Inventory Optimisation

Mon Dec 09 2019

by John Hill

Many firms are required to hold inventory of goods for sale, or for use in their day-to-day work. Significant cost savings can be achieved by ensuring that the correct amount of inventory is held, and that the ordering process for new inventory is well-matched to demand.

Why managing inventory is complex

Managing inventory is a complex challenge due to the number of interrelated factors involved in decisions to increase (or decrease) inventory, and the associated time-delays between measuring inventory, and delivery of additional stock. Spreadsheet models are unable to capture these dynamic effects, and often lead to dynamic instabilities - such as oscillations - in the quantity of inventory held.

To understand why this is, we can set up a simple stock-flow simulation in Sheetless, and illustrate the emergence of oscillations in inventory amounts following a demand shock.

Screenshot 2019-12-09 at 16.44.59

By varying the key parameters in this model, such as the length of the delay between placing new orders and their delivery, we can see how the inventory re-ordering system generates significant oscillations in the levels of inventory. These models are able to capture complex effects that simpler predictive solutions, or spreadsheet models, simply can't.

Optimising Inventory Levels

A well-calibrated simulation of the inventory re-ordering process should be able to accurately forecast inventory levels under a wide range of scenarios, including delayed orders, and demand shocks. There is a clear trade-off between efficiency and robustness.

The model can then be used to identify the optimal level of inventory that a firm should hold in order to ensure that it remains robust to unanticipated shocks.

You can get started with a simple model of inventory under demand shocks here. If you want to make and save your changes to the model, you can register for Sheetless for free.