TABLE OF CONTENTS

INTRODUCTION ............................................................................................................................................................. 2
VISION ............................................................................................................................................................................ 3
COMMITMENTS ............................................................................................................................................................. 3
ORGANIZATIONAL HIERARCHY ...................................................................................................................................... 4
COMPETITOR ANALYSIS ................................................................................................................................................. 5
MARKETING STRATEGY .................................................................................................................................................. 5
PRODUCT ................................................................................................................................................................... 5
PRICE .......................................................................................................................................................................... 6
PLACE ......................................................................................................................................................................... 6
PROMOTION .............................................................................................................................................................. 6
COLLECTION OF DATA ................................................................................................................................................... 7
FORECASTING TECHNIQUES AND CALCULATIONS......................................................................................................... 9
Moving Average (MA) ................................................................................................................................................ 9
A. 3-period moving average ............................................................................................................................. 9
B. Sixth-Period Moving Average .......................................................................................................................... 10
C. Ninth-Period Moving Average ......................................................................................................................... 12
Weighted Moving Average ...................................................................................................................................... 14
Three-period Weighted Moving Average ............................................................................................................ 14
B. Four-period Weighted Moving Average .......................................................................................................... 15
Exponential Smoothing ................................................................................................................................................ 18
Linear Regression Analysis ........................................................................................................................................... 20
Mean Absolute Deviation (MAD) ................................................................................................................................. 24
Mean Squared Error (MSE) .......................................................................................................................................... 25
Mean Absolute Percentage Error (MAPE) ................................................................................................................... 27
Conclusion ................................................................................................................................................................... 29

INTRODUCTION

Lafarge Surma Cement Ltd. is a sister concern of Lafarge and Cementos Mollins. It was
incorporated on 11 November 1997 as a private limited company in Bangladesh under the
Companies Act 1994 having its registered office in Dhaka. On 20 January 2003 Lafarge Surma
Cement Ltd. was made into a public limited company. The Company is listed in Dhaka and
Chittagong Stock Exchange. Today, Lafarge Surma Cement Ltd. has more than 20,000
shareholders.
In November 2000, the two Governments of India and Bangladesh signed a historic agreement
through exchange of letters in order to support this unique cross border commercial venture and
till date it is the only cross border industrial venture between the two countries. Since
Bangladesh does not have any commercial deposit of limestone, the agreement provides for
uninterrupted supply of limestone to the cement plant at Chhatak in Bangladesh by a 17 km long
belt conveyor from the quarry located in the state of Meghalaya. The company in Bangladesh,
Lafarge Surma Cement Ltd. wholly owns a subsidiary company Lafarge Umiam Mining Private
Ltd. (LUMPL) being registered in India, which operates its quarry at Nongtrai in Meghalaya.
This commercial venture with an investment of USD 280 million, which is one of the largest
foreign investments in Bangladesh, has been financed by Lafarge of France, world leader in
building materials, Cementos Molins of Spain, leading Bangladeshi business houses together
with International Finance Corporation (IFC The World Bank Group), the Asian Development
Bank (ADB), German Development Bank (DEG), European Investment Bank (EIB), and the
Netherlands Development Finance Company (FMO).
Lafarge Group, with 176 years of experience, holds worlds top-ranking position in Cement,
Aggregates, Concrete and Gypsum. It operates in 64 countries with around 68,000 employees.
Lafarge is named as one of the 100 Most Sustainable Companies in the World.
Now, after three years of production operations, they are producing world class clinker and
cement which is a demonstration of the sophisticated and state-of-the-art machineries and
processes of our plant at Chhatak. The Company is already meeting about 8% of the total market
need for cement and 10% of total clinker requirements of Bangladesh market whereas we
continue to enjoy strong growth rates. By supplying clinker to other cement producers in the
market, we contribute some USD 50~60 million per annum worth of foreign currency savings
for the country. We contribute around BDT 1 (one) billion per annum as government revenue to
the national exchequer of Bangladesh. About 5,000 people depend on our business directly or
indirectly for their livelihood.
They believe that cement is an essential material that addresses vital needs of the construction
sector. They are optimistic to meet the growing needs for housing and infrastructure in the
construction sector of Bangladesh.
VISION

To be the undisputed leader in building materials in Bangladesh through
- Excellence in all areas of operations in world class standards
- Harnessing our strength as the only cement producer in Bangladesh
- Sustainable growth that respects the environment and the community
COMMITMENTS

- Offering highest quality of product and services that exceeds our customers
expectations.
- Giving our people an enabling environment that nurtures their talent and opportunities to
give the best for the organization.
- Contributing to build a better world for our communities.
- Delivering the value creation that our shareholders expect.

