Safety stock is not set by falling orders
5 November 2026, 08:13
On 2 November S&P Global published the purchasing managers' index for Italian manufacturing, a monthly survey in which a value below 50 signals falling activity: in October it dropped to 49.4, from 50.4 in September, with output and new orders declining. Companies cut their purchases and, in some cases, used up the safety stocks built up when the conflict in the Middle East broke out, while supplies kept being interrupted.
Reducing inventory when orders slow down is a sensible choice, because every idle part has been or will be paid for, and that money comes back into the business only when the finished product is sold. The question is which part of the inventory to reduce, because not everything on the shelves is there for the same reason.
The stock of an item is made of two parts
The first is cycle stock: the parts that arrive with each delivery and that production uses up until the next one. If consumption falls and the same quantity keeps being ordered, each delivery lasts more days and the money paid to the supplier stays tied up for all that time.
The second is safety stock: the quantity that always stays in the warehouse, even when customer orders slow down. It has three jobs: absorbing the unexpected, that is, higher consumption than usual or a late delivery; serving customers who want the goods sooner than it takes to buy the materials and make the product; and never running short of strategic components, which must not be missing. It is measured in days of cover, that is, how long it can keep the line running: falling orders do not shorten it and only the quantity of parts changes, because if production slows, fewer are needed for the same period.
The reorder point, that is, the stock level below which a new order is placed, adds together the consumption expected over the days the supplier usually takes to deliver and the safety stock. Of this, the share that has to cover the supplier's delays is calculated as follows, if withdrawals are steady: average daily consumption multiplied by the number of days' difference between the slowest delivery and the usual one.
An example. Production uses 40 parts a day of a component; the supplier usually delivers it 20 days after the order, but in recent months it has sometimes arrived ten days late. During the usual wait the line uses 800 parts, and during the delay another 400: these last are the safety stock. The order therefore goes out when 1,200 are left in the warehouse: if the supplier is on time, the goods arrive with the safety stock still whole; if it is late, production carries on with that stock until the delivery.
The typical mistake is to cut everything by the same percentage
When the instruction comes to reduce inventory, minimum stocks and reorder points are often lowered by the same percentage for every item. This removes part of the safety stock precisely when suppliers are delivering less regularly, while the lots stay as they were. Inventory falls for a few months; then a component runs short, a line stops, and the emergency is paid for with express transport and overtime.
Cycle stock, on the other hand, is reduced by adapting lots to the new consumption, with smaller and more frequent orders where the supplier accepts it, and the reorder point is recalculated on the new consumption, which is lower. Safety stock is recalculated item by item: it falls where consumption has become more regular and the supplier delivers on time, and it rises where delays have lengthened.
How to set it up in the management software
Delivery times are not estimated from memory: the system records the date of each order to the supplier and the date on which the goods entered the warehouse with the delivery note. Whoever manages purchasing extracts those dates for the last six months and calculates, for each important supplier and item, the usual delivery time and the longest one. With daily consumption figures they then work out the safety stock and the reorder point. What to enter in the software depends on how it works. If it plans purchases starting from customer orders and the production plan, the item master data must hold the safety stock and the delivery time: the software proposes each purchase that many days in advance, so that the goods arrive before stock falls below the safety stock. If instead it only has a minimum stock field and flags the item when stock falls below that value, that field must hold the reorder point, that is, the 1,200 parts of the example, not the 400 of the safety stock: with 400 the order would go out when ten days of consumption are left, while delivery usually takes 20, and the line would stand idle waiting for the parts.
Safety stock and reorder point are calculated first for the items where a shortage costs most: those with a single supplier, those that come from far away and those without which a line stops. The calculation is repeated every quarter, or earlier if a supplier changes its delivery times. Periodically whoever is in charge of controlling divides the stock of each item by its average daily consumption: the result is its days of cover. If they exceed the interval between two deliveries plus the days covered by the safety stock, the parts beyond that threshold are surplus: multiplied by the unit cost, they show how much money is tied up for no reason. The surplus is the only cash locked up needlessly: cycle stock and safety stock, by contrast, serve production.
When demand falls and supplies become uncertain, therefore, lots are reduced while safety stocks are recalculated, often upwards: these are two separate decisions, taken with different numbers.
This article was prompted by a news item published in Milano Finanza on 2 November 2026.