# Assess the sensitivity of Loan-Service-Income (LSTI) ratios to increases in interest rates

Prepared by Emil Bandoni, Barbara Jarmulska and Jan Hannes Lang

In a context of rising interest rates in the Eurozone, it is important to understand how household LSTI ratios could be affected. To answer this question, it is useful to remember that at a given moment the LSTI ratio consists of two parts of payment: the amortization of the principal of the loan and the payment of interest on the outstanding balance of the loan:

$LSJIJ=LJIJδJ+IJ$

.

Factors that influence the interest rate sensitivity of LSTI include the initial loan-to-income ratio (LTI), loan term, interest rate fixation period, and initial interest rate. The definition of the LSTI ratio above suggests that the higher the LTI ratio is at the time of the interest rate reset, the higher the sensitivity of the LSTI will be to a given rise in interest rates.[1] In turn, the LTI ratio at the time of the interest rate reset will be higher: (1) the higher the LTI ratio at the origin; (2) the lower the amortization rate, which will depend on the maturity of the loan;[2] and (3) less time has elapsed since the loan was granted, which is determined by the interest rate fixation period.[3] These three factors give rise to an increase in the outstanding amount of loans in the event of a revaluation and therefore translate into a larger increase in the LSTI for a given change in the interest rate (Chart A, panel a). Finally, a lower interest rate at the origin will lead to greater sensitivity to the LSTI interest rate for a given market rate, because the change in the interest rate at the time of refixing will be greater.

Given the heterogeneous impact of interest rate changes on LSTI ratios depending on the underlying lending standards, loan-level simulations are essential to detect pockets of vulnerabilities. Simulations based on aggregate data will not be meaningful in identifying vulnerabilities, as the aggregate household debt-to-income ratio (~1 in the Eurozone) is several times lower than LTI ratios at the micro level (with values ​​typically between 2 and 8). Although full loan-level data for households is not available for the euro area, data from the European DataWarehouse (EDW) on the securitized household mortgage universe can be used to get a view at least partial pockets of vulnerability. With this data, the framework described above allows us to simulate shocked LSTIs based on mortgage loan characteristics, including LTI ratio and interest rate at origination, maturity, mortgage interest and type of amortization. Chart A, part b shows a simulated distribution of LSTI increases if interest rates increase by 200 basis points from end-2021 levels. Based on this exercise, we find that LSTI increases would be manageable for most loans, but pockets of vulnerability exist where existing loan LSTIs could increase by more than 8 percentage points. The granularity of the data also allows analysis of expected changes in LSTIs between countries and over time, or to focus on a subset of exposures.

#### Table A

The initial LTI ratio, the maturity of the loan and the period of interest rate fixation are key factors that influence the interest rate sensitivity of the LSTI ratio