ORGANIZATIONAL HIERARCHY




Mr. Tarek Samir
Ahmed Elba
CEO
Board of
Directors
Mr. Masud Khan
Finance Director
Ms. Sayeda Tahya
Hossain
Hr & Corporate
Affairs Director
Mr. Kazi Khalid
Mahmood
Commercial
Director
Mr. Asim
Chottopadyay
Senior Vice
President Operation
Mr. Mohammed
Arif Bhuiyan
Suppy Chain
Director
Mr. Narayan
Prasad Sharma
Vice President
and Director,
LUMPL
Mr. Kazi Mizanur
Rahman
Comapany Secretary

COMPETITOR ANALYSIS

There is only one type of competitor and that is Direct Competitors. There are no indirect
competitors because there are no substitutes for cement. There are many competitors of Lafarge
Surma Cement Ltd. and some them are the established local brands and also some multinational
brands. The companies which are the competitors of Lafarge Surma Cement Ltd are mainly the
multinational companies which are on the radical rating scale of Lafarge Surma Cement Ltd like
Heidelberg Cement Bangladesh Ltd, Cemex, Holcim, etc. be it in terms of brand image, product
quality and sales, each of the companies have their own unique features for which they are sited
in the higher ranks. However, Lafarge Surma Cement Ltd sees each of these companies as their
competitors and they are continuously fighting with these companies to reach their preferred
position in the market. Moreover, the cement companies which are ranked below the company
such as Shah Cement and Crown Cement are also viewed as competitors and they keep a look
out for these companies activities. Hence, Lafarge Surma Cement Bangladesh Ltd faces intense
competition in the market place.
MARKETING STRATEGY

PRODUCT

The products of Lafarge Surma Cement Ltd are Gray Cement and Clinker Cement. They are the
most preferred brands for long lasting constructions. They have the following features:
Consistent proven quality.
Long age strength ensuring continuous strength gain.
Higher durability- sustainable from generation to generation.
Higher workability.
Low heat of hydration.
Better resistance to sulphate and other chemical attack.
PRICE

The price of the products of Lafarge Surma Cement Ltd is a little higher than the other cements
in the market. This is due to their high quality of cement. However, commissions are given to
their customers especially the retailers based on the volume taken.

PLACE

Lafarge Surma Cement Ltd has only one factory of their own which is located at Chattak under
Sunamgonj district. Lafarge Surma Cement Ltd. will extract and procss the basic raw materials
like limestone and shale from its from its own quarry in Meghalaya, India. A 17 km crossborder
belt conveyor will be installed to link the quarry with the cement plant for transportation of raw
materials.
PROMOTION

Lafarge Surma Cement Ltd uses a combination of promotional activities to advertise their
products. The most widely used instrument is Newspapers. They keep on updating about their
products to the local customers through both English and Bangla newspapers like Daily Star,
Financial Express, Amader Shomoy, Prothom Alo, Bangladesh Protidin etc. Moreover, they also
go for Billboard advertisement to attract the customers. They also use brochures and posters to
inform their customers.
COLLECTION OF DATA

We have collected 20 months sales data from Lafarge Surma Cement Ltd. We have collected the
data to make the forecast and the compare between the forecast and the actual data by using
different forecasting and comparing techniques. Our purpose was to use these data for
forecasting and to compare between the forecast and the actual data of the sales of from Lafarge
Surma Cement Ltd. by using following forecasting and comparing techniques:
1. We used the data to find out the 3-period moving average to forecast the production for
the 4
th
period 21
st
periods.
2. We used the data to find out the 6-period moving average to forecast the production for
the 4
th
period 21
st
periods.
3. We used the data to find out the 9-period moving average to forecast the production for
the 4
th
period 21
st
periods.
4. We used the data to find out the 3 and 4- period weighted average to forecast the
production for the 4
th
period 21
st
periods.
5. We used the data to find out Exponential Smoothing with =.05, =.1 and = 0.4 to
forecast the sales for the 4
th
period 21
st
periods.
6. We used the data to plot the data to see whether the data follow a trend. Determine the
linear trend line for production. Use the trend equation to predict the production for the
5
th
period 30
th
periods. Use Excel to do this experiment.
7. We used the data to compare the performance of the five methods by using the mean
absolute deviation (MAD) as the performance criterion.
8. We used the data compare the performance of the five methods by using the Mean
Squared Error (MSE) as the performance criterion.
9. We used the data to compare the performance of the five methods by using the mean
absolute percentage error (MAPE) as the performance criterion.

