Sol America Diaz Eiserman

Kun Xue
Naga Thejus Sambasivan Mruthyunjaya
Mladen Toncev

Assignment 2
Group 14

Executive Summary
In this paper we analyze Vestas wind Systems A/S who are major players in the sustainable
energy segment. With their pioneering prowess and state of the art engineering capabilities, they
aim to become carbon neutral by 2030 and introduce zero waste wind turbines by 2040. They
have a good market share for both onshore and offshore wind turbines and services in around 80
countries. Their current plan is to expand their presence in China. In Vestas annual report they
expect ~ 9% of growth at a compound annual growth rate (CAGR) for global wind energy for
the year 2020-2026.

In this analysis we use the data from Vestas’ annual reports to estimate the demand function for
both onshore wind turbines and services using regression analysis. The proposed hypothesis for
onshore wind turbines is correlated with the regression runs, which shows the impact of pricing
and existing production capacity on the demand for wind turbines and that Vestas cannot lower
the price without scaling their production economies. The proposed hypothesis for services is
correlated with the regression runs, which can be deduced to say that the impact of accumulated
wind power has higher impact on the service price and its future demand is based on onshore
wind power.

For wind turbines using the price elasticity formula we see that the order intake demand is
inelastic in few terms (2006-2008, 2017-2018), unitary (2014-205) and elastic in rest of the
terms. Using the results from the regression analysis and computing for the year 2020 we find
that the demand is inelastic. This could mean that the price of onshore wind turbines per MW is
an inverse determinant of the quantity of yearly order intake for onshore wind turbines in MW.
And therefore, the ability to attract more orders is possible in very competitive market only with
continuous downward price pressure.

For services using the price elasticity formula, we see that the elasticity of demand is inelastic for
terms (2012-2013 and 2016-2017) and elastic for the rest of the terms. But considering the
results from regression we see that the demand for services is inelastic for year 2020, which can
mean that the future demand for services is inelastic and will be driven by total accumulated
Vestas onshore wind power.
For onshore wind turbines, the sales forecast for year 2021 was done using

- Visual time series which gives a forecast of 17000 MW

- Exponential smoothing which gives a forecast of 17202 MW
- Econometric model which gives a forecast of 20637 MW

For services, the sales forecast for year 2021 was done using

- Visual time series which gives a forecast of 135 GW

- Moving averages which gives a forecast of 99.67 GW
- Econometric model which gives a forecast of 133 GW

From above analysis we see that visual time series and exponential smoothing gives similar
values for onshore wind turbines. And visual time series and econometric model gives similar
values for services.

Firm, its goals and product offering.

Vestas Wind Systems A/S is a global company in the wind power generation systems with more
than 40 years in the market (Business, 2021). The company is originally from Denmark where is
the headquarters. Interesting to highlight is that Vestas started as a blacksmith family business
and with the years have evolved to what it is now. Today they have implemented their solutions
approximately in 80 countries. Their production plants are in more than 12 countries among
them are Denmark, The Netherlands, Spain, The United Kingdom, Sweden, Norway, Australia,
Brazil, the United States and China. The last one as a strategic move to have an entrance in the
Chinese market. They produce and develop their own products from wind turbine pieces, full
farms of windmills, turnkey solutions to innovative solutions like turnkey offshore wind
platforms. They are very well known as global leaders in their area due to their engineering,
design, and manufacturing of their products.

Vestas's ambition is to become a sustainable energy solution leader, with wind as center piece
(Vestas Wind Systems, 2020). The firm kept consistent with their initial approach when 1979
they considered the environment when implementing the first wind turbine.

The goal for the firm is “to lead the transition towards a world powered by sustainable energy”
(Vestas Wind Systems, 2020). The company strives to be leading the energy industry to safety,
inclusion, and social responsibility. By 2030 they set up to reach carbon neutrality, without
carbon offsets, and by 2040 they aim to introduce zero-waste wind turbines (Vestas Wind
Systems, 2021).

Vestas is making good use of their strengths: their global reach, technology and service
leadership, proven execution, and their cost-leading position (Vestas Wind Systems, 2021). They
have managed to grow their business with local governments as well as with providers, helping
the firm to keep a stable profit which they reinvest in innovative technologies that make Vestas
consolidate their position.

The firm strategy for their business is focus on three key areas (Vestas Wind Systems, 2021):

· onshore wind
· service solutions and
· offshore wind

Vestas keep moving the economy and the technology to the wind power for the position that
have in the industry (Vestas Wind Systems, 2021). They have a global efficient operation,
innovative solutions and investments that make them competitive and keep a leadership in their
cost in the market.

