Optimising Risk and Return of Non-Maturing Products by Dynamic Replication

Florentina Paraschiv and Michael Schürle

The risk management of non-maturing products is an important challenge for most banks, particularly for those with significant retail business. This task is complicated by the inherent options of these products: clients may add or withdraw volumes any time, and the product rate can be adjusted by the bank as a matter of policy. Both properties make future cashflows uncertain. Usually, non-maturing products are replicated by a portfolio of fixed-maturity instruments. We show with real examples that popular approaches based on static investment or funding rules are inefficient. As an alternative, we propose a novel method that we call “dynamic replication”. Here, new decisions are periodically made on the allocation of maturing tranches, corrected by volume changes, that are determined by a multistage stochastic optimisation model. We describe this approach conceptually before illustrating its performance in a case study. Because the method is new and combines concepts that have not been used in this combination so far, technical details are documented in the appendixes.

CHARACTERISTICS OF NON-MATURING PRODUCTS

The balances of most banks typically contain a significant portion of