Here is the raw data for the analysis that is collected from the Lafarge Surma Cement Ltd. at
Gulshan on Cement sells figure in BDT for 20 months.
**Sales in BDT (Crore).
Sales
(Taka Crore)
2010 JAN 72.6
FEB 72.1
MAR 73
APR 64
MAY 65.4
JUN 27
JUL 29
AUG 26
SEP 26
OCT 37
NOV 39
DEC 34.4
2011 JAN 39
FEB 41
MAR 36.1
APR 79
MAY 80
JUN 76.5
JUL 42
AUG 45

FORECASTING TECHNIQUES AND CALCULATIONS

MOVING AVERAGE (MA)

A. 3-PERIOD MOVING AVERAGE

Technique that averages a number of recent actual data values, updated as new values become
available is known as moving average forecast. The moving average forecast can be computed
using the following formula.

Formula
n
A
MA F
n
i
i t
n t

=

= =
1

Where,
t
F
= Forecast for time period t
i = an index that corresponds to time period
n = No. of period in the moving average
i
A
= Actual value in period t-i
MA = Moving average


Working Procedure of 3-Period Moving Average:

- STEP 1: Get actual data for Period 1 to 3 from the month of January.
Period 1 to Period 3 shows the actual data.

- STEP 2: Calculate the sum of last three months data (actual or forecast whatever we have).
We could find out the sum of last three months by using the formula-
=SUM (Period1:Period3) or
= (Period 1+Period 2+Period3)

- STEP 3: Calculate the average by dividing the sum of last three months actual data by the no.
of correspondent time period.

To find out the average of last three periods by using the Formula-
SUM (Period1:Period3)/3"or, =(Period 1+ Period 2+ Period 3)/3

STEP 4: To find out the next months forecasted data we have to simply calculate last three
months data (actual or forecasted whatever we have) and divide that by the no of correspondent
time period.

B. SIXTH-PERIOD MOVING AVERAGE

Technique that averages a number of recent actual data values, updated as new values become
available is known as moving average forecast. The moving average forecast can be computed
using the following formula.

Formula
n
A
MA F
n
i
i t
n t

=

= =
1


Where,
t
F
= Forecast for time period t
i = an index that corresponds to time period
n = No. of period in the moving average
i
A
= Actual value in period t-i
MA = Moving average

Working Procedure of Six-Period Moving Average:

- STEP 1: Get actual data for Period 1 to 6.
Period 1 to Period 6 shows the actual data.

- STEP 2: Calculate the sum of last six months data (actual or forecast whatever we have).
We could find out the sum of last six months by using the formula
=SUM (Period1:Period6).

- STEP 3: Calculate the average by dividing the sum of last six months actual data by the no of
correspondent time period.
To find out the average of last six periods by using the formula-
=SUM (Period1:Period6)/6"

- STEP 4: To find out the next months forecasted data we have to simply calculate last six
months data (actual or forecasted whatever we have) and divide that by the no of
correspondent time period.
C. NINTH-PERIOD MOVING AVERAGE

Technique that averages a number of recent actual data values, updated as new values become
available is known as moving average forecast. The moving average forecast can be computed
using the following formula.

Formula
n
A
MA F
n
i
i t
n t

=

= =
1


Where,
t
F
= Forecast for time period t
i = an index that corresponds to time period
n = No. of period in the moving average
i
A
= Actual value in period t-i
MA = Moving average




Working Procedure of Ninth-Period Moving Average:
- STEP 1: Get actual data for Period 1 to 9.
Period 1 to Period 9 shows the actual data.

