Quantitative Methods

The Newsvendor Problem and the Cost of Ordering Wrong

One ratio, computed from your own margins, tells you what service level an item actually deserves. Run the same 95% target across your whole assortment and some of those items lose money on every unit you stock.

  • Advanced
  • 12 min total
  • 14 chapters

What decision this helps you make: How many units to buy for a single selling season, and what in-stock target each item in your range is worth — derived from its economics rather than inherited from a policy.

What this topic is

The newsvendor problem is the single-period ordering decision under uncertain demand: you commit a quantity before you know what you will sell, being short costs you margin and being long costs you the write-down. The solution is a ratio of those two costs, converted into a quantity through the demand distribution — and it is the foundation under safety stock, capacity booking, and airline seat protection alike.

Why it matters

Most businesses set service levels by policy rather than by arithmetic. A single company-wide target of 95% or 98% is applied across items whose margins and write-down risks differ by an order of magnitude, which systematically over-buys the low-margin, perishable end of the range and under-buys the high-margin end. The newsvendor ratio is the correction, and on the worked example in this lesson the difference between the policy answer and the economic answer is the difference between a $2,400 profit and a $3,000 loss on one item.

Who should learn it

Buyers, merchandisers, inventory planners, restaurant and bakery operators, event and ticketing businesses, and anyone committing to a quantity before demand is known.

What you will understand

  • How to compute the critical ratio from underage and overage costs, and turn it into an order quantity
  • Why the profit-maximising in-stock probability is usually not 95%, and is sometimes far below the mean
  • The difference between in-stock probability and fill rate, which are routinely confused in service-level policy
  • Why the profit curve is flat near the optimum and asymmetric around it, and what that means for how much precision to chase

Prerequisites

Common misconception

"Higher service level is better; we target 95% across the range." Service level is an output, not an input. It falls out of the ratio between what a stockout costs you and what a leftover unit costs you, and for an item with a thin margin and a steep markdown that ratio can put the right in-stock probability well below 50% — meaning the profit-maximising order is smaller than expected demand. On the low-margin item worked through in this lesson, running a 95% policy instead of the economic answer converts a $2,400 seasonal profit into a $3,000 loss. The policy did not make service better. It bought inventory at a price the margin could never repay.