**4.3 Sales Forecasting**

(HL only)

(HL only)

**KEY LEARNING OUTCOMES:**

Calculate three period and four-period moving average

Identify sales trends and determine forecasts using given sales data

Evaluate the benefits and limitations of sales forecasting

Calculate three period and four-period moving average

Identify sales trends and determine forecasts using given sales data

Evaluate the benefits and limitations of sales forecasting

**Key concept: Strategy**

**Sales Forecasting**

Potential Benefits:

If marketing managers were able to predict the future sales accurately, the risks of business operations and business strategic decisions would be much reduced.

The operations department would know how many units to produce and what quantity of materials to order and how much stock level to hold.

The marketing department would be aware of how many products to distribute and whether changes to the marketing mix are needed.

Human resources workforce plan would be more accurate.

Finance could plan cash flows with much greater accurate amd make accurate profit forecasts.

Strategic decision-making such as developing new products or entering new markets would become much better informed.

In reality, such precision in forecasting is impossible to achieve, because of external factors that can influence sales performance.

Market forecasts form an essential part of the market planning process and of the screening process before new products are launched on to the market. These forecasts will be based on

market research data

, gained from both

primary

and

secondary sources

. A common way of assessing future demand for a product yet to be fully launched is to use test marketing in one particular area.

For existing products sales forecasts are commonly based on past sales data.

Quantitative sales forecasting methods - time-series analysis

This method of sales forecasting is based entirely on

past sales data

. Sales records are kept over time and , when they are presented in chronological order, they are referred to as "time series'.

Extrapolation

Extrapolation involves basing future predictions on past results. When actual results are plotted on a time-series graph, the line can be extended, or

extrapolated

, into the future along the trend of the past data. This simple method assumes that sales patterns are stable and will remain so in the future. It is ineffective when this is not true.

Moving Averages

This method is more complex than simple graphical extrapolation. It allows the identification of underlying factors that are expected to influence future sales.These are the trend,

seasonal variations

,

cyclical variations

and

random variations

.

Seasonal variations

: regular and repeated variations that occur in sales data withing a period of 12 months or less.

Examples - clothing winter and summer, Flights (holidays), textbooks.

Cyclical variations

: variations in sales occurring over periods of time of much more than a year - they are related to the business cycle.

Random variations

: may occur at any time and will cause unusual and unpredictable sales figures, eg exceptionally poor weather or negative public image following a high profile product failure.

The moving average is used to analyze these. This technique "smooths out" the fluctuations in time-series data and allows managers to identify the trend more easily.

CYCLICAL EXAMPLE

TREND, SEASONAL AND RANDOM VARIATION EXAMPLES

Trend: underlying movement of the data in a time series.

Calculating Moving Averages - Simple Example three-year moving average

Yearly sales of a calculator manufacturer:

Year

Sales (US$000)

1

400

2

600

3

800

4

650

5

700

6

850

7

950

8

1200

Steps calculate a three year moving average:

1. Calculate the mean sales for the first 3 years, then the second three sets and so on.

400 + 600 + 800

3

= 600

Years 1, 2, 3:

Years 2, 3, 4:

Years 3, 4, 5:

Years 4, 5, 6:

Years 5, 6, 7:

600 + 800 +650

3

800 + 650 + 700

3

650 + 700 + 850

3

850 + 950 + 1200

3

= 683.333

= 716.667

= 733.333

= 1,000.000

Remember in US$ 000

Sales revenue with the three-year moving average (trend): (in US$ 000)

Year 1 2 3 4 5 6 7 8

Sales 400 600 800 650 700 850 950 1200

Trend 600 683.333 716.667 733.333 833.333 1000

Years 6, 7, 8:

700 + 850 +950

3

= 833.333

Now what:

1. Plot the actual

sales revenue

and the

trend line (moving average)

on a time series graph.

Title

Sales

Years

2.

Extrapolate

- extend the trend line to predict future sales using a

line of best fit

.

PRACTICE: Do 3-period moving average practice questions A and B.

The benefits and limitations of sales forecasting

Benefits:

Improved working capital and cash flow

- by taking into consideration cyclical and seasonal variation factors, financial managers can better plan to improve the liquidity position of a business.

Increased efficiency and stock control

- sales forecasting greatly assists the production department in knowing the number of goods to produce and in planning for the amount of stock required in the future.

Better workforce planning

- accurate sales forecasting can help the human resources department in succession planning regarding the number of staff required in the future.

Improved budgeting

- will start the budgeting process.

Forecast costs and profits

Limitations:

Uncertain future demand and inaccuracy of predictions

- the business environment is constantly changing and this will impact sales.

