Contents

Abstract:......................................................................................................................................................1
Quantitative Analysis..............................................................................................................................1
Naive Model:...........................................................................................................................................1
Trend Analysis........................................................................................................................................1
Moving Averages....................................................................................................................................2
Exponential Smoothing...........................................................................................................................2
Holt's smoothing method.........................................................................................................................2
Simple linear regression..........................................................................................................................2
Multiple linear regression........................................................................................................................2
Summarize...................................................................................................................................................3
Appendix.....................................................................................................................................................3

Abstract:
Time-series forecasting models refer to statistical models that are employed to make predictions
about future values based on historical data points arranged in chronological order. For
forecasting purpose we have downloaded the data of sales, and fuel prices of German
manufactured cars from 2011 to 2022 from statista. To implement quantitative forecasting, we
have employ five approaches: the naïve approach, moving averages, trend projection,
exponential smoothing, Holt, simple linear regression and multiple linear regression. The results
of all the techniques shows that sales of German manufactured cars will raise in 2023 also found
trend in data except for the period of 2018 -2019. Further the regression analysis shows that time
frame is significant predictor of sales of German cars but not fuel prices.

Quantitative Analysis
Quantitative forecasting relies on measurable data, such as reliable and accurate historical data
such as past sales, labor reports, and a company's statistics, to eliminate the risks of inaccuracies
and biases in forecasts.

We have used following techniques for predictive analysis including: the naïve approach,
moving averages, trend projection, exponential smoothing, holt, simple linear regression and
multiple linear regression. These models scrutinize the trends and patterns present in the data and
extrapolate them to produce predictions about future values.

Naive Model:
The naive approach is one of the simplest forecasting technique which assumes that the outcome
of the previous period will repeat itself in the upcoming period. In 2022 the sales of German
manufactured cars was 506.1 (euros), we have forecasted that the same number of cars will be
sale in 2023 (See Table No.1).

Trend Analysis
Trend projection is a commonly used method of business forecasting that involves analyzing the
previous trends in a time-series and projecting them to continue in the future. This approach is
commonly utilized for budgeting and financial forecasting, as it enables predicting future market
trends by analyzing short-term market changes and revenue growth. The figure nO.1 below
shows that the rate of growth in German manufacture cars sales for a mature region begin to
decline in 2018 and then raise from 2020 till 2022 (See Figure No.1).

Moving Averages
The moving averages forecasting method calculates the average of past periods and assumes that
the next period's outcome will fall within that range. For analysis we have use 3 years moving
averages the average sales over the previous from 2020 to 2022 was 431.74 (euros), the moving
averages approach helps to predict that the sales of German manufactured cars in the upcoming
period would be around 431.74 euros (See Table No.2).

Exponential Smoothing
Exponential smoothing is a method that is commonly used for short-term forecasting.
Exponential smoothing assigns more substantial weights to recent observations, while the
weights decrease exponentially as the observations become more distant. This approach helps to
predict that the sales of German manufactured cars in the upcoming period would be around
428.94 euros (See Table No.3)
Holt's smoothing method
It is also referred to as linear exponential smoothing, is a popular forecasting model used to
analyze data that exhibits a trend. It indicates that sale of German manufactured cars will
increase arbitrarily into the future (See Figure No.2).

Simple linear regression

It is a method utilized by companies to predict revenues when there is only one dependent and
independent variable. The results shows that significant correlations between sales and time
period. Further the R-squared of the model is 0.61 which shows that model is mediocre fit (See
Table No.4).

Multiple linear regression

It is a method utilized by companies to predict revenues when there is a need for two or more
independent variables in the projection. In this part, a regression is performed on sales, time
period, and fuel prices to determine the significant correlations between these variables.

The analysis shows that fuel price is not significant predictor of sales of German manufactured
cars because its p-value 0.149 is greater than 0.05, whereas time frame is significantly use for
predicting the sales of German manufactured cars, further the R-squared of the model is 0.70
which shows that model is good fit (See Table No.5).

Summarize
The analysis above shows that forecasted sales of German manufactured cars will increase in
2023, this helps the manufactures to increase the vehicles’ production in 2023 for generating
more revenue. Further, the analysis also showed that fuel prices have not any impact on sales of
cars, so this also help them to know that the frequently raising fuel prices will not affect their
sales revenue.