- STEP 2: Calculate the sum of last nine months data (actual or forecast whatever we have).
We could find out the sum of last nine months by using the formula
=SUM (Period1:Period9).

- STEP 3: Calculate the average by dividing the sum of last nine months actual data by the no
of correspondent time period.
To find out the average of last nine periods by using the formula-
=SUM (Period1:Period9)/9"
- STEP 4: To find out the next months forecasted data we have to simply calculate last nine
months data (actual or forecasted whatever we have) and divide that by the no of
correspondent time period.
Sales(Taka Crore) 3-Month
MA
6-Month
MA
9-Month
MA
2010 JAN 72.6
FEB 72.1
MAR 73
APR 64 72.6
MAY 65.4 69.7
JUN 27 67.5
JUL 29 52.1 62.4
AUG 26 40.5 55.1
SEP 26 27.3 47.4
OCT 37 27 39.6 50.6
NOV 39 29.7 35.1 46.6
DEC 34.4 34 30.7 42.9
2011 JAN 39 36.8 31.9 38.6
FEB 41 37.5 33.6 35.9
MAR 36.1 38.1 36.1 33.2
APR 79 38.7 37.8 34.2
MAY 80 52 44.8 39.7
JUN 76.5 65 51.6 45.7
JUL 42 78.5 58.6 51.3




WEIGHTED MOVING AVERAGE

THREE-PERIOD WEIGHTED MOVI NG AVERAGE

Weighted moving average is similar to the moving average, except that it assigns more weight to
the most recent values in a time series. For example, the most recent value might be assigned a
value of 0.4, the next most recent value weight of 0.3, the next after that a weight of 0.2, the next
after that a weight of 0.1 where the total sum of the weights is 1. The weighted moving average
can be computed by the following formula.

Formula
1 1 2 2 ) 2 ( 2 ) 1 ( 1
......

+ + + + =
t t n t n n t n n t n t
A w A w A w A w A w F

Where,
t
F
= Forecast for time period t
n = No. of period in the moving average
n
w
= Weight of the Actual Value
t
A
= Actual value in period t



AUG 45 66.2 59.1 51.9
SEP 54.5 59.8 52.6

Working Procedure of Weighted Moving Average:
- STEP 1: Get actual data for Period 1 to 3.
Period 1 to Period 3 shows the actual data.

- STEP 2: Select the weight, what we are going to assign for the weighted moving average.
We have selected weight as 0.5, 0.3, 0.2, where-
0.5 represents the most recent periods weight.
0.3 represents the next recent periods weight.
0.2 represents the next recent periods weight.

- STEP 3: Multiply the data with their correspondent weight.
(Most recent periods * 0.5)
(Next recent periods * 0.3)
(Next recent periods * 0.2)

- STEP 4: Calculate the sum of multiplied the data.
(Most recent periods * 0.5) + (next recent periods * 0.3) + (next recent periods * 0.2)

- STEP 5: Step 4s calculated data will be the forecasted value for the next period. To find out
the next forecasted data we have to simply calculate last three periods data (actual or
forecasted whatever we have) and assign correspondent weights.

B. FOUR-PERIOD WEIGHTED MOVI NG AVERAGE

Weighted moving average is similar to the moving average, except that it assigns more weight
to the most recent values in a time series. The most recent value might be assigned a value of
0.4, the next most recent value weight of 0.3, the next after that a weight of 0.2, the next after
that a weight of 0.1 where the total sum of the weights is 1. The weighted moving average can
be computed by the following formula.

Formula
1 1 2 2 ) 2 ( 2 ) 1 ( 1
......

+ + + + =
t t n t n n t n n t n t
A w A w A w A w A w F

Where,
t
F
= Forecast for time period t
n = No. of period in the moving average
n
w
= Weight of the Actual Value
t
A
= Actual value in period t

Working Procedure of Weighted Moving Average:
- STEP 1 : Get actual data for Period 1 to 4.
Period 1 to Period 4 shows the actual data.

- STEP 2: Select the weight, what we are going to assign for the weighted moving average.
We have selected weight as 0.4, 0.3, 0.2, 0.1, where-
0.4 represents the most recent periods weight.
0.3 represents the next recent periods weight.
0.2 represents the next recent periods weight.
0.1 represents the next recent periods weight.

