Service overview

Reserves forecasting

Figure 5. Reserves: forecast vs. actual Confidence interval.

Banks create their own reserves evaluation systems. Their choice of reserving model depends on particular internal procedures and policies. Evaluation methods can vary greatly between banks, so for this short overview let us limit ourselves to one.

Quite commonly banks split up credit portfolios into segments for assessment. Then monthly write-off figures from every segment are combined in a years total. This is called coverage. For consumer credit annual coverage is all transitions to 120 DPD over 12 months divided by total value of loans at the start of the period. Once this is known, moving average coverages are calculated, e.g. for 6 months. Finally current segment distribution of a portfolio is determined and reserves are computed.

Forecasting reserves is much more challenging. A good assessment needs monthly predictions by segment and future annual coverage values. RRAS has a number of functions just for this purpose. It can forecast reserves for various business scenarios, macroeconomic scenarios and reserve models of credit portfolios.



The author presents the methods for studying credit portfolio behavior in the Modeling and Stress-Testing Credit Portfolio Behavior are partly based on so-called dual time dynamics method. This work suggests using dual time dynamics not for decomposing scalar values but for decomposing matrices. The author considers a credit portfolio as a process described by a first-order heterogeneous Markov chain. Starting from this premise, the author uses vintage analysis and the theorem of strong convergence of modified fixed-point algorithms to arrive at transition matrix decomposition. This method makes highly accurate forecasts of credit portfolios possible. Reserves can be estimated with excellent precision and relevant values for stress-testing obtained.

In his later article The Theory and practice of Retail Credit the author considers some successful practical applications of his methods described in the article Credit portfolios behavior modelling and stress-test.