A simple version of the trend extension method is the application of the historical rate to future growth. If demand during the last five years has grown at an average rate of, say, ten percent, it may not be unreasonable to estimate that demand is likely to grow at the same rate for the next year. Projected growth based on historical data can be modified by experts for the likely impact of new factors, eg: changes in produce prices, additional credit availability, increased area under irrigation or new seed varieties. An expert assessment of the net effect of these factors can be applied to the growth rate suggested by historical data. In a large country, with many agro-climatic zones, the procedure carried out for each province, aggregated and moderated at the national level, can yield reliable results. This method is useful for short-term forecasting of demand at the country level and of sales of companies.
The assumption that past growth will prevail in the future may often not be true in practice. In countries with large fertilizer consumption per hectare the growth rate is difficult to maintain. Similarly, countries with low consumption may grow at a much faster rate than in the recent past if conditions change. Different patterns of past growth can be converted into best-fitting mathematical equations which do, however, require expert intervention. Even with application of mathematical tools, it is advisable to have a final moderation of the results by a group of experts with good insight into the fertilizer situation. More than the need for statistical expertise, non-availability of reliable historical data is a major impediment to the use of the trend extension technique.
Demand estimation by the trend extension method is useful to a Ministry of Agriculture provided several years' data is available and growth trends are clear and well established. It is not suited for "new situations" when fertilizer usage is at an elementary stage. Considering the availability of computers and statistical skills, the method is not very expensive or difficult to handle provided data is available. If a less complicated approach is preferred, the simpler growth rate method can be adopted, amended to include the effect of new factors assessed by experts. Fertilizer marketing organisations and wholesalers will find the simple inexpensive growth rate method useful for short-term sales forecasting if suitable allowance is made for the fact that historical sales data of individual companies is not a reliable indication if the overall market itself is affected by major changes.
