MICROECONOMICS

Business Forecasting

Learning Objectives

After reading this chapter, you are expected to learn about:

• Know what is business forecasting

• Know methods of business forecasting

• understand to use appropriate method as per different situations

• Appreciate the use of forecasting in decision making.

Introduction

{{{2}}}

Historical Analogy

The important link between the art of management and forecasting is 'vision'. Long established business (probabilistic) forecasting techniques have been summarized as: Having a Time Horizon, e.g. Short Term / Medium Term / Long Term And follow one of the "Three Approaches" or methods: 1. Judgmental i.e. subjective assessments based on the opinions of experts, which can be further subdivided into: Personal Insight, Panel consensus, Market Surveys, Historical analogy and the Delphi method) 2. Projective techniques, (Time series, Constant series, Series with a trend, Seasonal series, Secondary characteristics, and Exponential smoothing) 3. Causal techniques, (Regression, Linear relationships etc.) What Can Be Forecast?

The External Environment e.g. National Economy Customer Behaviour Other producers Company distributors and the Internal Environment e.g. Marketing, Production, Financial Position, Executive goal-setting and planning etc.

Historical Analogies:

Historical analogies examine historical data from other businesses to find similarities. This is a qualitative forecast of possible futures through analogy to past data. Historical analogies are a variety of analogy often used by politicians and diplomats to explain or make a prediction about a current or future event based on events in the past. Historical analogy can be also described as an approach to sales forecasting in which the past sales results of a similar product are used to predict the likely sales of a similar new product. The past event is used as a source, while the present or future situation is the target of the analogy.

This theory is based on a more realistic assumption, i.e. that all business cycles are not uniform in amplitude or duration and as such the use of history is made not by projecting any fancied economic rhythm into the future, but by selecting some specific previous situation which has many earmarks of the present and concluding that what happened in that previous situation will happen in present also.

Steps in Forecasting Step 1 - Define the Problem What is needed? (E.g. future demand) Who will use it? (We will decide where to put a new plant) How does the forecast fit into the system? (At the beginning of the decision process) Step 2 - Gather Information & make a forecast. Do we have past history? Is it relevant? How much of it? Was there a strike at some point? When did we get a new major customer? What expert knowledge do we have? Someone who knows the business can tell us when data is irrelevant. Is all historical data equally valid? When is old data less valid? When is old data more valid? Observation and analysis of the past behavior is one of the most vital parts of forecasting. However, it should be carefully noted that though future maybe some sort of extension of the past, it may not be an exact replica. Changes in business and economic activity are caused by numerous forces or factors, which are often difficult to discover and measure. In historical analogy method, what is done is that a time series relating to the data in question is thoroughly scrutinized and from it such period is selected in which conditions were similar to those prevailing at the time of making the forecasts. The course which events took in the past under similar conditions is then studied which gives an idea of the likely course which the phenomenon in question would follow. Thus historical analogy, if not the best forecasting technique, is a vital tool for forecasting.

Field Surveys And Opinion Polls

Surveys are best suited for descriptive research. Possible relationships between the data and unknowns in the universe can be studied through surveys. Surveys may be either census (eg. opinion polls) or sample surveys (eg. field surveys). The method of data collection can be observation, interview, questionnaire (or opinionnaire), some projective techniques or case study methods.  Field Surveys: Field surveys may be conducted to obtain necessary information, which may constitute the basis for forecasting. Surveys may obtain qualitative as well as quantitative information. Field Surveys have larger samples and gather data from a relatively large number of cases at a particular time; it is cross-sectional. Companies undertake field surveys to learn about people’s knowledge, belief, preferences, satisfaction, etc and to measure these magnitudes in the general population. Questionnaire is the most commonly used instrument of primary data collection because of its flexibility.  Opinion Polls: Opinion polls are conducted related to a certain topic of interest or concern. They are typically carried out among a large, statistically significant population sample and are generally (though not always) administered through door-to-door visits or over the telephone. It is imperative that Indian agencies conducting opinion polls constantly innovate and improvise their research methodology for bridging the gap between forecasts and actual results. Examples of opinion polls:  In India, various websites conduct online opinion polls (www.indiatimes.com, www.mid-day.com, etc.).  Publication houses and magazines (like India Today, Business World, Outlook, etc.) also conduct opinion polls.  The winners for the various categories in the Economic Times Excellence Awards that were recently declared were selected through opinion polls.

