Sport Obermeyer

Michael Plasmeier

Michael Nackoul

15.761

March 31, 2011

Question 1:

In the past, the Buying Committee of Sport Obermeyer Ltd. made decisions by arriving at a consensus after spending several hours in a meeting. However, this year Wally Obermeyer asked each member to write down their own forecast. From this data, it is our job to instruct Wally how many of each of the 10 parka styles he should order for next year.

We choose a “wisdom of the crowds” approach to processing each person’s forecast. The wisdom of the crowds theory holds that various data points will assemble into roughly a normal distribution, through the law of large numbers.

Because Wally reported that the Buying Committee’s forecasts were usually off by a factor of two times the standard deviation, we used two times the standard deviation of each member’s estimates in our model to reflect the additional uncertainty.

We based our analysis off the “newsvendor” model. This model is most appropriate when all of the stock needs to be ordered before the season begins. Although Wally does not need to order all of his stock at once, he needs to order well before the season starts.

This model seeks to balance the cost of liquidating excess inventory with the lost revenue of running out and missing a sale. Although Obermeyer is able to liquidate unsold inventory at the end of the year, they do so at a loss. The overstocking cost o was given as 8% of the wholesale price. The understocking cost u is their gross margin – which is 24% of the wholesale price.

We are looking for the point where, given our sales forecast, the next marginal unit ordered will no longer make us money, but instead cost us money to liquidate. We want to set the probability of running out to the understocking cost per item over the sum of the understocking and overstocking costs per item.

We then calculate the CDF of the probability and then take the inverse. We use the inverse CDF (k) to...