Part III: Asset and Liability Management Using LSMC - Allocation Optimisation

In the previous articles of the series, an asset and liability management (ALM) framework using least-squares Monte Carlo (LSMC) was outlined. Moreover, the accuracy and performance of the LSMC method were cross-validated against a benchmark provided by a full nested Monte Carlo (MC) simulation.

In this third and concluding part of the series, the effects of moving capital between asset classes will be analysed, and an optimal allocation maximizing the risk-adjusted net asset value (RANAV) will be found. To determine the accuracy of the LSMC approximations, full nested benchmark evaluations are performed alongside the LSMC evaluations, and the corresponding optima compared.