How to simulate prices using historical returns rather than the normal distribution

In this course we walk-through the construction of a hybrid Monte Carlo simulation model for determining the path of future commodity prices. The model makes use of the Monte Carlo simulation structure but replaces normally scaled returns with randomly selected actual historical returns. In particular it replaces the z-scores (used in the terminal price formula of the MC model) determined using the normal distribution, with those calculated based on the actual historical returns. Results from the hybrid and original Monte Carlo simulation approaches are then compared:

Monte Carlo Simulator Using Historical Returns – Simulating Commodity Prices

The next stage is to test how the hybrid model fares in terms of Value at Risk measurement compared to the traditional Monte Carlo simulation and the historical simulation approaches. Tabular and graphical results are presented for all approaches to enable this assessment:

Simulation Models – Hacks – Results from switching Normal with the historical distribution

Leave a Comment