Pages - Menu

Wednesday, May 29, 2013

Are you still using Min-Max in your Inventory Management?

Some policies and practices that are prevalent in management are seriously counterproductive to the cause of companies, nevertheless they stay the norm. I used to be surprised whenever I stumbled against such practices with my clients, asking myself how is it that given the knowledge generated and advancements in technology companies stick to these practices that are keeping them away from the objectives they want to achieve. We term these practices as “disabling procedures”.

This article, is not about the reasons for the continued existence of these practices, it is about one of these practices that is common in the supply chain arena. A practice that delivers more damage than good and that is a major contributor to the exact phenomena it is supposed to be solving. This is the “Min-Max” practice.
 
When managing inventory one aspires to hold as little as possible, yet sufficient to ensure availability is appropriate. Min-Max is supposed to be providing the most economical quantity to hold for this purpose. And whether one uses mathematical equations like EOQ and its derivatives (EOQ stands for Economical Order Quantity) or plain judgment the resulting practice is one of the following:

  • Reorder Quantity – When the inventory reaches a low boundary (Min) an order to renew the inventory to its Max level is placed (no matter how much time has passed). Or,
  • Reorder Point – When a given amount of time has passed since the previous supply, an order to renew the inventory to its max level is placed (no matter how much was consumed).

 Graphically it will look like that:


As mentioned earlier we seek to “optimize” our inventory – hold as little as possible, yet ensuring sufficient availability. However, we commonly do not know in a good enough precision what the consumption will be in a given period of time, thus we use some form of forecasting to set the inventory targets of Min and Max. Again, it is not mandatory that we use a formal forecasting system, even just deciding about it, is a form of forecasting. As a result, the accuracy of our inventory is highly dependent on the accuracy of our forecast. So, what influences the accuracy of forecast?



Forecast accuracy is directly related to two factors:

  • Forecast horizon – The longer is the forecast horizon the less accurate the forecast is, and
  • Sample size (what portion of the consuming population is covered by the forecast?) – The smaller the sample size the less accurate the forecast is.

Any Min-Max policy is artificially elongating the forecasting horizon; the reorder quantity elongates the horizon by the time from one supply to the time the inventory level reaches the predetermined “Min” level. The reorder point elongates the horizon by the time from one supply to the predetermined reordering time. By elongating the horizon unavoidably the accuracy of the forecast deteriorates and thus the accuracy of the inventory deteriorates and, it shouldn't be a surprise that this results with misalignment between the inventories on hand to the inventories required. Statistically whether an item has surplus or shortage is a random event, still when using Min-Max it is a certain one.

Min-Max has other negative effects as well, some by itself and other when combined to other common practices in inventory management. Here are a few examples:
  • Often time Min-Max is done on a “line-item” consumption against some sort of Pareto analysis, resulting with differentiated Min-Max policies to the different Pareto categories. As a consequence of that different line-items even when are part of the same order, or same supplier, or same location etc. are handled with separate policies ensuring that the overall performance is negatively affected. Just consider – do you but from one supplier only “A” category products? Is it that your “A“ category are also “A” category for your supplier? When your customers buy from you, do they buy only products from one Pareto category? What is your level of service if on average a customer order has 2 “A” items, 3 “B” items and 5 “C” items, when your availability for “A” items is 99%, “B” items 90% and “C” items 85%? (it is 37%)
  • When you use Min-Max the quantity and timing of ordering is at your convenience (or at least your supposed convenience) and it does not consider the supply considerations resulting with misalignment that often times mandates violation of the Max considerations thus inflating inventories above the maximum desired level.

One of the original reasons for using Min-Max is the optimization of the cost of ordering. In the past placing an order required meaningful handling and manpower and thus there was an incentive to order as infrequently as possible so that the cost associated with reordering is minimized (of course without inflating inventories). However, have you noticed that for a very long while now, there is no cost associated with reordering? That your computer system can generate inventory renewing orders as frequently as you would like it to be, and for generating this order there is absolutely no cost?

What if we did reorder immediately (which does not necessarily means resupply is required immediately) as an item is consumed from our inventory? What will be the effect of that?

FROM     TO

   

This is exactly what we are looking for, aren't we? Lower levels of inventory, still meaningfully more accurate forecast thus more accurate inventory. The positive effects are numerous, can you see them from the picture above? While the cost and risk, are zero, and it is quite simple to implement. So, why are you still using Min-Max?

1 comment:

  1. I have been retired for a number of years now, but when I was on the Navy ships, we used both methods. For slow moving items (less than 1 demand in six months) we had an established allowance. When we issued anything to a customer, we immediately ordered a like quantity to replace in the inventory. For fast moving items (2 or more demand in six months), we referred to them as Selected Item Management or SIM items, we established min/max quantities based on usage as well as order/ship time. These were established to minimize the number of orders generated. To make the system better, once an item became a SIM item needed to maintain a minimum of 1 demand per year to remain as SIM although the min/max levels might be recalculated based on the most recent demand profile. This combination worked well in ensuring the inventory was able to provide continuous support to the ship. Today the ships have computers to do the manual work and have probably improved on what we were using but this was very effective for us.

    ReplyDelete