Business Barometer

Of great assistance impractical forecasting is a series that can be used “Index” or “indicator: is also widely, though loosely used in business statistics, sometimes the terms used to mean simply an indicator of the present economics situations & sometimes it is used to designate an indicator of future conditions.

The following are some of the important series, which aid businessmen in forecasting: 1 Gross national product 2 Employment 3 Wholesale prices 4 Consumer Prices 5 Industrial production 6 Volume of bank deposits & currency outstanding 7 Disposable personal income 8 Departmental store sales 9 Stock prices 10 Bond yields

Several of the above are composite averages or total or indexes of these averages or totals. Analysis also should be made of the major components of these series Index Numbers relating to different activities in the field of production, trade, finance etc may also be combined into a general index of Business Activity. This general index refers to the general conditions of trade & industry. But the behavior of individual industries of trades might show a different trend from that of the Composite Business activity Index. Also general boom or depression may be reflected in a majority of separate industries & trades, yet some industries & trades might show quite contrary tendencies. Hence the study of general business conditions, as revealed by the Composite Business Index, should be supplemented by special studies of Individual business will guide the businessman as to whether the stocks of goods should be increased or released or whether to increase investment or not etc.1

Consumer Prices: The Consumer Price Index (CPI) is a measure of the average change over time in the prices paid by urban consumers for a market basket of consumer goods and services. As an economic indicator. As the most widely used measure of inflation, the CPI (consumer price Index) is an indicator of the effectiveness of government policy. In addition, business executives, labor leaders and other private citizens use the index as a guide in making economic decisions. As a deflator of other economic series. The CPI and its components are used to adjust other economic series for price change and to translate these series into inflation-free dollars. 2 Stock prices: The stock markets are supposed to be the barometers of the economy. The barometer is a nineteenth century device consisting of a vessel, which contains mercury. In the present day Indian stock market, if the vessel is likened to the Indian economy, it looks sturdy, while the mercury if is likened to the Sensex, it looks just like that — mercurial. Gross National Product (GNP): The total value of final goods and services produced in a year by domestically owned factors of production. Final goods are goods that are ultimately consumed rather than used in the production of another good. For example, a car sold to a consumer is a final good; the components such as tires sold to the car manufacturer are not - they are intermediate goods used to make the final good. Only final goods are included when measuring national income. If intermediate goods were included too, this would lead to double counting - for example, the value of the tires would be counted once when they are sold to the car manufacturer, and again when the car is sold to the consumer.

Exponential Smoothing

Exponential smoothing

 It is a type of more sophisticated moving average forecasting technique of time-series which weighs past data in an exponential manner so that the most recent data/observations carry more weight in the moving average.

 Unlike moving averages, which give equal weightages to all the past data, exponential smoothing assigns decreasing (exponentially decaying) weightages to the past data. This is obviously more reasonable since the older the demand the less it holds relevance for the future.

Smoothing Process

 The exponentially smoothed value at time period t is denoted by Fi.  The smoothing process begins by selecting a smoothing constant α (0 < α < 1).  When ‘n’ observations are involved in a time series, the forecast for a subsequent period (i.e., n+1) is found as the weighted average of the observed value of the series at period ‘n’ and the forecasted value for the same period, i.e., Fn+1 = αXn + (1 – α) Fn where, Fn+1 is the forecast value of the subsequent period . Xn is value of the last observation. Fn is the forecast value of the last period in the time series.

 The forecast values of Fn and all previous periods are computed in an identical manner. Specifically,

Ft = α . Xt + (1 – α) Ft-1

	where      = smoothing constant lying between 0 & 1

Xt = observation at the ith point of time Ft = forecast for current time period t.

         	          Ft-1 =  forecast for the preceding period t-1

-Which is suitable for t = 2 to n, but unsuitable for F1 as X0 and F0 are not known. To

 overcome this problem, F1 is taken as equal to X1 .ie. F1 = X1

- Thus for the second time period :F2 =  X2 + (1-) F1

 Selection of a Smoothing Constant

In fact, any value of α can be selected. The coefficients of a regression equation are selected in such way so as to minimize the sum of squared deviations between observed and forecast values. The same principle can also be used for selecting the smoothing constant.

