People always wanted to know something about future events. However, after the development of business and industrial production, forecasting became the necessity.
Numerous real-life tasks require data on future trends as well. In the XXI century, forecasting is performed using mathematical modelling with computer programs.
Extrapolation methods use the data of the previous periods to generate information about the future. This means that the past movements of the variables are used to obtain the future values. There are extrapolation methods based on the regression models, exponential smoothing, moving averages, etc. Although some methods require complicated calculations, when the software is involved, all methods are relatively simple. Regression analysis is an individual case of extrapolation, which is based on developing an equation of the line that describes the past data, and applying it to the future (Albright, Winston, Zappe, & Albright, 2004).
Extrapolation forecasts are based on the assumption that "future [is] like the past" or "if such-and-such behaviour continues in the future, then " (Chatfield, 2001). For instance, the extrapolation forecast does not take into the account random future events, political situation, or other “structural breaks” (Chatfield, 2001), e.g. oil price fluctuations, a war in the neighboring countries, etc. Therefore, extrapolation is liable to provide only short-term forecasts for the stable conditions. Long-term forecasts are expected to be “way off target” (Chatfield, 2001). If a broker motivates a customer to invest and the assumptions are based on the extrapolation forecast, some ethical issues may arise due to an inaccuracy of long-term forecasts as it may seem that the broker deceives the customer.
In conclusion, extrapolation and regression are used to obtain data about the future events based on the previously obtained data. These forecasts do not take into account the probable random events and thus, only short-term forecasts are reliable. The long-term forecasts are usually only approximate and their usage is limited.
References
Albright, S. C., Winston, W. L., Zappe, C. J., & Albright, S. C. (2004). Data analysis for managers with Microsoft Excel. Belmont, CA: Thomson/Brooks/Cole.
Chatfield, C. (2001). Time-series forecasting. Boca Raton: Chapman & Hall/CRC.