Appendix
Table No.1

Sales(in Forecas Mean Absolute Squared Absolute Percentage

period Year Euros) t Error Error error
 1 2011 351.26
2 2012 356.97 351.26 5.71 32.6041 2%
3 2013 361.55 356.97 4.58 20.9764 1%
4 2014 367.95 361.55 6.4 40.96 2%
5 2015 404.77 367.95 36.82 1355.7124 9%
6 2016 404.61 404.77 0.16 0.0256 0%
7 2017 422.96 404.61 18.35 336.7225 4%
8 2018 426.21 422.96 3.25 10.5625 1%
9 2019 436.16 426.21 9.95 99.0025 2%
10 2020 378.17 436.16 57.99 3362.8401 15%
11 2021 410.91 378.17 32.74 1071.9076 8%
12 2022 506.15 410.91 95.24 9070.6576 19%
13 2023 506.15

Table No.2

3-
month
MA
Sales(in forecas
period Year Euros) t
1 2011 351.26 -
2 2012 356.97 -
3 2013 361.55 -
4 2014 367.95 356.59
5 2015 404.77 362.16
6 2016 404.61 378.09
7 2017 422.96 392.44
8 2018 426.21 410.78
9 2019 436.16 417.93
10 2020 378.17 428.44
 11 2021 410.91 413.51
12 2022 506.15 408.41
13 2023 431.74

13 2023 431.74

Table No.3

Alpha = 0.25
forecas ABS(error Squared Percent
period Year Revenue t error ) Error Error
1 2011 351.26 351.26 - - - -
2 2012 356.97 351.26 5.71 5.71 32.60 0.02
3 2013 361.55 352.69 8.86 8.86 78.54 0.02
4 2014 367.95 354.90 13.05 13.05 170.22 0.04
5 2015 404.77 358.16 46.61 46.61 2172.04 0.12
6 2016 404.61 369.82 34.79 34.79 1210.61 0.09
7 2017 422.96 378.51 44.45 44.45 1975.39 0.11
8 2018 426.21 389.63 36.58 36.58 1338.39 0.09
9 2019 436.16 398.77 37.39 37.39 1397.87 0.09
10 2020 378.17 408.12 -29.95 29.95 896.94 0.08
11 2021 410.91 400.63 10.28 10.28 105.64 0.03
12 2022 506.15 403.20 102.95 102.95 10598.44 0.20
13 2023 428.94 28.25 33.69 1816.06 0.08
BIAS MAD MSE MAPE
13 2023 428.94

Table No.4

SUMMARY
OUTPUT
 Regression Statistics
Multiple
R 0.785658
R Square 0.617259
Adjusted
R Square 0.578985
Standard
Error 28.4167
Observati
ons 12

ANOVA
Significa
df SS MS F nce F
Regressio 13022. 13022. 16.127
n 1 96 96 33 0.002455
8075.0 807.50
Residual 10 87 87
21098.
Total 11 05

Standa
Coefficie rd P- Lower Upper Lower Upper
nts Error t Stat value 95% 95% 95.0% 95.0%
17.489 19.456 2.81E- 379.24 301.30 379.24
Intercept 340.2761 27 28 09 301.3075 46 75 46
2.3763 4.0158 0.0024 14.837 4.2482 14.837
period 9.543042 24 85 55 4.248263 82 63 82

Table No.5
 SUMMARY OUTPUT

Regression Statistics
Multiple R 0.836968
R Square 0.700516
Adjusted R
Square 0.633964
Standard
Error 26.49638
Observations 12

ANOVA
Significan
df SS MS F ce F
7389.76 10.5258
Regression 2 14779.52 1 5 0.004402
702.058
Residual 9 6318.524 3
Total 11 21098.05

Coefficien Standard Lower Upper Lower Upper

ts Error t Stat P-value 95% 95% 95.0% 95.0%
5.80999 0.00025 376.679
Intercept 271.1179 46.66402 8 6 165.5565 2 165.5565 376.6792
1.58177 0.14815 1.31849
Fuel Prices 0.542559 0.343006 7 9 -0.23337 3 -0.23337 1.318493
3.99343 0.00314 14.0466
period 8.967095 2.245457 8 1 3.887518 7 3.887518 14.04667

Figure No.1
 Forecast vs. Actual

600.00
Sales

400.00 Sales(in
Euros)
200.00 forecast

0.00
1 2 3 4 5 6 7 8 9 1 1 1
0 1 2
Time

Figure No.2

Forecast vs. Actual

600.00
500.00
400.00
Sales

Revenue
300.00
forecast
200.00
100.00
0.00
1 2 3 4 5 6 7 8 9 10 11
Time