Forecasting Techniques

In order to make decisions that will enable an organization to be successful, managers need to be able to predict the needs of the future so that the organization can act appropriately in order to gain or maintain a competitive edge. Forecasting is the science of estimating or predicting future trends. Forecasts are used to support managers in making decisions about many aspects of the organization including buying, selling, production, and hiring. Statistical techniques are available to help managers examine the impact of trends, business cycles, seasonal fluctuations, and irregular or random events on future needs. However, in isolation, these methods are not sufficient for developing good forecasts. Expert judgment needs to be used in combination with statistical techniques in order to optimize the effectiveness of each.

Since the beginning of written history, human beings have been interested in learning what the future holds. From our vantage point in the 21st century, we read with bemusement in history books about signs and omens and oracles, and wonder why people ever thought they could read the future in the entrails of a goat. Despite our relative sophistication, however, the desire to know the future persists, and we still long to know what is going to happen. We listen to the television meteorologist to find out whether or not to carry an umbrella. We read the business section of the newspaper to find out the cost of a barrel of oil to determine whether we should wait a few days to buy gas or get it today. We plan our vacations for sunny climes at times when we expect to be knee-deep in snow at home.

Importance of Trends to Business Operations

However, it is not only in our daily lives or in these relatively trivial examples that we need to know what will happen. In the business world, organizations need to know the trends of the marketplace in order to best position themselves to leverage this knowledge into profits. The production manager needs to know if there will be a continuing need for widgets and how much raw material is needed to meet the anticipated demand. The marketing manager needs to know whether changing demographics in the marketplace mean that a new marketing strategy will be needed. The shipping manager needs to know whether or not the price of oil will continue to rise and how this cost affects the outsourcing for the production of gizmos. The human resources manager needs to know whether or not the turnover in the organization will continue and new sources of qualified employees need to be found. To answer these and other questions about the future of the business and how best to respond to the changing needs of the environment and marketplace, businesses rely on forecasting. This is the science of estimating or predicting patterns and variations. Forecasts are used to support managers in making decisions about many aspects of the business including buying, selling, production, and hiring. It is part of the responsibility of management to determine the goals and direction of the organization for both the short and long terms. To do this, it is helpful to be able to predict the variations of economic activity that may affect the business and plan to either leverage these into successes or prepare the organization to survive until the next boom.

Causes of Variation in Economic Activity

There are a number of causes of variation in economic activity: trends, business cycles, and seasonal fluctuations as well as irregular and random fluctuations. Trends are persistent, underlying directions in which something is moving in either the short, intermediate, or long term. Many trends tend to be linear rather than cyclic, steadily growing (or shrinking) over a period of years. For example, in the US there is an increasing trend for outsourcing and offshoring of technical support and customer service in the high tech industry. On the other hand, trends in new industries tend to be curvilinear as the demand for the new product or service grows after its introduction then declines after the product or service becomes integrated into the economy.

Business Cycles

Business cycles are continually recurring variations across the total economy. Such expansions or contractions of economic activity tend to occur across most sectors of the economy at the same time. For example, several years of a boom economy with expansion of economic activity (e.g., more jobs, higher sales) are frequently followed by slower growth or even contraction of economic activity. Business cycles tend not to occur only across one industry or business sector, but often occur across the economy in general.

Seasonal Fluctuations

Many industries also experience seasonal fluctuations -- changes in economic activity that occur in a fairly regular annual pattern. Seasonal fluctuations may be related to the seasons of the year, the calendar, or holidays. For example, office supply stores experience an upsurge in business in August as children receive their school supply lists for the coming year. Retail stores make a significant portion of their profits in the weeks between Thanksgiving and Christmas. Travel agencies experience a rise in clients in the winter who want to visit warmer climes and in the summer for families who need to go on vacation when their children are not in school.

Economic Fluctuations

In addition, there are irregular and random fluctuations in the economy that occur due to unpredictable factors. For example, natural disasters, political disturbance, strikes, and other external factors can cause fluctuations in the economy. In addition, there are unpredictable or random factors that can affect a business's profitability such as high absenteeism due to an epidemic. Although this category is by definition difficult if not impossible to predict, there are tools available that can help the manager recognize and predict the other kinds of variations. Identification of a these variations in economic activity allows managers to better plan to meet future needs and keep the business profitable.

Approaches to Forecasting in Business

There are a number of approaches to forecasting that are used in business. Subjective forecasting methods are used by many managers, particularly when a decision needs to be made quickly.