- STEP 3: Multiply the data with their correspondent weight.
(Most recent periods * 0.4)
(Next recent periods * 0.3)
(Next recent periods * 0.2)
(Next recent periods * 0.1)

- STEP 4: Calculate the sum of multiplied the data.
(Most recent periods * 0.4) + (next recent periods * 0.3) + (next recent periods * 0.2) +
(next recent periods * 0.1)

- STEP 5: Step 4s calculated data will be the forecasted value for the next period. To find out
the next forecasted data we have to simply calculate last four periods data (actual or
forecasted whatever we have) and assign correspondent weights.

















Sales(Taka Crore) 3-Month WMA 4-Month WMA
2010 JAN 72.6
FEB 72.1
MAR 73
APR 64 72.7
MAY 65.4 68.3 69.2
JUN 27 66.5 67.2
JUL 29 45.9 50.5
AUG 26 35.7 39.2
SEP 26 27.1 31
OCT 37 26.6 26.7
NOV 39 31.5 30.7
DEC 34.4 35.8 34.5
2011 JAN 39 36.3 35.5
FEB 41 37.6 37.4
MAR 36.1 39.1 38.9
APR 79 38.2 38
MAY 80 58.5 54.5
JUN 76.5 70.9 67
JUL 42 78.1 74
AUG 45 60 63.7
SEP 50.4 53.9
EXPONENTIAL SMOOTHING

Weighted averaging method based on previous forecast plus a percentage of the forecast error. It
is sophisticated weighted average method that is still relatively easy to use and understand.

Formula
Next forecast = Previous forecast + (actual previous forecast)
( )
1 1 1
+ =
t t t t
F A F F o

Where,
1 t
F
= Forecast for the previous time period
o
= The smoothing constant = % of the error
1 t
A
= Actual value in the previous period
t
F
= Forecast for time period t
The value of commonly used ranges from 0.05 to 0.5. Low values are used when the average
tends to be stable. Higher values of
o
are used when the average is not stable.
Working Procedure of Exponential Smoothing:
For = 0.05, 0.1 and 0.4
- STEP 1: Get actual data.
Period 1 to Period 20 shows the actual data.

- STEP 2: Find out the difference between actual and forecasted data. For the 2
nd
period, as we
dont have the forecasted value so we are going to take the period 1
st
actual value as
period 2
nd
forecasted value.
- STEP 3: Multiply the difference between actual and forecasted period with the error
constants.

- STEP 4: Add forecasted data and error constant considered difference of actual and
forecasted data.

- STEP 5: Step4s calculated data will be the forecasted value for the next period. To find out
the next years forecasted data we have to simply use this formula for all the next period


[Forecasted+ (smoothing constant*(actual-forecasted)]

Sales(Taka Crore) =.05 =.1 =.4
2010 JAN 72.6
FEB 72.1 72.6 72.6 72.6
MAR 73 72.6 72.6 72.4
APR 64 72.62 72.6 72.6
MAY 65.4 72.2 71.7 69.2
JUN 27 71.9 71.1 67.7
JUL 29 69.8 66.7 51.4
AUG 26 67.8 62.9 42.4
SEP 26 65.7 59.2 35.8
OCT 37 63.7 55.9 31.9
NOV 39 62.4 54 33.9
DEC 34.4 61.2 52.5 35.9
2011 JAN 39 59.9 50.7 35.3
FEB 41 58.9 49.5 36.8
MAR 36.1 58 48.7 38.5
APR 79 56.9 47.4 37.5
MAY 80 58 50.6 54.1
JUN 76.5 59.1 53.5 64.5
JUL 42 60 55.8 69.3
AUG 45 59.1 54.4 58.4
SEP 58.4 53.5 53




a t
y
LINEAR REGRESSION ANALYSIS

Analysis of linear regression involves developing an equation that will suitably describe the
trend. It may be linear or may not be. Linear trends are fairly common.

Formula:
A linear trend equation has the form
t b a F
t
+ =


Where,
t
F
= Forecast for time period t
t
= Specified no. of time periods from t = 0
a
= Value of
t
F
at t = 0
b
= Slope of the line



The coefficients of the line a and b can be computed from the historical data using the
formulas:
( )
2
2

=
t t n
y t ty n
b
&
t b y
n
t b y
a =

=



Where,
t is the period, n is the number of periods and y is the actual demand.
Working Procedure of Trend Equation:
- STEP 1: Get all the actual data and plot the data in the graph and find out n.
Period 1 to Period 20 shows the actual data where n is 17.
- STEP 2: Define, t and y and calculate the sum of t and the sum of y.
Here, period is expressed as t and volume nos as y. and
- STEP 3: Multiply t and y and calculate the sum of ty.

