Site Overlay


We were approached by a large municipality in the Netherlands with a long-term loan portfolio of about EUR 400 mln. Municipalities have a continuous budget cycle looking 4 years ahead, and therefore want to obtain a realistic projection of their future financing costs. As loans are typically refinanced at maturity, there is a potential financial risk in case of adverse interest rate risk movements. Therefore, one would want to know what the interest rate risk would be for refinancing the loan portfolio in a negative scenario and how likely this would be.

It is not always straightforward for a public institution to balance budget, keep costs low while not compromising on financing programs with a social impact. Large buffers mean more money would be put aside at the municipality to prepare for higher interest rate risk at time of refinancing. However, it would also mean that money would stay idle in case of the interest rate risk did not materialise.

What is a proper method for interest rate risk estimation then?

Although multiple models are available to estimate the risks given a certain confidence interval, it makes sense to leverage expertise from the insurance sector. This is a sector that by nature relies on having a proper, balanced calculation of it’s risks. It relies on EU-wide Solvency II regulations that provide concrete guidance on risk management practices for Value at Risk estimations. For this exercise we therefore applied the Solvency II standard formula for interest rate risk as used by the insurance sector.

The calculated interest rate component consists of two underlying risk-factors: 1) a risk-free base rate (with forward expectations into the future) and 2) a credit spread that depends on the rating of the organisation. The interest and credit spread shocks can be derived by applying the Solvency II methodology. These shocks enable us to obtain the loss (or gain) relative to the current interest rate under a given confidence interval. Essentially, we calculate worst, base, and best-case monthly scenarios for the interest rate in the next four years. For one of the loans, we obtained the following estimations using publicly available historical DNB swap rates and the derived credit spread:

interest rate risk - historical interest rates and risk projection in 2021

Once we have created the three scenarios, we are better able to estimate the amount due for each loan based on its refinancing date. As a result, we have an exact estimation of differences in payments due relative to worst- and best-case scenarios. For example, for loan 2 below, based on the Value at Risk (90th percentile) interest rate estimation, the amount of additional interest expected to be paid in 2025 in case of a worst-case scenario was estimated at 91,5k Euros.

Interest rate risk - single loan refinancing costs per scenario and year

Perhaps not that detrimental if you look at one loan but taking the full refinancing portfolio into account provides a different perspective. The total additional amount (relative to a base case scenario) to cover for all loans expected to be refinanced by 2025 in case of a negative interest scenario was close to 876k.

Interest rate risk - total financing cost per scenario and year

Overall impact varies substantially depending on the confidence interval (80, 90, 95 or 99% VaR) we choose to use. Using the 80% VaR means that statistically in 20% of the years we expect to have interest costs higher than the calculated number, while a confidence interval of 99% means that there is a 1% chance that the costs are higher.

Interest rate risk - Value at Risk per percentile

What is the best Value at Risk level for an adverse interest rate scenario?

The confidence level we choose depends a lot on the risk appetite an organisation can or wants to take. In case a scenario materialises that is worse than the selected confidence level, the corresponding buffer will not suffice, and to maintain a balanced budget would inherently lead to successive cost-cuts or halt of other activities. So, one could say that it is a trade-off between having high (‘idle’) buffers with almost no probability of cost cuts or having relatively low buffers in combination with frequently occurring cost cuts.

After implementation of the methodology we obtain a manageable and repeatable process that enables us to project the future financing costs for the loan portfolio and the interest rate risk involved. It can become a simple as uploading a file with the current loan portfolio.

To illustrate how the method works we have created a tool that helps to better understand what is calculated:

We provide various interest rate risk reports to our clients that include the risk of the long-term and short-term loan portfolio and the pension provisions. In the case above we did so for a municipality, however the same process could of course be applied to any other organisation.

Data analytics and automation will help you to get proper and timely insight in the probability and size of loan portfolio risk. This will give you more control and makes it a lot easier to communicate with all stakeholders about adverse events.

If you want to receive more information about our risk reporting services please feel free to contact us. We would be happy to schedule a video-call to show you more about these risk reports, what they contain and how they can help you to get more control over your financial risks.