Professional Documents
Culture Documents
The
backorder is triggered when the inventory level hits the reorder point. The EOQ is calculated
in order to minimize a combination of costs such as the purchase cost (which may include
volume discounts), the inventory holding cost, the ordering cost, etc. The order quantity
optimization is complementary to the safety stock optimization that focuses on finding the
optimal threshold to trigger the reorder.
$H$ be the carrying cost per unit for the duration of the lead time (1).
$\delta$ be the delta inventory quantity needed to reach the reorder point (2).
$\mathcal{P}$ be the per unit purchase price, a function that depends on the order
quantity $q$.
(1) The time scope considered here is the lead-time. Hence, instead of considering the more
usual annual carrying cost $H_y$, we are considering $H = \frac{d}{365}H_y$ assuming
that $d$ is the lead time expressed in days.
(2) The delta quantity needs to take into account both the stock on hand $q_{hand}$ and the
stock on order $q_{order}$, which gives the relationship $\delta = R - q_{hand} - q_{order}
$ where $R$ is the reorder point. Intuitively, $\delta+1$ is the minimal quantity to be ordered
in order to maintain the desired service level.
Then, the optimal order quantity is given by (the reasoning is detailed below): $$Q =
\underset{q=\delta+1..\infty}{\operatorname{argmin}}\left(\frac{1}{2}(q-\delta-
Cost function
In order to model the cost function for the order quantity which takes into account volume
discounts, let's introduce $R$ the reorder point. The inventory cost is the sum of the
inventory carrying cost plus the purchase cost, that is: $$C(q)=\left(R+\frac{q-\delta-1}
{2}\right)H+Z\mathcal{P}(q)$$ Indeed, taking an amortized viewpoint over the lead time
period, the total quantity to be ordered will be $Z$ the lead demand.
Then, the inventory level is varying all the time, but if we consider strict minimal reorders
(i.e. $q=\delta+1$) then, the average stock level over time is equal to $R$ the reorder point.
Then, since we are precisely considering order quantity greater than $\delta+1$, those extra
ordered quantities are shifting upward the average inventory level (and also postponing the
time when the next reorder point will be hit).
The $(q-\delta-1)/2$ represents the inventory shift caused by the reorder assuming that the
lead demand is evenly distributed for the duration of the lead time. The factor 1/2 is justified
because an increased order quantity of N is only increasing the average inventory level of
N/2.
In practice, unit price rarely increases with quantities, yet, some local bumps in the curve may
be observed if shipments are optimized for pallets, or any other container that favors certain
package sizes.
In the Excel sheet attached here above, we are assuming the unit price to be strictly
decreasing with the quantity. If it is not the case, then edit the macro EoqVD() to revert back
to a naive range exploration.
Wilson Formula
The most well-known EOQ formula is the Wilson Formula developed in 1913. This formula
relies on the following assumptions:
The rate of demand is known, and spread evenly throughout the year.
$S$ be the fixed flat cost per order (not a per unit cost, but the cost associated to the
operation of ordering and shipping).
little or no benefits in marginal unit cost afterward. In such a situation, the Wilson Formula is
more appropriate.