The company offers more than 25 years of experience which makes them world leaders at
offering offshore wind turbines with quality and safety (Vestas Product Portfolio, 2021). Among
their experience in their portfolio, they also offer fixed-bottom and floating offshore wind
solutions, including the services associated to this type of product to secure quality and on time
results (Vestas Product Portfolio, 2021). Vestas's services portfolio includes parts & repair,
maintenance, fleet optimization, digital services, among others. In addition, the company have
Options & Solutions that have the focus to boost the implementation and performance of their
products (Vestas Product Portfolio, 2021).

Market Demand for the firm’s products

From Vestas point of view businesswise they face the challenge and the opportunity that will
bring the projections on the increase of the electrical global demand (Vestas Wind Systems,
2020). This will require an increase in the capacity to generate electricity which will attract
investors that are expected to choose the wind and solar energy. The expectation is an annual
investment in new capacity for wind. It is estimated to be approximately the doubled in 10 to 15
years in comparison to the past 2 years.
The graph above represents the market expectations and the projection in CAGR from 2020 to
2025 for the different business areas that Vestas have. The figures indicated in the graph were
presented in their Annual report from 2020 (Vestas Wind Systems, 2021).

For the period 2020-2026 the market overview for global win energy indicates a ~ 9% of growth
at a compound annual growth rate (CAGR) for the wind turbine market (Gupta & Singh Bais,
2020). For the installation there is an expectation to go above 101 GW by 2026. Restrictive
regulations that have been put in practice by different governments to control the increase on the
carbon mark will influence the growth of the wind energy solution in the different markets
(Gupta & Singh Bais, 2020). Furthermore, the expansion of some growing countries shows that
the future need of energy will increase.

During 2020 the global economy has been affected by the COVID-19 pandemic reducing the
capacity to progress in farm projects. Even considering this the wind power business has grown
and indicated a positive trend during the year 2020 (Gupta & Singh Bais, 2020).

When it comes to the European market, they are the one dominating the offshore wind business
globally (Research and Markets, 2020). This market indicates an approximately GAGR growths
of 6,11%. The Offshore section of the wind energy business is showing a tendency of growth.
This offshore market keeps been stable since 2018 with a tendency on increase in approximately
23,74% in progressive installed capacity in relation to other years.

In Europe, the focus for the growth of the offshore wind business are the progress and growth in
technology what indicates that will decline the manufacturing value of the offshore wind farms
producing some comparisons with the onshore wind farms on the financial behavior (Research
and Markets, 2020). For the forecast period, the offshore wind business has a projection to
increase and keep in a positive trend.

Demand Function Estimation for Onshore Wind Turbines

We have used regression analysis to estimate demand function for onshore wind turbines. The
demand for onshore wind turbines corresponds to total order intake per year and is expressed in
number of Mega Watts (MW). We have extracted total order intake in both billion EUROs and
MW from Vestas’ annual reports (Vestas Wind Systems, 2021a) covering period of 15 years
(2006-2020). Selling price per MW for each year is calculated by dividing order value in EUROs
by order intake in MW and is presented in million EUROs. Total added global onshore capacity
in MW from 2009-2019 is taken from Statista (n.d.) and Pitteloud (2018). The source for years
2006-2008 and 2020 is Vestas’ annual reports (Vestas Wind Systems, 2021a). Vestas’ annual
reports (Vestas Wind Systems, 2021a) are used as source for the following data: number of MW
in onshore wind turbines shipped to customers during each year. All extracted values are
presented in table below. Order intake in MW is our dependent variable. Selling price in mEUR
per MW, total added onshore capacity in MW, and shipped previous year MW are three
independent variables. Three graphs show scatter plots for each of independent variables paired
with order intake in MW for the set of data from the table.

Our regression analysis hypothesis:

Hypothesis 1: The price of onshore wind turbines per MW is an inverse determinant of the
quantity of yearly order intake for onshore wind turbines in MW (the sign of the coefficient is
expected to be negative). We expect that ability to atract more orders is possible in very
competitive market only with continuous downward price pressure.

Hypothesis 2: The total added global onshore wind turbines capacity in MW is a direct
determinant of the quantity of yearly order intake for onshore wind turbines in MW (the sign of
the coefficient is expected to be positive). We expect that demand for Vestas onshore wind
turbines follows the changes in the global demand for power turbines in each calendar year. Due
to very competitive market and strong players from China our expectation is that Vestas’ market
share will grow slightly slower than global market.

