look for potential risk beyond one year using such measurements as Economic Value of Equity (EVE) and duration.