This summer, Supply & Demand Executive published an article on how to make EOQ relevant again. Going through the recent archives, it got my attention because it is simultaneously a metric that should have never lost relevance and a metric that loses relevance when too much emphasis is placed on JIT or avoiding stock-outs (at all costs).
EOQ, short for Economic Order Quantity, is an old-school metric whose function is to identify the optimum order size that has the lowest cost. Defined as the square root(2UA/IC), where:
- U is the (annual) usage,
- A is the acquisition cost per order,
- I is the inventory carrying rate, and
- C is the cost per item
if demand is relatively constant, or at least known and predictable, the item is purchased in lots or batches, and the order preparation and inventory carrying costs are constant and known, then the formula tells you how much to order (and, as a result, how often to order) to minimize your overall order cost.
But, as the article points out, even assuming you can mesh this with your JIT production schedule (timing the orders so that you don’t carry too much inventory but still keep production running smoothly), these are not all of the costs associated with an order. Other costs include:
- purchase order processing cost (which should be included in acquisition cost)
- a true ICC rate (and not the holding cost computed as the extra cost of money invested in stock) that takes into account opportunity costs (which should be included in I)
- taxes paid on inventory (which can be substantial and why many auto dealers, for example, have year end clear-outs) (which should also be included in I)
- stock quantity shrinkage loss due to handling, pilferage, and theft (which should also be included in I)
- stock risk losses due to product obsolescence, deterioration or shelf life expiration (which should be factored into C)
Thus, in order to use EOQ, you need to first insure that all of these costs can be accounted for. Then, as per the article, you need to insure that:
- the product(s) are offered at a single price,
- demand is predictable,
- prices will not change (significantly) during the time the order is in stock,
- stock will not exceed the shelf-life,
- a single order can be placed, and
- freight is included in the purchase price, or can be factored in.
In this situation, if the additional costs identified above are factored in, EOQ is still very relevant and should be used. However, if multiple assumptions are violated, EOQ may not be appropriate, and, more specifically, if demand is slow, or very unpredictable, then JIT is the preferred method.
And, furthermore, if you
- establish minimum quantities to reflect minimum purchase volumes,
- set a maximum ceiling stock for difficult, bulky, or large items, and
- base adjustments on multiples of packaged lots
then you may find that EOQ is still right for you.