Dear Colleagues,
What do you think about what kind of the methods of forecasting of the economic indicators of an enterprise's activity are the best in variable environment condition?
I'm interested in quantitative methods of evaluation.
Thank you for your answer
hello,
In the following link yo can find a good information that summarize in several tables
https://hbr.org/1971/07/how-to-choose-the-right-forecasting-technique
Good Luck
It depends on the data. If the mean is changing with time and the variance is constant a model of the ARIMA could be proposed. If in addition it is seasonal a SARIMA could be fitted. If it has long memory then an ARFIMA model is suggestive. If it shows volatility (that is, it is heteroscedastic) a GARCH model is fitted.
To supplement Dr.Ette, Artificial Neural Network model can be tried if you are not sure about the data pattern. However, try some combination and choose the best based on forecast evaluation
Forecasting (and the method of) depends on your position. If you are outside the enterprise, you may apply the methods recommended by Ette and Sendhi, although I would prefer to find out (first theoretically or by common sense), whether there are leading indicators, before. If you are in the staff, you will likely have some information about the plans for the future, and you will be more interested in the economic environment to draw up realistic scenarios. If the enterprise for which you should make forecasts is a competitor or client of your own firm or your firm is a customer of that firm, you have to take account of influences of measures of your own firm, too.
- Badges
- Science topic