By Darnell Canada, managing director, Darling Consulting Group
“The future ain’t what it used to be.”
There was a time, not so long ago, when the market pundits could never have imagined the economy receding to the state it is in today. In hindsight, we all should have taken more seriously the words of a famous baseball player with only an eighth-grade education when he made that well-known Yogiism.
In an effort to maintain margin and reach earnings goals, community banks struggle with the notion of adding longer-duration loans to their portfolios. While this article will not advocate or denounce the merits of asset extensions, I will put into context the manner in which you can develop confidence in that strategy. I will also try to ensure that your worst fears will not come true, if Murphy’s Law plays out and rates begin to move sharply higher in the near future.
Playing long game isn’t new
Heated debates centered on the idea of holding long-duration assets date back to the mid-1990s.
Back then, the economy had only recently recovered from an environment that was—at the time—labeled “the worst mortgage crisis in history.” Thirty-year mortgage rates were declining toward a then-historical low of 7%.
Many held the view that these rates could go no lower for any sustained period. This speculative concern sparked widespread resistance to long-duration asset classes, in particular fixed-rate mortgages, as well as an explosion in the size of the mortgage investment marketplace.
This, in turn, led to the creation of new derivative products, followed by … well, you know how the story played out.
Twenty years later, market rates are significantly lower. Economists continue to struggle to find evidence that warrants any meaningful northward movement in Federal Reserve monetary policy.
Where community banks are now
This history proves of particular interest today because our observations of interest rate risk models suggest the average community banking institution is asset sensitive, exhibiting worst-case exposure to lower rates (or a flatter yield curve). This type of rate-risk position typically demands that banks hold fixed-rate assets and search for call protection.
For years, historically low rate levels have remained the primary deterrent to executing such a strategy. Meanwhile, softness in both the domestic and global economies has prevented the Fed from hiking short-term rates and has influenced long-term (i.e. 10-year) Treasury rates closer toward 1%.
Further, long-term rates in three of the other six largest economies in the world are currently negative!*
And still, most bankers are convinced that market rates in the U.S. have hit their floor and cannot decrease further. Accordingly, there remains a common view that medium- and long-duration assets remain a “toxic” product not to be held on balance sheets.
It is clear why bankers are concerned about building (or even maintaining) a portfolio of long-term assets at these historically low rate levels.
However, a key question remains: “How concerned should they be?”
One size doesn’t fit all
The answer varies for each institution. It will depend on how each bank’s interest rate risk position is postured. And it will hinge on the specific individual drivers affecting each bank’s exposure.
The last few decades have proved that markets and economies can travel meaningfully in directions that contradict and/or far exceed our expectations and imagination. More importantly, balance sheet strategies should be based primarily upon risk posture, not taking speculative viewpoints on rates.
Today, strategies that defy common sense may actually be appropriate and suitable.
It is in this context that bankers should embrace, rather than contest, regulatory guidance that forces them to stress the risk characteristics of their balance sheets. Stress analysis is the only way that ALCO can develop an understanding in the appropriateness of a long asset strategy and fully appreciate the conditions under which that strategy could become problematic to overall financial performance and balance sheet health.
Interest rate risk stress testing
Commonly banks construct a balance sheet model that projects margin income trends under a flat/current rate scenario as well as +/-200 basis point parallel rate ramp scenarios (typically over 12 months). Based upon the historical tendencies of the Federal Open Market Committee, this would seem to be a reasonable set of benchmark parameters for gauging interest rate risk.
Of course, many banks are currently modeling -100 basis points given the low-level term structure of today’s yield curve. Beyond this benchmark model, interagency guidance stipulates that banking institutions must also understand the impact of:
1. Instantaneous and significant changes in the level of interest rates (i.e. rate shocks).
2. Substantial changes in rates over time.
3. Changes in the slope and the shape of the yield curve.
4. Changing relationships between key market rates (i.e. basis risk).
5. Variations in balance sheet behavioral assumptions.
If this process is not administered properly, examiners can give your bank nightmares.
It is not unusual to hear that a set of exam MRAs includes expectations for additional scenarios that cover what seem like the full universe of rate paths and/or sensitivity stresses. This is usually a signal that your examiner is unsatisfied with the manner in which the bank represented its understanding of the business issues related to rate risk.
This could be because you are not, in fact, running stress tests that add meaningful insight to your analysis, or because your bank has not adequately articulated the value that your stress test scenarios present. Accordingly, they will default to having you run more scenarios to ensure you capture any and every potential worst case circumstance.
At this point, it is important to remember that the objective behind stress testing is not to capture as many rate paths as possible in order to identify “where the bank breaks.” Instead, the point is to tell a fully comprehensive story about your risk posture, something that the benchmark simulation analysis cannot offer.
The goal is to identify which stress scenarios offer the greatest value in helping you learn something about your rate-risk posture that you didn’t already know. In turn, this will help your bank understand the level of urgency you should have in making decisions about hedging strategies.
Putting things into plain English
Now, let’s take a closer look at federal stress testing requirements.
Below is our translation of the language in the interagency guidance for stress testing requirements. This will help you think through the business issues related to rate-risk stress testing in the current economy and rate environment.
