Turning Another Corner: Freeway or Blind Alley?

Most of the forecast information I’ve read about 2007, our own included, indicates that at worst, 2007 will be as good as 2006, and at best it will see some growth. The indicators—employment, inflation, capacity utilization and capital expenditure—all point to positive numbers for at least the first half of the year.

Inexorably, life’s odometer has clicked over another year. It’s January again—so soon. Traditionally, it’s a time of year for reflection and optimism. I was thinking about the year past and that it was pretty good for most everyone. At shows, in shop visits, at conferences and other gatherings, the word was optimistic and the feeling upbeat.

Moving forward, 2007 looms large, and like most new years, is largely unknown. Prognosticators peer into their crystal balls and read surveys to try to get a handle on the future, but the almost infinite variability of a complex industry like ours makes these educated guesses at best. That said, we’re always looking for better tools to help in the planning process.

Most of the forecast information I’ve read about 2007, our own included, indicates that at worst, 2007 will be as good as 2006, and at best it will see some growth. The indicators—employment, inflation, capacity utilization and capital expenditure—all point to positive numbers for at least the first half of the year.

However, is this sufficient data to plan your business through the year? My company, Gardner Publications, has been working on this problem for a while now, and we’ve learned a few things. Bear in mind, we are only interested in metalworking manufacturing and have built models that are specific to the vagaries of this industry.

One model that we’re working on with some success is a trend analysis that uses, among other things, a plot called a rate of change curve. It looks at specific economic indicators in an effort to predict when something has happened in the economy rather than looking at the change in hindsight.

I talked to Steve Kline, Jr., a financial analyst here at Gardner, who has been using this contextual method to forecast our ad pages with quite a bit of success for several years. In 2006, he used this method to forecast machine tool sales as well.

There are basically three kinds of forecasting. The first Steve calls the “gut” method. Being somewhat statistically challenged, this method is my personal favorite, but I’m not running a business.

The “gut” forecasting method predicts sales through conversations with customers, business experience and a gut feel about where the industry is headed. The more experience this person has and the more the person can pull disparate facts together, the more accurate this type of forecasting is.

One of the main problems with this type of forecasting though is the psychology of the predictor. Typically, when business is going up, people tend to think it will continue to go up. And, when business is going down, people tend to think it will always go down. The philosophy at work is the same where people know they should buy low and sell high, but seem to have a hard time doing it.

The second type of forecasting we’ll call the “formulaic” method. With this method, the individual uses some kind of mathematical formula, whether it is a simple moving or complicated regression analysis, and plugs in a variety of past economic data to forecast the market. One of the problems here is that while manufacturing clearly has regular business cycles, the numerous variables that affect the timing and heights and depths of the cycle don’t always work exactly the same way a formula would.

The third type of forecasting we’ll call “contextual.” We’ve found this method uses the best of the gut and formulaic method. By using a small number of variables, such as the Fed funds rate, consumer spending and industrial production, and graphing them using rate of change curves, an individual can get a good feel (the gut method) for how these variables interact with each other. Then, using the simple formula for the rate of change curve (the formulaic method), one can forecast future sales 12 to 18 months out. A benefit of this method is an ability to see trends in the data ahead of the down or up curve and therefore, be in a position to respond more quickly.

Steve is available to talk more in depth about the new tools he is bringing to manufacturing forecasting. Call him at (513) 527-8837 or send an e-mail to skline2@gardnerweb.com to learn more about it.