Can anyone personally recommend a good approach at forecasting in Excel
(either using inherent features or a third-party application). Thanks.
The phrase "trustworthy forecasting" is an oxymoron, IMHO -- unless
you believe the results of tarot readings, crystal balls and
astrology.
Forecasting is as much an art as it is a "science" (if statistical
extrapolation can be called a science). Whether or not there are
"good" approaches is debatable; but there are certainly __bad__
approaches. The worst approach is to use the Excel forecasting
functions (e.g. FORECAST) blindly.
IMHO, choosing a "good" approach to forecasting first requires an
understanding of the nature of what you are trying to forecast. For
example, forecasting seasonal sales is very different from forecasting
the equity prices.
But if you "must" forecast without an understanding of the nature of
you are trying to forecast, with Excel I would start by using Chart
Wizard and Trendline to look at the numbers and some possible
regression approaches. (Caveat: Some people are critical of the
Excel regression algorithms.) Then you might be able to choose an
appropriate Excel forecasting function to use and the appropriate
(sub)data to use for your extrapolation.