Thus,

Σ (Xt – Ft )2

t=1

will be minimum for the selected value of α.

For determing the value of , the rule of the thumb followed is that when the magnitude of the random variations in the series is large, we should select a small  so that the smoothed value Ft will reflect Ft-1 to a greater extent than it reflects Xt. Thus when we have a moderately stable process, a larger  should be selected.

Advantages  This method gives more weight to the current data and it can update forecasts easily.  More useful in short term forecasting of inventories and studying seasonals and trends in sales. Disadvantages  It cannot make accurate forecasts when there is significant trend in the data.

Seasonal adjustment to Exponential Forecast:

 Where a seasonal pattern exists it may be desirable to seasonally adjust an exponentially smoothed forecast. The procedure is: 1. Deaseasonalise the actual demand 2. Compute a desesonalised forecast 3. Seasonalise (adjust) the forecast by multiplying by the seasonal index.

Extrapolation Technique

Extrapolation Techniques are "curve fitting" techniques in which the analyst plots total population data from past years, chooses a best-fitting curve for that data , and then extends that curve to project future values. It is a very simple procedure that is useful for small areas that do not have access to detailed data and for general projection figures for all areas. The extrapolation procedures used in forecasting may vary from simple extrapolation to the use of more complex mathematical equations and analog methods based on theory. The forecaster should extrapolate past and present conditions to obtain future conditions. Extrapolation is the simplest method of forecasting both long and short wave movement. However, there are few limitations from the Extrapolation Techniques. Total population figures for past years are being used to project future conditions; there is no inclusion of housing trends, economic changes, growth management, or any other external pressures on population in this technique. Any factors other than past population totals are external to the method. A further limitation to the model is that in Extrapolation Techniques the analyst assumes that past conditions will help to predict the future. Since there is simply no assurance that past trends will continue into the future, the technique does not provide any accurate information, in practice. Therefore, extrapolation techniques should be used carefully and with a full understanding of their limitations. The basic procedure for Extrapolation Techniques: 1. Acquire population data for past years 2. Plot data to determine the best fitting curve 3. Extend the curve into the future The common curves used in the extrapolation technique are linear, geometric, parabolic and exponential. Various methods used to extrapolate a particular value depending upon the requirement of the question and the nature of the given data, are: Graphical method Binomial Expansion method Newton’s method Lagrange’s method Parabolic Curve method Extrapolation Techniques can be used for various practical purposes, such as, predicting the population of a country, etc. It is also the most practical prognostic technique used by the tropical meteorologist who extrapolate the past movement of the synoptic features on his or her chart into the future. This technique is basically used when time, data, information, and expertise in more advanced techniques are limited.

Lead Lag Analysis

Leading indicators are industrial and economic statistics from which an indication of the value or direction of another variable (for example, a sales forecast) might be obtained. They are called "leading" because their direction or magnitude historically "leads" the focal variable. For example, we may find that money supply indicates (leads) the future level of consumer spending. This technique is most useful for identifying the turning points or cyclic nature of a variable. PROCEDURE: Start with common sense in searching out likely leading indicators. You might start by considering such leading indicators as GNP, interest rates, capital spending, inflation, and/or the unemployment rate - looking for those that appear to have a meaningful "leading" relationship with the variable (e.g. sales) you are forecasting. For example, a very good leading indicator of computer sales might be corporate profits whereas the unemployment rate may not. In addition to data sources available within many firms which may serve as useful leading indicators, there are several published sources of statistics that are frequently used by forecasters and market analysts, including: 1. Economic indicators (council of economic advisers) 2. Federal reserve bulletin 3. Handbook of basic economic statistics 4. Survey of current business 5. Survey of industrial purchasing power

Having identified a likely leading indicator, then you must experiment to find the degree of lead-lag (for example, how many months or years) in the indicator-sales relationship. A simple way to identify whether a relationship exists and the degree of lead-lag is to graph the values of both variables over a number of time periods and then line up their peaks and troughs. The amount of adjustment need to match the profiles depicted on the graph by both sets of data indicates the length of time the leading indicator "leads" the variable of interest.

