Daily Archives: August 21, 2007

Forecasting, Part II

Back in January, I wrote an introductory post on demand Forecasting that discussed a really good article called Outlook Warm and Sunny that ran in APICS Magazine.

In the post, I noted that the proper combination of judgmental and statistical and statistical methodologies can often be used to create better forecasts than either method alone, because humans can make mistakes and statistical methods are very slow to react to changing market conditions.

Well, a month or so ago, Purchasing ran an article titled Commodities forecasting: It’s all in your head, that addressed the subject of commodity price forecasts. The article indicated that of the three types of commodity price forecasts – those based on judgment, those based on historical price data, and those based on commodity futures prices – the judgmental forecasts have the best record of accuracy.

Although this worries me, since the risk of human error is just as real in commodities forecasting as it is in demand forecasting, and I would strongly recommend that you use a good statistical model that incorporated historical prices and futures pricing before making a judgment as to what the price will be in the future, the article does have some good advice.

First of all, it notes that if you are going to use personal judgment in commodity forecasting, you have to base it on the right factors. Specifically, you need to review the right data, and the most critical data is:

  • market intelligence
  • global economic trends
  • supplier safeguards against volatility
  • your own company’s strategies

Market intelligence is probably the most critical. Be sure to watch the international marketplace to determine supply, demand, pricing, and trading trends. Economic conditions that affect supply and prices are changeable, and usually global.

The article concludes with ten forecasting tips.

  1. determine corporate price goals and adjust them to economic realities
  2. reduce purchasing pricing strategies to monthly or quarterly intervals to represent the rapid dynamics of the marketplace
  3. adjust actual timing of buys to monthly or quarterly events
  4. analyze and adjust the structure of supply contract agreements
  5. pay close attention to inventory levels
  6. ensure true supply tie-in with operational action plans
  7. study global pricing and sourcing trends
  8. survey primary suppliers’ operating rates, inventory, and costs
  9. analyze the secondary sourcing market for alternate suppliers
  10. determine potential supply alternatives for commodity products