Subjective Approaches

Subjective approaches to forecasting are qualitative rather than quantitative and based on the manager's experience and intuition about a situation rather than on the application of mathematics or an attempt to reduce the situation to quantifiable terms. Depending on the manager and his/her experience, however, subjective approaches to forecasting may or may not be effective. Further, because they are subjective, even when they are effective, subjective approaches are typically not reproducible. There is also no way to quantify the variables used in the forecasting process so that the process can be applied to future situations. The quality of a subjective forecasting approach is completely dependent on the skill and expertise of the manager using it.

Structural & Economic Model Approaches

A second category of approaches to business forecasting are structural and economic models. These approaches use mathematical and statistical techniques to support the development of a forecasting model. Structural models are sets of mathematical functions that are designed to represent the causal relationships within the organization's environment. For example, if an organization was interested in investing in a new venture, it would be helpful to know the future price of its current product in order to forecast its profits and the resultant availability of funds for the new venture. To do this, the manager or analyst might build a specification model of the factors affecting the supply and demand for the current product and the relationship between the factors. There are, however, a number of sources of error associated with this approach to forecasting. First, actual future wages and income will more than likely differ from the estimated values used in the development of the model. Second, the predictive value of the model will be affected by the sampling error in the estimates used in the construction of the model. Third, the model may not take into consideration some important variable which, in turn, could skew the results.

Deterministic Model Approach

A third approach to business forecasting is the use of deterministic models. These models assume that the variable of interest is a deterministic function of time and does not include the effects of any underlying data uncertainty or variability in the time series. Although this type of model means that the observed changes are due solely to changes in the components of the model, there are some drawbacks. Perhaps the biggest objection to deterministic models is that they assume that the time series used as a basis for the model is systematic and highly predictable. This assumption is typically not valid when dealing with real life problems, particularly when they are complex.

Ad Hoc Forecasting Formulas

Another approach to forecasting is the use of ad hoc forecasting formulas based solely on past history. This approach typically uses weighted moving averages. Although this approach simplifies computation, there is no way to determine whether the formula chosen is the most appropriate. This means that there can be no concomitant confidence in the worth of the resultant forecasts.

Time Series Analysis

A final approach to forecasting is the use of time series analysis. Time series data are data gathered on a specific characteristic over a period of time. To be useful, time series data must be collected at intervals of regular length. In time series analysis the sequence of observations is assumed to be a set of jointly distributed random variables. Unlike the ad hoc approach to forecasting where it is impossible to tell whether or not the formula chosen is the most appropriate for the situation, in time series analysis one can study the structure of the correlation of variables over time to determine the appropriateness of the model. The model can be adjusted as needed to make it more representative of the real world situation.

Applications

There are a number of statistical techniques that can be used to analyze time series data and forecast industry or marketplace trends or other factors of interest to the business. Some of these techniques are regression models to analyze trends and decomposition techniques to determine seasonality. Statistical methods are not the only important factor in forecasting, however. The analyst and decision maker both must integrate statistical techniques with human judgment in order to maximize the utility of time series forecasting.

Trend Analysis

In business, it is often helpful to be able to forecast trends in the industry, the supply chain, applicant pool, or other factors that affect the ability of the organization to do its job. Using time series data, it is possible to analyze trends to give managers the information that they need to make decisions about the direction the business should take. One of the ways that times series data can be analyzed is through the use of regression analysis. This is a statistical technique used to develop a mathematical model for predicting one variable from the knowledge of another variable.

Lumber Industry & Regression Analysis

An example of the use of regression analysis in determining trends comes from the lumber industry. Black spruce is the most important source of structural lumber in eastern Canada as well as the most important tree for reforestation. There is currently a higher demand for structural lumber than there was in the past, and this trend is expected to continue. To help the lumber industry better choose trees that will meet industry standards for structural lumber, it would be helpful to have a model to predict lumber grade yield for the black spruce. Liu, Zhang, and Jiang used regression analysis to develop a model for forecasting lumber grade and yield. A number of tree characteristics are important to the yield and quality of black spruce lumber including tree size, tree taper, stem form, crown size, and branchiness. To help the industry better identify trees to meet industry standards for structural lumber, they developed a model to identify the variables that significantly influence the yield and quality of black spruce lumber. The purpose of this model was to identify which characteristics are most predictive of lumber grade yield and help improve the prediction of lumber grade yields for the visual grading system.

To develop the model, measurements of various characteristics of interest were taken from 139 randomly selected trees. The trees were then processed and the quality of the resultant lumber graded. The authors used stepwise multiple regression analysis to build a model to predict lumber grade and yield using these data. The first attempt at model building utilized data from all the variables collected. A subsequent model was developed that used only those variables that seemed to be most directly related to yield and lumber grade. The latter model proved almost as effective as the first model. It was determined that either model developed in the study could provide an estimation of black spruce lumber grade and yield either from individual trees or stands of trees using only forest industry data. These models can aid significantly in choosing the trees most likely to have the highest yield of lumber and can also be helpful for forest management and preservation of black spruce supplies.

