4.3 Sales forecasting (HL) Flashcards
What does the sales forecasting imply and what is the purpose?
Quantitative methods, which rely on the past data in an attempt to predict the future and estimate:
● Future sales levels and
●Trends over a specified period of time.
v Reduces uncertainty,
v helps in management of stock and cash
flow, and
v ensures better planning.
What is a trend?
A visible pattern noted after inputting the past sales data. This may indicate the rise and fall of sales over a given period.
Seasonal fluctuations
Changes in demand because of the varying seasons in the year.
Seasonal variations are usually repeated and occur within one year or less.
Cyclical fluctuations
Variations tied to the business cycle in an economy.
For example, sales could be on the rise during the growth phase but declining during a recession.
Cyclical variations can extend for more than one year.
The are much more unpredictable than seasonal variation.
Random fluctuations
(Unpredictable) changes or fluctuations that stand out from a given trend.
For example, there may be a sudden increase in the demand for ice-cream during a rare warm day in winter.
Moving average
One of the methods to find/understand a trend. More accurate and complex method than simply predicting future sales from actual sales data.
● helps to smooth out fluctuations from sales data.
Extrapolation? Is it reliable?
Once the trend line has been drawn, this line can be extended using a “line of best fit”.
The problem with extrapolation is that you have nothing to check how accurate your model is outside the range of your (past) data. … Because there are no (future) data to support an extrapolation, one cannot know whether the model is accurate or not.
How will be a variation in each period calculated?
Actual past sales (moving average) minus the trend.
How will be cyclical variation calculated?
We take variations of several years (e.g.10) calculate average.
How will be seasonal variation calculated?
Normally, we divide a year by seasons, the weeks/months/quarters/other, depending on business model; Calculate variations of these seasons within a year; Compare seasonal variations of 2+ years and calculate average.
How to use the average variations?
We add them to the extrapolated trend value.
What is meant by Centring?
Use of 4-year-moving center 2 times (in total 8), the second time shifting by one period, to find a center
Limitations of sales forecasting?
● May be time-consuming
● Ignores qualitative external factors (political, social, economic etc)
How we use seasonal variations?
They are used to generate more accurate prediction of quarterly/seasonal sales.
