Chapter 1

Forecasting
 • Apply high-low method to estimate the fixed and
variable element of costing.
• Explain the concept of correlation coefficient and
the coefficient of determination.
• Calculate and interpret correlation coefficient and
the coefficient of determination.
Objectives • Explain the principles of time series analysis
(cyclical, trend, seasonal variation and random
elements).
• Calculate moving averages.
• Describe the product life cycle and explain its
importance in forecasting.
 1. Correlation, correlation coefficient and
the coefficient of determination.
2. Linear relationship between dependent
Contents and independent variables.
3. High-low method.
4. Time series analysis.
5. The product life cycle.
 Correlation, correlation coefficient and the
coefficient of determination

• Correlation
• Two variables are
said to be correlated
if a change in the
value of one variable
is accompanied by a
change in the value
- có mqh tương quan cùng tăng nhưng ko có mqh nhân quả
of another variable.
Correlation, correlation
coefficient and the coefficient
of determination
• Correlation between
two related variables
could be shown by a
scattergraph or
scatter diagram
Correlation, correlation coefficient and
the coefficient of determination
• Degrees of correlation
• perfectly correlated ( r = 1) 100%

• partly correlated 0< r < 1

• Or uncorrelated r = 0 => ko tương quan

- hệ số tương quan: r (-1 < r < 1)
- outliner: dị biệt
1. mẫu nhỏ ko mang tính đại diện
2. có mẫu dị biệt
• Correlation can be positive or negative 3.
positive: hiện tượng cùng tăng => tương quan đồng biến => 0 < r < 1
negative: tương quan nghịch biến => -1 < r < 0
đối với khảo sát mẫu càng rộng thì càng đáng tin cậy
chiều cao và cân nặng => positive đồng biến
r = 1 & r = -1 khi các dấu chấm cùng nằm trên 1 đường thẳng
Correlation, correlation coefficient and
the coefficient of determination
• Positive correlation: low values of one variable are associated
with low values of the other, and high values of one variable are
associated with high values of the other.
• Negative correlation: low values of one variable are
associated with high values of the other, and high values of one
variable with low values of the other.
Correlation, correlation coefficient and
the coefficient of determination
• Correlation coefficient (r): degree of linear correlation between
two variables.
• -1 ≤ r ≥ 1
• Formula: n: số lượng quan sát (sample size)
Correlation, correlation coefficient and
the coefficient of determination
• Coefficient of determination (r2):
strength of the linear association
between variables.
• If r = 0.9 → r2 = 0.81 then we could
conclude that 81% of the dependent
variable (y) can be explained by the
independent variable/variables (x),
leaving 19% of y to be explained by
other factors.
Correlation, correlation coefficient and
the coefficient of determination
• Correlation ≠ Causation
 Linear relationship between dependent
and independent variables
• Correlation does not
offer any method of
forecasting values for y
for a given x.
• The key is to determine
the line of best fit or to
determine the equation
of a straight line (Y = a
+ bX) which is a good fit
for the available data.
- khoảng cách từng điểm tới đường thẳng bình phương, tổng bình phương là điểm nhỏ nhất
- Least Squared Regression: pt đường thẳng nhỏ nhất
Linear relationship between dependent
and independent variables
• Linear regression analysis (or the least squares method) is one
technique for estimating a line of best fit.
Linear relationship between dependent
and independent variables
X
Y = ax + b
X X X
Total mixed cost Y

X X X
X
X X
X X
X X
X
X
X

Unit-of-production X
Linear relationship between dependent
and independent variables
- It is our goal to set out the line y = ax + b
- Therefore, we have to solve the following simultaneous linear
equations:
∑xy = a ∑x² + b ∑x (1) a trên slide là b trên sách
∑y = a ∑x + n b (2)
y: mixed cost
x: unit of production
a: unit variable cost
b: total fixed cost
n: observation times
Linear relationship between
dependent and independent
variables
• Factors affecting the reliability of regression analysis:
1. It assumes a linear relationship exists between
the two variables.
2. Y might depend on several other variables, not
just X. giả định
3. It assumes that what has happened in the past
will provide a reliable guide to the future.
4. Problem with extrapolation. dùng x ko thuộc khoảng data
5. The amount of data available. (The more the
better!) cỡ mẫu
6. The reliability of collected data.
3. những chuyện xảy ra trong quá khứ có khả năng xảy ra trong tương lai
High-low method pp cực đại - cực tiểu
• The high-low method is a simple forecasting technique. The key
principle is that the “best line” would cross the highest volume
of activity and the lowest volume of activity.
• This method may be applied to annual sales figures or any
other activity as well as costs.
High-low method

Advantages: Disadvantages

• It is easy to use and understand. • It uses two extreme data points

• It needs just two activity levels. which may not be representative
of normal conditions.
• Using only two points to
determine a formula may mean
that the formula is not very
accurate.
 A time series is a series of figures or
values recorded over time.

Time series
analysis
There are four components of a time
series:
Seasonal Cyclical Random
Trend
variations variations variations.
 Time series
• Trend: the underlying long-term analysis
movement over time in the
values of the data recorded.
• Trend of given data could be
found by the use of moving
averages.
Time series analysis

• Seasonal variations: short-term

fluctuations due to different
circumstances which affect results at
different times
 • Cyclical variations:
fluctuations which take place
Time series over a longer time period
analysis
Time series analysis

• Random variations:
fluctuations caused
by unforeseen
circumstances.
• Moving averages: remove
seasonal (or cyclical) variations Time series analysis
from a time series by a process
of averaging so as to leave a set
of figures representing the trend.
• Apply in case of:
• Odd number of periods
• Even number of periods
Time series analysis
• Seasonal variations are the difference between actual and trend
figures.
• An average of the seasonal variations for each time period
within the cycle will help to determine the final estimation for
period variation (Remember that to check the total of the
seasonal variations sums to zero)
Time series analysis

• Weakness of moving average

analysis
• the result is less accurate if
there is steeply rising or a
steeply declining trend
The product life cycle
• Sales of a product can be
expected to vary with the
passage of time.
• Stages in a product life
cycle:
➢Introduction
➢Growth
➢Maturity
➢Decline
The product life cycle
• Introduction stage:
✓A slow growth in sales.
✓Unit costs are high.
✓Loss-maker.
 The product life cycle
• Growth stage:
✓Sales rise more sharply.
✓Start to make profits.
✓Unit costs fall.
The product life cycle
• Maturity stage:
✓Sales will peak.
✓Sales growth slows down
(market saturation).
✓Profit at its best!
The product life cycle
• Declining stage
✓Sales growth becomes negative.
✓Profits decline.
✓Product might reach its ‘end’.