Introduction to

Forecasting
Prerequisites of a
Good Forecast
Forecasting
Forecasting is the process of making predictions of the future based on past and present data and
most commonly by analysis of trends. A commonplace example might be estimation of some
variable of interest at some specified future date. Prediction is a similar, but more general term.
Both might refer to formal statistical methods employing time series, cross-
sectional or longitudinaldata, or alternatively to less formal judgmental methods. Usage can
differ between areas of application: for example, in hydrology the terms "forecast" and
"forecasting" are sometimes reserved for estimates of values at certain specific futuretimes, while
the term "prediction" is used for more general estimates, such as the number of times floods will
occur over a long period.
Risk and uncertainty are central to forecasting and prediction; it is generally considered good
practice to indicate the degree of uncertainty attaching to forecasts. In any case, the data must be
up to date in order for the forecast to be as accurate as possible. In some cases the data used to
predict the variable of interest is itself forecast.

Categories of forecasting methods

Qualitative vs. quantitative methods

Qualitative forecasting techniques are subjective, based on the opinion and judgment of
consumers and experts; they are appropriate when past data are not available. They are usually
applied to intermediate- or long-range decisions. Examples of qualitative forecasting methods are
informed opinion and judgment, the Delphi method, market research, and historical life-cycle
analogy.
Quantitative forecasting models are used to forecast future data as a function of past data. They
are appropriate to use when past numerical data is available and when it is reasonable to assume
that some of the patterns in the data are expected to continue into the future. These methods are
usually applied to short- or intermediate-range decisions. Examples of quantitative forecasting
methods are last period demand, simple and weighted N-Period moving averages,
simple exponential smoothing, poisson process model based forecasting [2] and multiplicative
seasonal indexes. Previous research shows that different methods may lead to different level of
forecasting accuracy. For example, GMDH neural network was found to have better forecasting
performance than the classical forecasting algorithms such as Single Exponential Smooth,
Double Exponential Smooth, ARIMA and back-propagation neural network.
Average approach
In this approach, the predictions of all future values are equal to the mean of the past data. This
approach can be used with any sort of data where past data is available. In time series notation:

where
is the past data.

Although the time series notation has been used here, the average approach can also be used for
cross-sectional data (when we are predicting unobserved values; values that are not included in
the data set). Then, the prediction for unobserved values is the average of the observed values.

Naïve approach

Naïve forecasts are the most cost-effective forecasting model, and provide a benchmark against
which more sophisticated models can be compared. This forecasting method is only suitable for
time series data. Using the naïve approach, forecasts are produced that are equal to the last
observed value. This method works quite well for economic and financial time series, which
often have patterns that are difficult to reliably and accurately predict. If the time series is
believed to have seasonality, the seasonal naïve approach may be more appropriate where the
forecasts are equal to the value from last season. In time series notation:

Drift method
A variation on the naïve method is to allow the forecasts to increase or decrease over time, where
the amount of change over time (called the drift) is set to be the average change seen in the
historical data. So the forecast for time
is given by

This is equivalent to drawing a line between the first and last observation, and extrapolating it
into the future.
Seasonal naïve approach
The seasonal naïve method accounts for seasonality by setting each prediction to be equal to the
last observed value of the same season. For example, the prediction value for all subsequent
months of April will be equal to the previous value observed for April. The forecast for time is

Where  =seasonal period and  is the smallest integer greater than 

The seasonal naïve method is particularly useful for data that has a very high level of seasonality.
Time series methods
Time Series methods use historical data as the basis of estimating future outcomes. They are
based on the assumption that past demand history is a good indicator of future demand.

 Moving average
 Weighted moving average
 Exponential smoothing
 Autoregressive moving average (ARMA)(forecasts depend on past values of the variable
being forecast and on past prediction errors)
 Autoregressive integrated moving average (ARIMA) (ARMA on the period-to-period
change in the forecast variable)
e.g. Box–Jenkins
Seasonal ARIMA or SARIMA or ARIMARCH,
 Extrapolation
 Linear prediction
 Trend estimation (predicting the variable as a linear or polynomial function of time)
 Growth curve (statistics)
 Recurrent neural network
Causal / econometric forecasting methods
Some forecasting methods try to identify the underlying factors that might influence the variable
that is being forecast. For example, including information about climate patterns might improve
the ability of a model to predict umbrella sales. Forecasting models often take account of regular
seasonal variations. In addition to climate, such variations can also be due to holidays and
customs: for example, one might predict that sales of college football apparel will be higher
during the football season than during the off season.
Several informal methods used in causal forecasting do not rely solely on the output of
mathematical algorithms, but instead use the judgment of the forecaster. Some forecasts take
account of past relationships between variables: if one variable has, for example, been
approximately linearly related to another for a long period of time, it may be appropriate to
extrapolate such a relationship into the future, without necessarily understanding the reasons for
the relationship.
Causal methods include:

