As a former graduate student in history -- and later as a consultant -- I was always taught that the process one took to arrive at an analysis or forecast was more important as the answer itself (and being able to explain that process was the most important piece). That's why I often find the curt pricing forecasts from Purchasing and others that I get in my inbox more dangerous than anything else. Personally, I'd rather get a combination of raw data, explanation, and expert interpretive opinion to guide my own forecasts and analysis -- not something in between that is supposed to be taken as commodity forecast gospel.
But sometimes forecasts are even worse than useless (which maybe is better than dangerous depending on your opinion). Consider this forecast from the National Weather Service that the U.S. Midwest is "given an equal chances of hot or cool June". I suppose if you work for a national agency that you can get away with such prognosticative drivel. But as a company, could you imagine telling your customers or the street that "next month our component prices have an equal chance of being above or below normal." Go figure! Hat-tip: The Cynical Sorcerer.