The loss optimization of loan recovery decision times using forecast cashflows

A theoretical method is empirically illustrated in finding the best time to forsake a loan such that the overall credit loss is minimized. This is predicated by forecasting the future cashflows of a loan portfolio up to the contractual term, as a remedy to the inherent right-censoring of real-world incomplete portfolios. Two techniques, a simple probabilistic model and an eight-state Markov chain, are used to forecast these cashflows independently. We train both techniques on different segments of residential mortgage data, provided by a large South African bank, as part of a comparative experimental framework. As a result, the recovery decision’s implied timing is empirically illustrated as a multiperiod optimization problem across uncertain cashflows and competing costs. Using a delinquency measure as a central criterion, our procedure helps to find a loss-optimal threshold at which loan recovery should ideally occur for a given portfolio. Furthermore, both the portfolio’s historical risk profile and the forecasting thereof are shown to influence the timing of the recovery decision. This work can therefore facilitate the revision of relevant bank policies or strategies toward optimizing the loan collections process, especially that of secured lending.