Variable business – proven model, new implementation

Variable business – mix ratios in the low interest rate phase

The first fundamental departure from the concept of moving averages occurred during the low-interest rate phase.

Previously, mix ratios were determined using historical time series analyses and/or expert estimates as part of forward-looking analyses.

Although the consideration of volume fluctuations was technically sound, it was rarely applied in practice.

In the low interest rate environment, significant volume growth was frequently observed in demand deposit accounts. The reason for the growing volume of demand deposits was the lack of alternative products offering interest.

This brought the consideration of risk from volume fluctuations into focus: “What happens if interest rates rise again and our customers withdraw their funds from demand deposits?”

In practice, assumptions were made about how the total volume of demand deposits might have developed under normal circumstances. Based on the balance sheet before the low interest rate phase, growth assumptions were made and, based on these assumptions, the balance sheet was divided into “real demand deposits” and “potentially at risk of outflows.”

This approach was technically sound, rule-based, and withstood regulatory scrutiny.

Questioning previous assumptions in a positive interest rate environment

Acceptance within banks and by supervisory authorities, as well as the easily comprehensible set of rules for implementation, tend to encourage banks to continue to proceed in accordance with existing process descriptions and internal checklists.

However, it is important to recognize the reasons behind the approach and whether these circumstances still apply.

The original question, “What happens when interest rates rise again?”, can now be answered.

Market interest rates have risen sharply since mid-2022 and have been clearly positive ever since. In addition to shifts within banks, there have also been outflows to other institutions.

This does not make it unnecessary to consider risks from volume fluctuations. However, in order to continue to accurately reflect interest rate risks, the procedural model from the low interest rate phase must be fundamentally changed.

Adjustments to the current process model

The division of variable products into a long-term base amount and a shorter-term portion that is subject to volume fluctuations continues to make technical sense for many products. However, the approach to determining this base amount should be adapted: Away from a consideration of the volume before the low-interest phase with the application of a set of formulas to extrapolate a possible development, towards observable data. The focus should be on the following questions:

• How strong were the volume outflows due to market interest rate changes in the individual positions?
• What proportion can still be considered at risk of outflow in the short and medium term?
• Do these figures match the internal sales and overall corporate planning under the interest rate forecasts relevant to management?

In addition, the interest rate mix itself was often not dealt with in detail during the low-interest phase. This is also due to the atypical data situation during this period.

Historical analyses lost a great deal of their informative value during the low-interest phase due to the floor in the customer interest rate and atypical sliding interest rates.

However, a future analysis – as an alternative or supplement to a historical analysis – was equally difficult. As no products with significant interest rates were offered during this period, it was not possible to derive information about any changes in customer behavior. Consequently, this lack of current data often led to a confirmation of the existing mix.

Time series analysis vs. expert estimate

The period after 2022 is not yet long enough for statistical time series analyses to be carried out reliably. However, a mixture of an analysis of the existing data and an expert estimate in the form of a future analysis can lead to better and in some cases significantly changed mixing ratios.

The procedure for determining the mixing ratios must also be adapted.

The phase of low interest rates represents a fundamental break in the product time series, meaning that an analysis based on long time series does not make technical sense. Shorter time series can and should be used to enable a technically correct illustration. Here, seasonal fluctuations and medium-term planning should again be placed in the foreground. Additional future analyses and expert estimates supplement this.

Liquidity mix ratios

Liquidity mix ratios are another technical aspect that is attracting increasing attention. Here, too, the situation was such that it was difficult to derive technically accurate results due to a lack of reallocations and constantly growing portfolios.

Similar to the considerations regarding base amounts, however, the liquidity assumptions can now also be reviewed in this respect:

• How long will funds remain in a particular variable product?
• What proportion of the total volume is permanently available and what proportion is at risk of outflow?

Impact on key risk figures and sales controlling

The described adjustments to the procedure in line with the changed market situation are necessary in order to adequately reflect the risks.

However, the direction in which calculated risk indicators move as a result of the adjustments should not be relevant here.

A clear trend has been identified in various consulting projects: In total across all of a bank’s products, the duration of passive interest rate mix ratios has increased. Although these were often made shorter in the interest rate adjustment context in order to strengthen customer loyalty, this was clearly overshadowed by the effects of the base adjustment.

The high buffer share in a short term could lead to strongly fluctuating margins in sales controlling. This effect is reduced by the revision of the base amounts.

In summary, it can therefore be stated that the rise in interest rates and the resulting need to adjust the mix ratios leads to a reduction in assumptions and thus to an improvement in the quality of risk measurement, also stabilizes the sales ratios and often even relieves the institutions’ risk utilization.

Requirements for documentation and decision templates

A model change with a positive impact on risk indicators should always be very well justified, as such changes must be able to withstand critical questions from internal and external auditors.

The documentation effort therefore does not only consist of adapting the existing procedure – from internal checklists and process descriptions to documentation and decision templates – at individual points.

Rather, the analysis procedure changes so significantly compared to the low-interest phase that the documentation must also be adapted in large parts. However, as the above explanations have shown, a clear argumentation is possible. During the low-interest phase, assumptions had to be made in order to reflect the risks as well as possible given the current level of information. After the end of the low-interest phase, these assumptions must be questioned and a break in the procedure is unavoidable.

In practice, when managing sales via margins, it makes sense to implement this break by changing the mix ratios at the turn of the year.