 Regression analysis includes a large group of methods for predicting future values of a

variable using information about other variables. These methods include both parametric (linear
or non-linear) and non-parametric techniques.
 Autoregressive moving average with exogenous inputs (ARMAX)[6]
Quantitative forecasting models are often judged against each other by comparing their in-
sample or out-of-sample mean square error, although some researchers have advised against this.
Different forecasting approaches have different levels of accuracy. For example, it was found in
one context that GMDH has higher forecasting accuracy than traditional ARIMA 
Judgmental methods
Judgmental methods include:

 Composite forecasts
 Cooke's method
 Delphi method
 Forecast by analogy
 Scenario building
 Statistical surveys
 Technology forecasting
Artificial intelligence methods
 Artificial neural networks
 Group method of data handling
 Support vector machines

Often these are done today by specialized programs loosely labeled

 Data mining
 Machine learning
 Pattern recognition
Other methods
 Simulation
 Prediction market
 Probabilistic forecasting and Ensemble forecasting
A good forecast should satisfy the following criteria:

 Time frame: The first factor that can influence the choice of forecasting is the time
frame of the forecasting situation. Forecasts are generally for points in time that may be a
number of days, weeks, months, quarters, or years in the future. This length of time is called the
time frame or time horizon. The length of the time frame is usually categorized as Immediate,
Short term, Medium or Long term. In general, the length of the time frame will influence the
choice of the forecasting technique. Typically a longer time frame makes accurate forecasting
more difficult with qualitative forecasting techniques becoming more useful as the time frame
lengthens.
 Pattern of the data: The pattern of the data must also be considered when choosing a
forecasting model. The components present i.e. trend, cycle, seasonal or some combination of
these will help determine the forecasting model that will be used. Thus it is extremely important
to identify the existing data pattern.
 Cost/Economy of forecasting: Though the firm is interested in accurate forecasts, the
benefits of accurate results must be weighed against the cost of the method. While choosing a
forecasting technique, several costs are relevant. First, the cost of developing the model must be
considered. Second the cost of storing the necessary data must be considered. Some forecasting
methods require the storage of a relatively small amount of data, while other methods require the
storage of large amounts of data. Last, the cost of the actual operation of the
forecastingtechnique is obviously very important. Some forecasting methods are operationally
simple, while others are very complex. The degree of complexity can have a definite influence
on the total cost of forecasting.
 Accuracy desired: Accuracy in forecasting is very important. The previous method must
be checked for want of accuracy by observing that the predictions made in the past are accurate
or not. The accuracy of past forecasting can be checked against present performance and of
present forecasts against future performance. In some situations a forecast that is in error by as
much as 20% may be acceptable. In other situations a forecast that is in error by 1% might be
disastrous. The accuracy that can be obtained using any particular forecasting method is always
an important consideration.
 Availability of data: Immediate availability of data is an important requirement and the
method employed should be able to produce good results quickly. The technique which takes
much time to produce useful information is of no use. Historical data on the variable of interest
are used when quantitative forecasting methods are employed. The availability of this
information is a factor that may determine the forecasting method to be used. Since various
forecasting methods require different amounts of historical data, the quantity of data available is
important. Beyond this, the accuracy and the timeliness of the data that are available must be
examined, since the use of inaccurate or outdated historical data will obviously yield inaccurate
predictions. If the needed historical data are not available, special data-collection procedures may
be necessary.
 Plausibility/Ease of operation and understanding: The ease with the forecasting
method is operated and understood is important. Management must be able to understand and
have confidence in the technique used. It has to understand clearly how the estimate was made.
Mathematical and statistical techniques should be avoided if the management cannot understand
what the forecaster does. Managers are held responsible for the decisions they make and if they
are to be expected to base their decisions on predictions, they must be able to understand the
techniques used to obtain these predictions. A manager simply will not have confidence in the
predictions obtained from a forecasting technique he or she does not understand, and if the
manager does not have confidence in these predictions, they will not be used in the decision-
making process. Thus, the managers understanding of the forecasting system is of crucial
importance.
 Durability: The forecast should be durable and should not be changed frequently. The
durability of the forecasts depends on the simplicity and ease of comprehension as well as on
continuous link between the past and the present and between present and the future.
 Flexibility: The technique used in forecasting must be able to accommodate and absorb
frequent changes occurring in the economy.