- STEP 4: Find out t
2
and calculate the sum of t
2
.

- STEP 5: Calculate the square of sum of t
- STEP 6: Calculate b by using formula,
( )
2
2

=
t t n
y t ty n
b
- STEP 7: Calculate a by using formula, t b y
n
t b y
a =

=


- STEP 8: Find out the trend line by using this formula
t b a F + =
and plot that in the graph.

- STEP 9: Calculate each years forecasted data by using the same
Formula
t b a F + =
Where t is the time for each and every period.


Here, b= 1.19, and a= 35.5
The regression equation is
C2 = 35.5 + 1.19 C1


Predictor Coef SE Coef T P
Constant 35.527 9.444 3.76 0.002
C1 1.1924 0.9216 1.29 0.215

S = 18.6153 R-Sq = 10.0% R-Sq(adj) = 4.0%

Month t y(At)
t
Ty
Apr 1 64 1 64
May 2 65.4 4 130.8
Jun 3 27 9 81
Jul 4 29 16 116
Aug 5 26 25 130
Sep 6 26 36 156
Oct 7 37 49 259
Nov 8 39 64 312
Dec 9 34.4 81 309.6
Jan 10 39 100 390
Feb 11 41 121 451
Mar 12 36.1 144 433.2
Apr 13 79 169 1027
May 14 80 196 1120
Jun 15 76.5 225 1147.5
Jul 16 42 256 672
Aug 17 45 289 765
153 786.4 1785 7564.1

From the above equation we can see the R (coefficient of determination) is only 10%. This
means that only 10% variation in sales is explained by the changes in the number of periods.

18 16 14 12 10 8 6 4 2 0
80
70
60
50
40
30
20
C1
C
2
Scatterplot of C2 vs C1

Trend Analysis

16 14 12 10 8 6 4 2
80
70
60
50
40
30
20
Index
C
2
MAPE 34.910
MAD 15.718
MSD 305.761
Accuracy Measures
Actual
Fits
Variable
Trend Analysis Plot for C2
Linear Trend Model
Yt = 35.53 + 1.19*t



MEAN ABSOLUTE DEVIATION (MAD)

It is a significant factor when deciding among forecasting alternatives. Accuracy is based on the
historical error performance of a forecast. It is very important to find the historical error
performance of different techniques. So, we can easily find which the best method is. There are
different methods of measuring historical errors. Mean absolute deviation (MAD) is the first one.

Formula
Mean absolute deviation (MAD): Average absolute error.
MAD =
n
F A
t t


Where,
t
F
= Forecast for time period t
n = No. of period
t
A
= Actual value in period t
Here, we have to take absolute value of the error.
Working Procedure of Mean Absolute Deviation (MAD):

- STEP 1: Get actual data and all the forecasted data for every techniques
Here we have four techniques of- 30 days moving average, linear forecasting method, weighted
moving average and Exponential smoothing.

- STEP 2: Calculate the difference between actual data and forecasted data for every
techniques by using formula, (
t t
F A
).

- STEP 3: Take the absolute value of the differences of actual and forecasted data for every
techniques by using formula,
t t
F A
.

- STEP 4: Calculate the sum of absolute value of the differences of actual and forecasted data
for every techniques.

t t
F A


- STEP 5: Divide the sum of absolute value of the differences of actual and forecasted data by
the no of the period, n for every six technique Using formula,
n
F A
t t

.

- STEP 6: Follow STEP 1 to STEP 5 for every techniques and find out the MAD.
Results
MAD of 3-period moving average

15.13
MAD of 6-period moving average

17.19
MAD of 9-period moving average

15.47
MAD of 3-period weighted period moving average

13.31
MAD of 4-period weighted period moving average

14.94
MAD of Exponential Smoothing = .05, 21.82
= .1, 18.93
= .4, 12.89
MAD of Linear Regression Analysis

15.72

Recommendation
We recommend MAD of exponential smoothing where = 0.4 because the value of it is lowest,
12.89 compare to others.
MEAN SQUARED ERROR (MSE)

Mean Squared Error is the second methods of measuring error or accuracy of the forecasted data.

