What is the Newsvendor Model?

Richard F.


The classic newsvendor’s dilemma is how many newspapers to order for a given day. If they order too few they miss out on sales and could lose the customer to the competition. If they order too many they end up with goods they cannot sell. The ebb and flow of breaking news complicates things further.

Many businesses with short shelf-life products face a similar dilemma. Restaurants take a chance when they let their protein defrost naturally for greater succulence, as do ice-cream parlors when they put their goods out on display. Ordering summer swimsuits in the latest fashion colors can be risky too. The newsvendor model is a way of predicting demand based on known variables.

The Newsvendor Formula

Unless you are a statistician who gets a kick out of page-long formulas, I recommend you buy a computer program or pay a college kid some pocket money. I’ll introduce the concept here. Just don’t ask for a live demonstration! There are a number of variations on the newsvendor model. Common denominators in all formulae include:

1. The degree of uncertainty / certainty of demand as expressed in the distribution, mean and standard deviation

2. The cost of buying one more unit greater than the demand that actually occurs

3. The loss associated with a single stockout in terms of reduced profit and goodwill

4. The decision generated by the model in terms of order quantity, safety stock and overbooking level.

The following graph, provided by Brunel University in London, illustrates the critical tipping point this approach is trying to isolate.

Practical Example

Here’s a worked example provided by the industrial engineering & logistics management department of HKURST University in Hong Kong. It relates to a bouquet flower seller trying to decide how many blooms to order in.

Number of bouquets 3 4 5 6 7 8 9
Probability of Selling 0.05 0.12 0.20 0.24 0.17 0.14 0.08
Expected Profit 45 57.5 64 60.5 45 21 -10

The model predicts that profit will be maximized with stocks limited to five bouquets. This is a simple example but it illustrates the principle.

Other Applications

Although the newsvendor model is based on assumptions, it has been proven to be more accurate than guessing. One example would be a big ticket open air sports meeting. Attendance and program sales depend on variables such as the weather and competing events, but the decision must be taken days before because printing and binding take time.

Next time you visit your newsvendor and find your favorite journal is out of stock, spare a thought for the fellow behind the counter who may not even know the newsvendor model exists. His method of estimating demand is less accurate than the newsvendor model described in this article.