Par Level Calculator

Written out

A par level is the standing quantity a location should hold, replenished back up to that level on a fixed schedule: (average demand per period × periods between deliveries) + safety stock.

It is the periodic-review cousin of the reorder point. Rather than ordering a fixed quantity when stock hits a trigger, you order a variable quantity on a fixed day, whatever brings the shelf back to par.

In practice

The product sells 40 units a week, deliveries arrive weekly, and the lead time is one week. The exposure period is the review interval plus the lead time, two weeks, so expected demand is 80 units.

Safety stock over two weeks at 95% service: 1.65 × 12 × √2 = 28 units. Par level is 108.

Each week, count what is there and order the difference. Twenty-two units on the shelf means ordering 86; sixty units means ordering 48. The order size varies and the shelf position does not.

The limitations

Par levels need more safety stock than reorder points for the same service level, because the exposure period includes the review interval as well as the lead time. Moving from continuous review to weekly review on this product raises the required buffer from 34 units to a larger cover over a longer window.

Fixed pars also decay. A par set when the product sold 40 a week is wrong at 65, and because the system still replenishes to the same number, the stockouts look like a supply problem rather than a stale parameter.

Putting it to use

Review pars on a schedule, quarterly for most products, monthly for anything seasonal or growing. The whole benefit of par levels is that they need no thought day to day, and that only holds if the numbers behind them get thought periodically.

Set them per location rather than centrally where multiple sites are involved. A par that works for the busy site starves the quiet one of space and the quiet site's par stocks the busy one out.

When par levels beat reorder points

Par levels win where ordering is naturally periodic: a supplier who delivers on Tuesdays, a consolidated shipment, a route-based replenishment. Consolidating many items into one scheduled order captures freight and handling savings a per-item reorder point cannot.

They also win where counting is easy and continuous tracking is not. A stockroom shelf, a kitchen, a van, a retail backroom, anywhere a person can look and see the gap, works better on pars than on a system trigger that depends on perfect transaction capture.

Reorder points win where inventory is tracked accurately in software and items are ordered independently, which describes most ecommerce. The two can coexist: pars for the consumables and packaging, reorder points for the products, and each applied where its assumptions actually hold.

One practical refinement: express the par as a physical marker rather than a number where the location allows it. A line on the shelf, a divider in the bin, or a two-bin system where emptying the first bin triggers the order removes the counting step entirely and is far more reliable during a busy period than any figure on a sheet.

That is the same calculation implemented differently. The par is still lead time demand plus safety stock; it is simply held in the physical layout rather than in the system, which means it works whether or not anyone remembers to run the report.