Seasonal Effects

Another important factor in forecasting is industry seasonality. Seasonal effects are patterns that occur in cycles of less than one year and that are associated with time cycles such as the calendar, seasons, or holidays. However, seasonal effects are not the only source of variability in time series data. As discussed above, time series data can also be affected by trends, business cycles, and irregular or random fluctuations. To determine the effects of these variables, a technique called decomposition is often used. This technique allows the analyst to break down time series data into the component factors of trends, business cycles, seasonal fluctuations, and irregular or random fluctuations.

Seasonality & the Non-OPEC Oil Supply

An example of the importance of determining seasonality and the techniques to do so is given in a recent study of the seasonality of non-OPEC oil supply. For reasons that are obvious at the gas pump each week, short-term changes in oil supply are constantly under scrutiny by analysts. The perceptions of immediate supply have implications not only on energy prices but on crude oil transport prices as well. Therefore, it is important to forecast oil supply fluctuations as accurately as possible. Two-thirds of the oil supply is produced by non-OPEC countries. Whereas OPEC decisions about supply are coordinated between the member countries, this is not true for non-OPEC countries. Decisions in these countries are made by each producer individually. This fact makes analysis of this industry segment particularly important for forecasting supply.

There are a number of reasons for seasonality in oil supply in non-OPEC countries. These include demand for oil, price of oil, stock levels, annual maintenance schedule at production facilities, the psychology and manipulation of the market, timely completion of development projects, as well as irregular and random factors including severe weather, floods, earthquakes, and strikes. There is also a seasonal component to the non-OPEC supply. Cold weather in the winter in the northern hemisphere where the major consumers are located leads to a higher demand for heating oil than in the spring months. During the summer, cooling systems and vacation travel again raise the demand for oil. The resultant seasonal fluctuations in demand for oil yield concomitant variations in non-OPEC oil supply. Because of this cycle, regularly scheduled maintenance -- another seasonal factor in oil supply -- is frequently scheduled for periods of lower consumption and more temperate weather.

Jazayeri & Yahyai performed an analysis of seasonality of nonOPEC supply in order to help improve the accuracy of short-term supply forecasts. Their study was based on the assumption that observed seasonality cycles are independent of other factors and that they will continue into the future. To analyze the time series data on oil supply in non-OPEC countries, the authors decomposed the data into the four components of trends, business cycles, seasonal fluctuations, and irregular or random variables. Using the decomposition technique of Fourier spectral analysis, it was found that non-OPEC supply follows a seasonal pattern that repeats annually irrespective of other trends. These results can be very useful for analysts forecasting oil supply at various times of the year.

Integrating Statistics & Judgment

Opinion about the best way to forecast for business decisions often seems to be sharply divided between those that rely on statistical methodology and those that prefer to use their "gut" to determine where the industry, supply chain, or market is going. However, both approaches have advantages and disadvantages. Statistical methods are less prone to bias than are judgments. In addition, they tend to be more reliable and can more efficiently make use of historical data. However, statistical techniques can only work with the data they are given. Judgmental decision making, on the other hand, can be useful particularly when there are recent events about which the decision maker is aware but which have not yet had sufficient time to result in observable data in a time series. This type of data includes information about events that have happened in the past but are not expected to recur in the future or events that will affect the future but have not occurred in the past (e.g., the effect of an innovation on the marketplace; governmental or industry policy changes). There are, however, risks with judgmental-only forecasts. Human error may make the analyst or manager more optimistic (or pessimistic) than actually warranted, trends or factors may be read into the data that are not actually there, or the effects of correlated variables may not be taken into account.

Three Ways to Integrate Judgment & Statistical Forecasting

There are three ways in which judgment can be integrated into statistical forecasting using time series data. First, judgment is key to determining which data are relevant to the model. Potential variables are virtually unlimited. However, a statistical model or analysis cannot take all these into account. Even if they could, spurious positive results would be seen due to the effects of probability alone. Therefore, it is important that expert judgments be used to reduce the inputs into the process. Another way in which judgment is important in forecasting is the determination of which statistical technique is most appropriate to analyze the data. There are a number of statistical techniques available for model building, analysis, and forecasting. Many of these are closely related. However, the worth of the end result depends heavily on correctly choosing the most appropriate analytical method. Third, expert judgments can be helpful in aiding the analyst to understand the situation and give insight into the parameters in which the data and subsequent analysis should be interpreted (e.g., "given the current economy and our market plan, we expect widget sales to rise by two percent over the next quarter"). Figure 1 shows some of the considerations for integrating judgment and statistical methods for forecasting.

Terms & Concepts

Business Cycle: A continually recurring variation in total economic activity. Such expansions or contractions of economic activity tend to occur across most sectors of the economy at the same time.