Change in costs affecting price

- if the cost changes this most likley impact price which will affect sales forecast

Complex moving average calculations

- difficult and time-consuming to calculate

External influences

- the external environment causes change that may not be predictable

To what extent does knowing assist us in predicting?

How do we know that our predictions are reliable?

"

Prediction is very difficult, especially if it is about the future

" Neils Bohr, Nobel Laureate

This statement highlights some of the problems of using mathematics in forecasting.

Do you think there is any point in

managers forecasting future sales?

**Calculating a four-year moving average**

Calculating the four-year moving average is a bit more complex than calculating a three-year moving average. Four-year moving average is the most widely used technique as it is often used when forecasting from quarterly data. Much business data is released quarterly.

In this case, it makes use of

centering

. This involves the use of a four-year and a eight year moving total to establish a mid-point. This is because if a four-quarter moving total was divided by four in order to calculate the average it would not lie alongside any one quarter. It would not make sense to have a result which did not belong to any time period. This is overcome by

centering

the average so that it lies alongside one actual quarter. This is done by adding two four-quarter moving totals together. This is divided by eight to give the moving average.

TOK

Four-year moving average using the previous example:

Yearly sales of a calculator manufacturer:

Year

Sales (US$000)

1

400

2

600

3

800

4

650

5

700

6

850

7

950

8

1200

Steps to calculate a four-year moving average:

1. Four-year moving total

Sum the sales of year 1, 2, 3 and 4. (400 + 600 + 800 + 650 = 2450)

Sum the sales of year 2, 3, 4 and 5. (600 + 800 + 650 + 700 = 2750)

.... (continue for years 3-6, 4-7 and 5-8)

2. Calculate eight-year moving total

Add 2 sets of 4 year moving totals = 2450 + 2750 = 5200

.... continue

3. Calculate the four-year centered moving average.

Divide the eight-year moving total by 8. 5200/8 = 650. Place this in the line where year 3 (or Quarter 3 is positioned).Continue using same approach.

Complete the four-year moving average in table format.

Practice 4-year moving average:

Do C on 4.3 exercises.

**Calculating Variations**

Variation - the difference between actual sales and trend values (moving average)

Continuing with the example:

Year

Sales (US$ 000)

Trend - 4-year moving average

Variation in each year

1 2 3 4 5 6 7 8

400 600 800 650 700 850 950 1200

650 718.75 768.75 856.25

150 -68.75 -68.75 -6.25

1. Calculate the

variation

: Sales - Trend. (800-650 = 150) ...

Average cyclical variation:

This is calculated as the sum of the variations over the period divided by the number of years within the period.

sum of variations

number of years

Example:

150000 + (-68750) + (-68750) + (-6250)

4

= 1,562.50

BE CAREFUL WITH ADDING + AND - NUMBERS. THE SIGN MUST BE CORRECT

=

IF you were reading off the trend line for year 9 you would

adjust

it for the average seasonal or cyclical variation. If it was 1350000 then it would be 1350000 - 1562.50 = 1348437.5.

DO:

MORE Practice -

Four Period Moving Average Questions

Apple iPad sales + Chill-Out and Genius

Headline: October 16th, 2015

Market Trends

Bloomberg Business

"US Retail Sales Rose Less than Forecast."

Sales forecasting

- is where a business uses data and other information to predict future sales.

It is impossible to predict the future with complete certainty but most business will attempt to forecast future sales - if only for a short time into the future.

A moving average is where the mean average in a set of data is continuously recalculated over time to establish a trend in the data.

Three Point Moving Average - Steps

Calculate the three- point moving total

- add up the sales revenue for the first three years and repeat.

Calculate the three- point moving average

- calculate the average for each three year period in the data.

Establish the trend:

Plot the sales data

Plot the 3 year moving average

Extrapolate the trend

Set up a table with the following headings:

Year Sales Revenue 3-year total 3-year moving average

Four-period moving average - STEPS

Set up a table with the following headings:

Year - Quarter - Sales Revenue - 4 Quarter moving total - Quarterly moving average - Seasonal Variation

Calculate the four quarter moving total for each quarter.

Calculate the eight quarter moving total for each possible quarter.

Calculate the four-quarter moving average - for each possible quarter

the 8 quarter moving total

8

Calculate the seasonal variation -

this is the way data changes in a repeatable and predictable way during the year.

Actual sales revenue - quarterly moving average = seasonal variation

Calculate the average seasonal variation for each quarter.

This is a way of smoothing out the actual seasonal variations in the data to give a seasonal variation that makes it easier to forecast sales. It is calculated by adding together seasonal variations in a quarter and dividing by the number of times there is data for this quarter.

Tip: Try skipping a line between numbers will help you to visualize the centering