Formula
Mean squared error (MSE): Average of squared errors.
MSE =
( )
1
2

n
F A
t t

Where,
t
F
= Forecast for time period t
n = No. of period
t
A
= Actual value in period t

Working Procedure of Mean Squared Error (MSE):
- STEP 1: Get actual data and all the forecasted data for every techniques
Here we have five techniques of- 30 days moving average, linear forecasting method, weighted
moving average and Exponential smoothing.

- STEP 2: Calculate the difference between actual data and forecasted data for every
techniques by using formula, (
t t
F A
).

- STEP 3: Do squares of the differences of actual and forecasted value for every techniques by
using formula,
2
) (
t t
F A
.

- STEP 4: Calculate the sum of the squares of the differences of actual and forecasted data for
every techniques,
( )

2
t t
F A
.

- STEP 5: Divide the sum of squares of the differences of actual and forecasted data by the n-
1, no of the period for every technique using formula,
( )
1
2

n
F A
t t
.

- STEP 6: Follow STEP 1 to STEP 5 for every techniques and find out the MSE.


Result

MSE of 3-period moving average
438

MSE of 6-period moving average
507

MSE of 9-period moving average
506

MSE of 3-period weighted period moving average
369

MSE of 4-period weighted period moving average
420

MSE of Exponential Smoothing
= .05, 674.4
= .1, 549.6
= .4, 342.5

MSE of Linear Regression Analysis
305.76

Recommendation
We recommend MAD of exponential smoothing where = 0.4 because the value of it is lowest,
342.5 compare to others.

MEAN ABSOLUTE PERCENTAGE ERROR (MAPE)

Mean Absolute Percentage Error is the third most common methods of measuring error or the
accuracy of the forecasted data.

Formula
Mean absolute percent error (MAPE): Average absolute percent error.

MAPE =
n
A
F A
t
t t
100

Where,
t
F
= Forecast for time period t
n = No. of period
t
A
= Actual value in period t
Working Procedure of Mean Absolute Percentage Error (MAPE):
- STEP 1: Get actual data and all the forecasted data for every techniques

- STEP 2: Calculate the difference between actual data and forecasted data for every
techniques by using formula, (
t t
F A
).

- STEP 3: Take the absolute value of the differences of actual and forecasted data for every
techniques by using formula,
t t
F A


- STEP 4: Divide the absolute value of the differences of actual and forecasted data by the
actual value for every techniques.
t
t t
A
F A

- STEP 5: Multiply, the division of the absolute value of the differences of actual and
forecasted data by the actual value for every techniques.
100

t
t t
A
F A


- STEP 6: Divide the division of the sum of absolute value of the differences of actual and
forecasted data by the no of the period, n for every five technique using formula,
n
A
F A
t
t t
100

.
- STEP 7: Follow STEP 1 to STEP 6 for every techniques and find out the MAPE.
Results

MAPE of 3-period moving average
36.3

MAPE of 6-period moving average
40.9

MAPE of 9-period moving average
26.1

MAPE of 3-period weighted period moving
average
32.0

MAPE of 4-period weighted period moving
average
36.7

MAPE of Exponential Smoothing
= .05, 61.4
= .1, 51.3
= .4, 31.7

MAPE of Linear Regression Analysis
34.91
Recommendation
We recommend MAPE of 9-period moving average because the value is lowest, 26.1 compare to
others.
CONCLUSION
Forecasting is a useful method for predicting future demand in all sectors. It helps managers plan
the system and help them plan the use of the system. Forecasts play an important role in the
planning process because they enable a manager to anticipate the future so he/she can plan
accordingly. A wide variety of forecasting techniques are in use. In many respects, they are quite
different from each other. In our analysis we have used some techniques of forecasting and they
are moving average, weighted average, exponential smoothing and linear trend method. Each
one of them has a unique advantage, easy to understand, provide good forecast and are reliable.
We have also used MAD, MSE and MAPE to analyze the accuracy of the forecasts. At the end
we would like conclude that we have tried our best to give the best results of forecasting and
hope that it will help the organization in the future.