Correlation: The degree to which two events or variables are consistently related. Correlation may be positive (i.e., as the value of one variable increases the value of the other variable increases), negative (i.e., as the value of one variable increases the value of the other variable decreases), or zero (i.e., the values of the two variables are unrelated). Correlation does not imply causation.

Data: (sing. datum) In statistics, data are quantifiable observations or measurements that are used as the basis of scientific research.

Decomposition: The process of breaking down time series data into the component factors of trends, business cycles, seasonal fluctuations, and irregular or random fluctuations.

Forecasting: In business, forecasting is the science of estimating or predicting future trends. Forecasts are used to support managers in making decisions about many aspects of the business including buying, selling, production, and hiring.

Moving Average: A method used in forecasting in which the average value from previous time periods is used to forecast future time periods. The average is updated in each ensuing time period by including the new values not available in the previous average and dropping out the date from the earliest time periods.

Regression: A statistical technique used to develop a mathematical model for use in predicting one variable from the knowledge of another variable.

Sampling Error: An error that occurs in statistical analysis when the sample does not represent the population.

Seasonal Fluctuation: Changes in economic activity that occur in a fairly regular annual pattern. Seasonal fluctuations may be related to seasons of the year, the calendar, or holidays.

Time Series Data: Data gathered on a specific characteristic over a period of time. Time series data are used in business forecasting. To be useful, time series data must be collected at intervals of regular length.

Trend: The persistent, underlying direction in which something is moving in either the short, intermediate, or long term. Identification of a trend allows one to better plan to meet future needs.

Bibliography

Armstrong, J. S. & Collopy, F. (1998). Integration of statistical methods and judgment for time series forecasting: Principles from empirical research. In Wright, G. & Goodwin, P. (Eds.). Forecasting with Judgment. New York: John Wiley & Sons.

Black, K. (2006). Business statistics for contemporary decision making (4th ed.). New York: John Wiley & Sons.

Dauten, C. A. & Valentine, L. M. (1978). Business cycles and forecasting (5th ed.). Cincinnati: South-Western Publishing Co.

Jazayeri, S. M. T. & Yahyai, A. (2004). An analysis of seasonality of non-OPEC supply. Maritime Policy & Management, 30(3), 213-224. Retrieved May 23, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=14352172&site=ehost-live

Liu, C., Zhang, S. Y., & Jiang, Z. H. (2007). Models for predicting lumber grade yield using tree characteristics in black spruce. Forest Products Journal, 57(1/2), 60-66. Retrieved May 23, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=24255405&site=ehost-live

Ma, Y., Wang, N., Che, A., Huang, Y., & Xu, J. (2013). The bullwhip effect on product orders and inventory: a perspective of demand forecasting techniques. International Journal of Production Research, 57(1), 281-302. Retrieved November 15, 2013, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=83404185&site=ehost-live

Nelson, C. R. (1973). Applied time series analysis for managerial forecasting. San Francisco: Holden-Day.

Sperrazza, C. A., & McManus, D. J. (2013). Net job creation using time series forecasting. International Journal of Business Management & Economic Research, 4(3), 714720. Retrieved November 15, 2013, from EBSCO Online Database Business Source Complete. http://search.ebsco-host.com/login.aspx?direct=true&db=bth&AN=89462676&site=ehost-live

Voineagu, V., Pisica, S., & Caragea, N. (2012). Forecasting monthly unemployment by econometric smoothing techniques. Economic Computation & Economic Cybernetics Studies & Research, 46(3), 255-267. Retrieved November 15, 2013, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=85469422&site=ehost-live

Suggested Reading

Dishman, P. (2006). A typology of psychological biases in forecasting analysis. In Lawrence, K. D. & Geurts, M. D. (Eds.). Advances in Business and Management Forecasting. Boston: JAI Press.

Makridakis, S. & Wheelwright, S. C. (1982). Introduction to management forecasting: Status and needs. In Makridakis, S. & Wheelwright, S. C. (Eds.).The Handbook of Forecasting: A Manager's Guide. New York: John Wiley & Sons.

Morrell, J. (2001). How to forecast: A guide for business. Burlington, VT: Gower.

Nazem, S. M. (1988). Applied time series analysis for business and economic forecasting. New York: Marcel Dekker.

Pindyck, R. S. & Rubinfeld, D. L. (1998). Econometric models and economic forecasts. Boston: Irwin/McGraw-Hill.

Wynne, B. E. & Hall, D. A. (1982). Forecasting requirements for operations planning and control. In Makridakis, S. & Wheelwright, S. C. (Eds.).The Handbook of Forecasting: A Manager's Guide. New York: John Wiley & Sons.