Statistical Forecasting in Business - Term Paper Example

Applications of Statistical Forecasting in Business April Executive Summary Business forecasting models which include moving averages, simpleaverages, exponential smoothing and some decomposition (Lapide, L. 2000). These models are easy to use as the managers’ only needs to insert data and then use predetermined formulas for the time series model values. According to many researchers in the field of business forecasting, businesses which have embraced the application of forecasts are more likely to have a competitive edge than the other who have not since their decisions are backed by tangible data (foundation). It is always advantageous to use historical data to project future positions and then plan accurately. This is positive planning and positions a business strategically in the market. Introduction According to Allen, D. (2000), growing a business has become such an enormous challenge although the greatest challenge to him is how to support human resources and financial commitments without accurate forecast data. Un-reliable forecasts compromise on the kind of data developed and the profits realized by a business venture. According to Allen further, organizations/companies which are good at producing forecasts are aid to be stable and better partners by donors. Lapide, L. (2000) on his part argues that the current competition organizations/businesses face in addition to tight profit margins, advanced planning is encouraged in order to survive. Lapide further asserts that, companies/businesses that don’t plan accurately are more likely to incur extra production costs which translate to low profits and thus less growth. In summary, Lapide argues that taking advantage of future business forecasts guards a business against unwanted costs/expenses in addition to giving the manager greater understanding of the business activities. Applications of Statistical Forecasting in Business To place the business/company strategically in the market, business forecasting which is a process used by many managers to take advantage of past business/company information for future benefits through the estimation of expected future patterns using the existing business/company data. These estimations as Lapide puts it include; Quarterly sales/marketing data; Product supply and demand functions; Customer prevalence and behaviours; Potential customer base; Competitor capabilities; Future human resource requirements; and Future business risk. Useful business/company forecasts are done using time series analysis (through the applications of moving averages, Holt’s methods among many others) and regression analysis (both simple and multiple regression analysis). Well documented and supported (statistically) business forecasts supports managers and other business analysts in making tangible decisions which drive businesses notches. Further, such forecasts form part of the business/company business intelligence which managers build on for competitive advantage (Lapide, L. 2000). In the use of moving averages, then moving averages of a certain order (depending on the researchers/business analysts liking) are developed and then compared with the actual data by plotting the values (see table 1 below). Table 1: Sales Data Vs Predicted Data Sales MA (2) 45345 - 34545 17272.5 4545 2272.5 675 337.5 67567 33783.5 7567 3783.5 6767 3383.5 67567 33783.5 7567 3783.5 67567 33783.5 57565 28782.5 7687 3843.5 87878 43939 98798 49399 Note: two (2) moving averages of the sales data Figure 1: Sales Vs Predicted Data (MA (2)) Figure 2 shows the relationship between the predicted and the actual sales values. From the graph, it is evident that forecasted data can be used to provide future insights on how the business with perform. With such information, then Lapide assertions will be true. On the other hand, regression analysis in business forecasting answers the question “by how much”. In this regard, both simple and multiple regression analysis shows the relationship between the dependent variable (DV, for example volume sales, market share etc) against the independent variables (IV: within subject factors e.g. products differentiation, product prices etc.). A business normally comes up with a model based on its operations (what it deals with) and then using the available historical data formulates a regression equation which can be used ably to predict the future values. An example of a regression model is shown below; SV = β0 + β1P + β2M+…+ βnXi In this model, β’s are the coefficients of the various independent variables and which signify the amount of dependent variable change (in this case sales volumes), either positive or negative, once one of the independent variable’s changes by a single unit. P, product differentiation, and M, market share, are examples of independent variables. βn shows that the variables can be as many as possible depending on the business itself. According to Jain, C. L. (2000), business forecasting is normally completed through the application of data mining processes where models are used to predict future using existing/historical data from which business potential if projected based on different business activities as well as transactions. To Jain, the sole purpose of data mining is to establish the presence of predictive criteria which might be present if large volumes of historical data are used. Business Forecasting Initially, business forecasting used to be done using common sense but over the years it has changed so much towards scientific approach where many methods and advanced techniques have been developed each meant for the forecast of certain types of data (Jain, C. L. 2000). According to Jain, the internet had had positive contributions towards forecasting as it has brought about so many tools which aid in forecasting. Further statistical packages are nowadays available and include; Excel; STATA; SPSS; Minitab; Among many others. To Culbertson & Lee (2001), many businesses have recognized the advantages of forecasting and are spending quite a good chuck of their profits to come up with useful forecast. According to them further, many businesses have tailored their profits to forecasts such that a certain fraction of their profits are made forecasts alone. Conclusion Business analyst are better placed today since with historical data and knowledge to forecast, It is easier to come up with tangible decisions which are valid and reliable; It is faster to get a deeper understanding of business performance; It is faster and easier to plan accurately; It is possible to evaluate different business strategies at different situations and quantify how much the dependent variables changes with a unit application of change on the independent variables; and It is possible to understand the market and plan adequately for the different consumers. Personal Opinion Personally, all business decisions need to have a full backing which is only available through the use of historical data. However, the following skills are important if statistical business forecasting is to have an impact; Statistical packages knowledge for faster analysis data presentation are important; Regression skills are important; Further, EDA enables the manager to understand the variables better; and Data analysis and interpretation of the results is an important skill if the results are to achieve their intentions In this regard, there is need for more research in the field of; EDA in business data; and Ability to interpret statistical outputs by managers. References Allen, D. (2000). "Looking Forwards." Management Accounting. John Hopkins University. Baltimore Culbertson, S.; Jim, B. & Lee, B. (2001). “Control System Approach to E-commerce Fulfillment.” Journal of Business Forecasting Methods & Systems, 23(1), pp.34-45 Jain, C. L. (2000). "Which Forecasting Model Should We Use?" Journal of Business Forecasting Methods & Systems, 45(2), pp. 23-41 Lapide, L. (2000). "New Developments in Business Forecasting: The Internet Does Not Eliminate the Need to Forecast." Journal of Business Forecasting Methods & Systems, 56(1), pp. 21-34 Read More