Hypothesis 3: Vestas’ shipped number of onshore wind turbines in MW in previous year is a

direct determinant of the quantity of yearly order intake for onshore wind turbines in MW (the
sign of coefficient is exepcted to be positive). We expect that increase in Vestas’ production in
previous year will positively impact the order intake in the current year as Vestas will be able to
respond more quickly to the increased market demand for green power.
Results of correlation analysis of three independent variables as well as regression analysis for
seven different combination of independent variables are shown here:

The results of the first, second and third simple regressions reveals that our hypothesis correctly
explained the nature of correlation between the demand for onshore wind turbines power and our
three independent variables with high 0.01 level of statistical significance. The sign of the
regression coefficients is not changed in multiple regression analysis (reg4, reg5, reg6 and reg 7).

Due to correlation coefficient being higher than 0.7 there is multicollinearity between Vestas’
selling price per MW of onshore wind turbines power and amount of MW of onshore wind
turbines power that Vestas shipped in previous year. That means there is close association
between increase in capacity and decrease in average selling price. However, F-test has passed in
multiple regression analysis reg5 and reg7 that include these two independent variables. As
expected, t-test for each regression coefficient is impacted due to reduced t-values and it is
harder to reject the null hypothesis in multiple regression analysis. We have also checked for
autocorrelation in reg5 and reg7. Durbin-Watson test is inconclusive at 5% significance points.
On the other hand, Breusch-Godfrey test results show that there is no autocorrelation.

As per Keat, Young and Erfle (2014, pp.153-154) when using results of regression analysis for
future demand forecasting, we should not be concerned about multicollinearity and therefore we
propose that results of the regression analysis with all three independent variables is used for
demand forecasting. In this case adjusted R2 shows that the explanatory power of the regression
equation is at 81.95%. The result shows that impact of pricing and existing production capacity
at the beginning of the year impact the demand for Vestas wind turbines as Vestas cannot lower
the price without reaching the production economies of scale. On the other side impact of
increase in global demand is softened and muted by limited production capacity and therefore,
Vestas market share is under pressure.

orderintakeMW = 164345 – 16102*sellingpriceinmEURperMW +

0.041*totaladdedonshorecapacityMW + 0.785*shippedpreviousyear

On the other hand, if we use regression coefficients for analyzing magnitude of independent
variable impact as well as for calculating elasticity coefficients it is important t-test passes for the
independent variable. Hence, when calculating price demand elasticity for the onshore wind
turbines we recommend regression coefficient of -17,198.312, which is result of regression
analysis reg5 as it has higher t-value than reg7.

Demand Function Estimation for Service

We have used regression analysis to estimate demand function for Vestas wind turbines’ service.
The demand for the corresponds to total amount of onshore wind turbines’ power serviced during
a calendar year expressed in Giga Watts (GW). We have extracted amount of serviced GW from
Vestas’ annual reports (Vestas Wind Systems, 2021a) covering period of 9 years (2012-2020).
Service price per GW for each year is calculated by dividing Vestas service revenue in EUROs
by service in GW and is presented in million EUROs. Vestas’ annual report (Vestas Wind
Systems, 2021a) is used as source for the following data: accumulated GW delivered by Vestas
and service order backlog. Total global onshore capacity in MW from 2009-2019 is taken from
Statista (n.d.) and Pitteloud (2018). The source for year 2020 is Vestas’ annual reports (Vestas
Wind Systems, 2021). All extracted values are presented in table below. Service in GW is our
dependent variable. Service price per MW in mE, accumulated GW delivered by Vestas at the
start of each calendar year, service order backlog at the start of each calendar year and global
onshore capacity in GW are four independent variables. Four graphs show scatter plots for each
of independent variables paired with order service in GW for the set of data from the table.
Our regression analysis hypothesis:

Hypothesis 1: The price of service per GW is an inverse determinant of the quantity of yearly
service of onshore wind turbines in GW (the sign of the coefficient is expected to be negative).
We expect that ability to service more onshore wind turbines is possible in very competitive
market only with contnuous downward price pressure.

Hypothesis 2: The accumulated onshore wind power in GW delivered by Vestas is a direct

determinant of the quantity of yearly service of onshore wind turbines in GW (the sign of the
coefficient is expected to be positive). We expect that demand for Vestas service follows the
changes in the accumulated delivered onshore wind power turbines. We assume that customers
who bought Vestas wind turbine prefer Vestas service of the turbines.

