Posts

/by

Interest Rate Risk Management: Timing of Earnings Matters

,

Would it surprise you that some interest rate risk mitigation strategies actually add risk for a period of time?  In the end, these strategies may be effective at mitigating interest rate risk but it may take years for the effectiveness to be felt.  The challenge is that many methods used to test mitigation options, like NEV, don’t show decision-makers the full picture.  The reality is that earnings and the timing of earnings matter when understanding the impact of different decisions.

Take for example, a credit union that decides to sell $90.0 million of its 30-year fixed rate mortgages earning 3.61% and reinvest in auto participations earning 1.90% to mitigate interest rate risk.  No surprise that such a move hurts earnings today.  In fact, it would hurt earnings by 23 bps.

What is not obvious, though, is that in a +300 bp interest rate environment they would give up more cumulative revenue with a strategy to hold auto participations than with a strategy to hold fixed-rate mortgages over a four-year time horizon.  Year-by-year and cumulative revenues for both strategies are summarized below:

The credit union would earn more revenue holding fixed-rate mortgages than holding auto participations for each of the next three years.  It is not until Year 4 that the strategy change pays off and they earn more holding auto participations.  Over the four-year period, the credit union would earn a cumulative $15.0 million with mortgages versus $13.5 million with auto participations in the +300 bp increase.  Timing of earnings matters.

Using only NEV to evaluate this strategy might lead the credit union to pull the trigger on it.  In this case, the credit union is using NCUA’s NMS values from the NEV Supervisory Test.  When comparing the proposed strategy to the base, the NEV in the current rate environment is unchanged at 10.75% – meaning the restructure neither helps nor hurts the current NEV.

However, after selling the mortgages, the volatility decreased in the +300 bp environment leaving a higher NEV ratio (see example above).  An added benefit is that the credit union would now be considered low risk under NCUA’s NEV Supervisory Test thresholds.  Said differently, the NEV results show decision-makers would not have to sacrifice a thing in the current rate environment while significantly reducing risk in a +300 bp interest rate change.

Earnings (see Beginning ROA), on the other hand, show there is sacrifice in the current rate environment and more risk to earnings and net worth in a +300 bp rate environment over a four-year time frame.

The moral of this story is not that all credit unions should always hold mortgages because they will earn more than other alternatives.  Certainly, not.  It should be clear, though, that NEV provides credit union decision-makers with a limited picture.  Earnings and the timing of earnings matter.  In this particular example, the credit union may decide to execute the strategy, but decision-makers understand, in advance, that even if interest rates rise quickly, it could take about four years before the alternative strategy will contribute positive earnings and start to reduce interest rate risk.

For more in-depth ways to use ALM as actionable business intelligence, please click here for our c. notes.

/by

Observations from ALM Model Validations: NEV – Loans Devalue in Rate Shocks – or Do They?

When considering valuation as a measure of interest rate risk, and value volatility as an indicator of changes in interest rate risk, many institutions perform net economic value (NEV) analysis. When working with credit unions, or performing model validations, a concern many have is ensuring the models have the “right” assumptions. What is the “right” discount rate? Should credit risk spreads be incorporated? What effective discount rate or what yield curve should be used to discount cash flows – which method is “more right”?

All of the above may be questions to consider but they are distractions from simple analyses credit union management teams can perform when determining if answers are reasonable. For example, take an auto loan portfolio in which the valuation methodology derives a value of $210M in the base rate environment. This same portfolio devalues to $200M in a +300 bp shock. Said differently, the value volatility in a +300 bp shock is -5.00%. From a quick reasonableness test, this is within a 4-6% devaluation range in a +300 bp shock – very reasonable for an auto loan portfolio.

However, does it change the reasonableness answer if the current book value of the auto loans is $199M? While the devaluation of the loan portfolio is certainly reasonable, the resulting answer implies that the loan portfolio could be sold at a 0.50% gain if rates increased 300 bps instantly. That answer is certainly less reasonable. It is important to remember, in Chapter 13 of NCUA’s Examiner’s Guide, NEV is defined as the fair value of assets less the fair value of liabilities. Would it be reasonable to assume a fair value gain on an auto loan portfolio if rates increased 300 bps?

When measuring NEV volatility, the starting value still matters. High starting values can be driven by low starting discount rates. It is good to evaluate both the effective discount rate and the difference between value and book in the current environment. Some models are unable to calculate an effective discount rate. We have found that sometimes in this situation the effective discount rate does not match what the user intended. If you are in the situation of the model not being able to show the current discount rate, extra attention should be given to the value versus book and how the value compares to book in different environments. Optimistically high starting and shocked values can hide risk and volatility; this connects with the cautions brought out in our blog regarding high starting NEV ratios posted on September 25, 2015.

/by

Observations From ALM Model Validations: High Starting NEV Ratios

When performing model validations, it is common to see a net economic value (NEV) ratio that is considerably higher than the credit union’s current net worth ratio. Understanding NEV and net worth are two completely different concepts; there are reasons why starting with a high NEV ratio in the base environment may not be reasonable.

First, let’s discuss some of the reasons why this can occur:

  1. Non-maturity deposits are assumed to have long average lives. Given the positive slope of the yield curve, this assumption results in higher discount rates and optimistic market value premiums
  2. The credit union does not incorporate transaction spread costs when valuing deposits, which overstates the value to the credit union
  3. Optimistic loan discount rates, that ignore credit and other market risks, results in overly optimistic loan market values

Consider that NEV is intended to show the fair value of a credit union. Therefore, mergers can be used as a reasonableness check of this critical component of modeling. Mergers over the last several years do not support the assertion that an acquiring institution would pay a significant premium to the net worth.

It is important to understand that optimistic base NEV results also impact volatility ratios in various rate shocks. To keep the math simple, consider a $100 credit union performing NEV with two different sets of assumptions. For example purposes, the dollar volatility in a +300 basis point (bp) shock is assumed to be the same while, in reality, the more optimistic assumptions in Assumption B would result in a lower dollar volatility.

Model validations with high NEV ratios inaccurately predict volatility

Many have said that the reasonableness of the starting NEV doesn’t matter; it is the volatility that should be the focus. Notice that while the two sets of assumptions in this example have the same dollars of volatility in a +300 bp shock, the percent volatility and NEV ratio in a +300 bp shock are dramatically different. The NEV in Assumption A may be considered high risk, while the NEV in Assumption B may be considered low to moderate risk.

If using NEV, credit unions should focus not only on NEV volatility but should also understand what their base NEV ratio is showing and if it is reasonable. If the starting NEV ratio is considerably higher than the net worth ratio, the credit union needs to understand why. If it is not defendable, credit union management should consider making adjustments to assumptions.

/by

Borrowing Rates Are Up – Other Impacts to NEV

Continuing with our message from last week, if you are running NEV, there are some other changes you will notice as a result of borrowing rates increasing. As pointed out last week, NEV calculations for the current rate environment will result in shares showing increased value as “market rates” (generally represented by borrowing rates) have risen relative to dividend rates. However, in a +300 bp rate change, the percentage decline in value of shares will actually decrease when compared to your last NEV calculation. This is because a 300 bp change represents a smaller proportional change when you are starting out with a higher discount rate. The overall impact to your NEV results will depend on changes in your balance sheet, but there will be less decline in share values to offset declines in asset values when compared to using lower borrowing rates in your last NEV. Understanding this may help you interpret some of the nuances of your NEV results; however, it will not tell you if your credit union is positioned to make money or not.