COMMENTS: • Leading indicators are many times difficult to apply in practice because the lead-lag relationship tends to be quite volatile. • Turning point may be difficult to identify quickly; for example, does a decrease in a leading indicator represent a turning point or just a temporary fall? • Also, regresssion analysis can be used to identify the nature of the relationship. Simply try running a series of regresssions of the variable of interest (i.e. sales) on the indicator variable at varying time periods. The appropriate lead-lag period is the one for which the regression produces the highest value of R2 (a statistic provided in the output of the regression that indicates the strength of the relationship between the independent and dependent variables).

Regression Analysis

The statistical tool with the help of which we are in a position to estimate or predict the unknown values of one variable from known values of another variable is called regression. Regression thus reveals average relationship between two variables and this makes possible estimation or prediction.

In regression analysis there are two types of variables. whose value is influenced or is to be predicted is called dependent variable and the variable which influence the value or is used for prediction, is called independent variable. Regression analysis is a statistical technique for quantifying the relationship between variables.

Simple and Multiple The regression analysis confine to the study of only two variables at a time is termed as simple regression. On the other hand regression analysis for studying more than two variables at a time is known as multiple regression. In this one variable is the dependent variable, the remaining variables are dependent ones. For Example :- Turnover may depend on advertising expenditure and the income of the people.

Linear and Non linear

If the given bivariate data are plotted on a graph the points so obtained will form a path called the 'curve of regression'. if this path is a straight line, it is linear regression. The regression is termed as non-linear if the curve of regression is not a straight line.

Total and Partial

In the case of total relationship all the important variables are considered. thay take the form of multiple relationship because most economical and business phenomenon are affected by multiplicity of causes. In the case of partial relationship one or more relations are considered, but not all, thus excluding the influence of those not found relevant for the given purpose. For Example: considering the example of sales (y) influenced by , advertising expenditure (x), income of people (i) and price of products (p). as the three important independent variables, the Total relationship is expressed as : y = f (x,i andp) Partial relationship is expressed as : y = f (x but not i&p)

         				             y =f (i but not x&p)

y = f (p but not i&x)

Utility of regression analysis:

• It helps in establishing the functional relationship between two or more variables; hence it can be used for various advanced analytical purposes. • Since most of the problems of economic analysis are based on cause and effect relationship, the regression analysis is a highly valuable tool in economics and business research. • It is very useful for prediction purposes or estimation of future production, prices, sales, investments, income, profits and population. Once a functional relationship is known, the value of the dependent variable can be predicted from the given value of the independent variables. • It is widely used in statistical estimation of demand curves, supply curves, production functions, cost functions, consumption functions etc. Regression relationship between personal consumption expenditure and disposable personal income are regarded very important for various policy matters.

There are two dangers in using regression analysis for forecasting: 1) There is a possibility of a mechanistic approach accepting with little question the relationship, which the calculations reveal. There are many possibilities for spurious correlation among time series as many series move together over time even where there is no conceivable connection between them. 2) There is a risk that the estimated regression is false.

Time Series Analysis

Time series analysis is one quantitative method we used to determine patterns in data collected over time  .In other words the first step in making estimates for the future consists of gathering information from the past. In this connection one usually deals with statistical data which are collected, observed or recorded at successive interval of time .Such data are generally referred to as “ TIME SERIES ”.Thus when we observe numerical data at different points of time the set of observation is known as “Time series “ For example if we observe production ,sales ,population ,imports ,exports ,etc at different points of time, say , over the last five or ten years the set of observation  formed shall constitute “time series”. Time series techniques are the most popular quantitative method. Two major types of time-series methods are moving average and exponential smoothing

UTILITY OF TIME SERIES ANALYSIS:

The Analysis of time series is of great significance not only to the economists and businessman but also to the scientists , astronomist ,geologist ,sociologist ,biologist ,research worker ,etc for following reasons:

1) It helps in understanding past behaviour 2) It helps in planning future operations 3) It helps in evaluating current accomplishments 4) It facilitates comparison

COMPONENTS OF TIME SERIES:

1) Secular trend: With the first type of change , secular trend, the value of variable tends to increase or decrease over a long period of time. The steady increase in the cost of living recorded by the consumer price index is an example of secular trend 2) Cyclical Fluctuation: The most common example of cyclical is the business cycle . 3) Seasonal Variation: The third kind of change in time is seasonal variation . as we might expect from the name, seasonal variation involves pattern of change within a year that tends to be repeated from year to year .For Eg A physician can expect a substantial increase in number of flu cases every winter. 4) Irregular Variation: it’s the fourth type of change in time series analysis .In many situation the value of a variable may be completely unpredictable, changing in a random manner. Irregular variation describes such movements. Eg. The Iraqi situation in 1990 on gasoline prices in the US .