1. What is the impact on IRR if rates move “more quickly” than I assume in my benchmark model scenarios?
Shock scenarios, as extreme and improbable as they are, help banks gauge the impact of a faster rate path in the event the Fed decides to tighten monetary policy. Margin levels could change more sharply with aggressive Fed action.
Inversely, our firm also recommends that banks model scenarios that depict the impact of a more sluggish shift in rates, such as +200bps over 24 months. Our observations suggest that, had this scenario been examined by more banks, fewer would have been inclined to put on aggressive rising rate hedges in 2012-2014 that today place undue pressure on income.
Given actual Federal Open Market Committee rhetoric and global economic circumstances, a “delayed rising rate scenario” remains very appropriate within the context of contemporary stress test analysis.
2. What is the impact on rate risk if rates move “more substantially” than I capture in my benchmark model scenarios?
When dealing with financial modeling we need to define in quantitative terms what “substantial” means. In this case, we might define this term as “greater than 200 basis points”—say +400-500 basis points.
Only a few years ago the market expected a Fed funds rate as high as 3% by now. The Fed has surprised everyone on the low end. Every cycle changes so, eventually, there may be a time when the Fed surprises to the upside.
In this regard, we might view these extreme rising rate scenarios as a measuring stick to ensure the balance sheet doesn’t exhibit too much duration or convexity risk in the asset base. If we conclude that too much duration and/or convexity does exist, ALCO should closely monitor yield curve slope and other signs that the Fed has taken on a more hawkish sentiment.
3. Does rate risk increase or decrease meaningfully if long-term rates shift differently than short-term rates?
A steeper yield curve is typically helpful to bank margins. A flatter curve is typically harmful.
This is important to remember, given the current flat nature of the current yield curve. One could reasonably argue that the yield curve will steepen in the early stages of the next rising rate cycle. If this is the case, benchmark models that assume a parallel yield curve shift may be overstating rising rate exposure for liability-sensitive balance sheets and exaggerating margin upside for traditionally asset sensitive balance sheets.
4. How does our understanding of rate risk change if the relationship between loan and deposit pricing behaviors differ from what is assumed in the benchmark model?
Most benchmark models assume that loan pricing moves in perfect alignment with market rates. In reality, this is not true.
In general, credit spreads widen when the economy is suffering and market rates fall. As the economy recovers and begins to expand with rising rates, credit spreads tend to tighten. This suggests that the degree to which loan rates rise and fall varies from market rate movements.
This illustrates one aspect of “basis risk.” If coupled with a flattening non-parallel yield curve shift, this basis risk can have a meaningful impact on exposure to rising rates and, therefore, warrant caution with regard to fixed-rate asset strategy.
Deposit pricing assumptions may be the single most influential variable affecting your model results. They are also likely the most uncertain of all the variables within the model. It should be commonplace to periodically test the degree to which different deposit pricing betas can impact your conclusions about interest rate risk.
5. Does my level of concern for rate risk change if actual cash flow variability (i.e. option risk) differs materially from what I assume in my benchmark model scenarios?
The uncertainty of borrower and depositor optionality presents another significant potential risk to model accuracy. This is especially true as it relates to depositor behaviors, where the impact will be most material.
Even if these assumptions are formulated with the help of a core deposit study, it is a good idea to test and quantify the impact of potential balance migration or attrition on your model results. The level of stability in your core deposit base plays a critical role in your capacity to hold long duration assets.
Most banks estimate loan and investment options (prepayments and calls) using a third-party model, which may not represent your local lending market and/or borrowers or the true risk characteristics of individual bonds.
In this regard, it is hard to judge the level of potential error in the model results. Therefore, it is a good idea to identify the degree to which your snapshot on rate risk changes if prepayments (or bond calls) are materially slower or faster than assumed in your benchmark model.
Applying this to lending
Of course, model results will change for each stress test. But to what magnitude?
If the magnitude of change does not change the direction of exposure(s) or shift the materiality of exposure to a level of concern, then you can be confident in your assessment of long-term lending capacities.
If this isn’t the case, ALCO should investigate further; decide whether to change the assumptions in the benchmark model; and/or determine whether the level of rate risk warrants a greater caution.
If your bank is able to speak to these business issues, you are sure to have developed a full comprehensive perspective on the rate risk posture of the balance sheet.
This, in turn, will ensure that you’re in the best possible position to defend the practicality of your modeling process and the utility of your risk analysis.
Most importantly, you won’t “make too many wrong mistakes” with your balance sheet strategy, as Yogi claimed the Yankees did during one of their baseball games.
* For more about life under negative rates, see my colleague Jeff Reynolds’ “Real risks of negative rates”
About the author
Darnell Canada is a Managing Director for Darling Consulting Group, a solutions firm that specializes in the area of asset/liability management for financial institutions.
Canada works directly with C-suite executives helping them to understand the complexities of their balance sheet financial risks and providing guidance and unbiased advice on strategies that strengthen earnings performance. A former FDIC field examiner, he is a speaker and active educator, who enjoys participating in a wide range of educational programs for banks, including the ABA Stonier School of Banking and the Graduate School of Banking at Louisiana State University.
Canada holds a B.S. in finance from Bentley University and an M.S. in finance from the Carroll School of Management at Boston College.