Inventory Risk : A Monte Carlo Simulation Approach

Sammanfattning

This thesis examines inventory imbalance risk at a Swedish company warehouse under uncertain demand and supply conditions. Historical warehouse data from 2023–2025 is used to model demand and supply as stochastic processes in assumptions that conditions will remain the same for 2026. Event frequencies are represented by Poisson distributions, while event sizes are represented by Gamma distributions. The fitted distributions are then used in a Monte Carlo simulation with 10,000 runs to estimate possible inventory outcomes for 2026. The results show that the selected warehouse is more exposed to excess inventory than to shortages. The simulated final inventory remains positive in all scenarios, with an expected surplus of approximately 32 800 units. This indicates that, under the model assumptions, the main risk is not unmet demand but capital tied up in inventory. The financial impact is reported using a masked cost-to-sales parameter, α, where α represents the average cost price divided by the average sales price. This protects the company’s margins and sensitive operational information. With a WACC of 15%, the expected annual holding cost is expressed as approximately 820,668α. Under an illustrative 30% margin assumption, corresponding to α = 0.70, this equals approximately 574,468 SEK. Qualitative interviews support the interpretation of the results by showing that the company prioritizes high product availability, but also faces a trade-off between service level and capital efficiency. The study shows how historical transaction data can be used to estimate inventory imbalance and translate it into a financial risk framework for the company. For the selected warehouse, the simulation mainly points to excess stock rather than shortage as the relevant risk.