Time series analysis example in real world: 1) Time series analysis is an integral part of financial analysis. The topic is interesting and useful, with applications to the prediction of interest rates, foreign currency risk, stock market volatility, and the like. There are many varieties of econometric and multi-variate techniques.

In order to study inflation in India, a macroeconomic multivariate model was developed using ANN (artificial neural network) methodology. The data consisted of twenty-one years of average monthly values of five significant variables. These five variables were: real effective exchange rate (REER), money supply (M3), index of industrial production (IIP), food arrivals (F), and the wholesale price index (WPI). It was found that the usual statistical approach was reasonable only when making a univariate time series study of M3 and WPI. All other variables had nonlinear features. The multivariate ANN model was the best in making very accurate (mean error 1-2%) short-term forecasts and long-term trends (mean error up to 4%). The multivariate model also helped generate alternative scenarios. For example, it was possible to forecast the change in inflation rate and industrial activity by giving impulses to REER and M3.

2) Land-Cover Change in China using Time Series Analysis, 1982 - 1999: A more overall and newer understanding of China's land-cover change dynamic can be achieved by a longer time series analysis (TSA). The results have proved that PCA/TSA is a very effective method to identify both macro and micro factors driving the change of NDVI (normalized difference vegetation index) Especially the thesis to pave a way to detect the impacts of extreme physical accidents and human-induced activities upon the NDVI change.

Econometric Models

The name ‘Econometrics’ was introduced in 1926 by a Norwegian economist and statistician, Ragnar Frisch’. To provide better guidance for economic policy-making, we need to know the quantitative relationships between the different economic variables. If empirical data verifies the relationship proposed by theory, we may accept it: otherwise we must reject the relationship. For example: if investment is proposed to be increased by 15%, we must know by how much national income will be expected to increase.

“Econometrics is the science, which combines economic theory with economic statistics and tries by mathematical and statistical methods to investigate the empirical support of the general schematic law established by economic theory.”

Goals of Econometrics:

Econometrics helps us to achieve the following three goals: 1. Judge the validity of the economic theories; 2. Supply the numerical estimates of the coefficients of the economic relationships which may be then used for sound economic polices; and 3. Forecast the future values of the economic magnitudes with certain degree of probability. Theoretical econometrics deals with the development of the appropriate methods for measuring economic relationships described by econometric models. These methods can be classified into two groups: • Single-equation techniques which are applied to one relationship at a time; and • Simultaneous-equation techniques, which are applied to all the relationships of the models simultaneously. Also it is the concern of theoretical econometrics to spell out the assumption of these methods, their properties, and what happens to these properties when one or more of the assumptions of the methods are not fulfilled.

Applied econometrics describes the practical value of econometric research. It deals with the application of econometric technique developed in theoretical econometrics to different field of economic theory for its verification and forecasting. Presently more and more empirical studies in the fields of market demand and supply, production function, cost function, consumption and investment function, are being carried out with the help of econometrics. The applied econometrics has made it possible to obtain numerical result from these studies which are of great importance to our planner.

Methodology of Econometrics:

1. Specify mathematical equations to describe the relationships between economic variables as proposed by economic theory. 2. Design methods and procedures based on statistical theory to obtain representative samples from the real world. 3. Development of methods of estimating the parameters of the specified relationships 4. Development of statistical methods of testing the validity of theory by using estimated parameters obtained. 5. Development of methods of making economic forecasts or policy implications based on estimated parameters in case the theory stands out the test evolved.

For Example: The chain of events, as shown by the activities numbered 1 through 5 in figure is a diagrammatic model of the operation of a private enterprise economy. To represent magnitudes conveniently, a mathematical model is to be employed, which basically is a set of equations that describe various relationships between variables.

In activity number 4, if W is the amount of wages and salaries earned by households, and C is household expenditures on clothing, then the equation C = .12W states that households spend 12 percent of their wages and salaries on clothing. Each of the activities pictured in figure can be represented in the form of an equation. Doing so may take a blend of economic theory, physical and institutional realities, and mathematical sophistication, but once done, with real data, the result would be a "quantitative economic model" or an “econometric model”