Hypothesis 3: The service order backlog in GW delivered by Vestas is a direct determinant of

the quantity of yearly service of onshore wind turbines in GW (the sign of the coefficient is
expected to be positive). We expect that increase in demand for Vestas service each year reflects
the service order backlog at the beginning of the year.

Hypothesis 4: The total global onshore wind power capacity in GW is a direct determinant of the
quantity of yearly service of onshore wind turbines in GW (the sign of the coefficient is expected
to be positive). We expect that demand for Vestas service may be impacted by the increase in
total global onshore capacity. However, we believe that this impact will be much less
pronounced due to Vestas decreasing market share and assumed customer preference for the
service delivery of the origina wind turbine producer.

Results of correlation analysis of four independent variables as well as regression analysis for
eight different combination of independent variables are shown here:

The results of the first, second, third and fourth simple regressions reveals that our hypothesis
correctly explained the nature of correlation between the demand for service of onshore wind
turbines power and our four independent variables with high 0.01 level of statistical significance.
The sign of the regression coefficients is not changed in multiple regression analysis (reg4, reg5,
reg6 and reg 7). However, once all four independent variables are included, service order
backlog at the beginning of the year and total global onshore capacity become inverse
determinant of the demand for service.

Due to correlation coefficient being higher than 0.7 there is multicollinearity between all four
independent variables. The correlation is slightly lower if one of the variables is service price
(around -0.80). The other three independent variables are very highly correlated (correlation
coefficient close to 1).

However, F-test has passed in multiple regression analysis reg5-8 that include these independent
variables. As expected, t-test for each regression coefficient is impacted due to reduced t-values
and it is harder to reject the null hypothesis in multiple regression analysis.
As per Keat, Young and Erfle (2014, pp.153-154) when using results of regression analysis for
future demand forecasting, we should not be concerned about multicollinearity. However, we
have chosen to avoid result of reg8 that includes all 4 independent variables as the sign of
regression coefficient has changed for two out of three highly correlated independent variables in
order to compensate for the multicollinearity. Our of 3 regression analysis that include service
selling prices and one of the other 3 independent variable we have chosen one with the highest
adjusted R2, the highest F value, the highest t-values and the lowest Root MSE, which is reg5. In
this case adjusted R2 shows that the explanatory power of the regression equation is at 98.84%.
We have also checked for autocorrelation in reg5 and reg8. Durbin-Watson test is inconclusive at
5% significance points. On the other hand, Breusch-Godfrey test results show that there is no
autocorrelation.
The results show that impact of accumulated Vestas onshore wind power in GW at year start
explains the change in demand with higher degree that the impact of service price. Hence, we
expect that the future demand for service will be driven mostly by total accumulated Vestas
onshore wind power.

serviceGW = 47 – 2.89*servicepriceperGWmE + 1.058*accumulatedVestasGWdel

On the other hand, if we use regression coefficients for analyzing magnitude of independent
variable impact as well as for calculating elasticity coefficients it is important that t-test passes
for the independent variable. Hence, when calculating price demand elasticity for the service of
onshore wind turbines we recommend regression coefficient of -2.894, which is also the result of
regression analysis reg5 as it has higher t-value than reg6-8.

Price Elasticity for Onshore Wind Turbines

To analyze the price of elasticity for the onshore wind turbines, we look at the data obtained
from the Vestas annual report between the years 2006 to 2020 (Vestas Wind Systems, 2021a).
We calculate the price of elasticity using the below equation from Keat, Young and Erfle (2014,
pp.69-72):

∆Q
Q
E p= avg
∆P
Pavg

Where ∆ Q is the change in quantity.

∆ P is the change in price.
 Elasticity of
Sl.no Year ∆Q/Qavg ∆P/Pavg Demand (Ep)
1 2006-2007 0,009667025 0,106394412 0,090860268
2 2007-2008 0,069807428 0,081253059 0,859135994
3 2009-2010 -0,648333517 -0,019952494 32,49385839
4 2010-2011 0,953767561 -0,049164208 -19,39963218
5 2011-2012 -0,158805227 -0,005053057 31,42755445
6 2012-2013 -0,657207005 0,02994012 -21,95071396
7 2013-2014 0,458874459 -0,045248869 -10,14112554
8 2014-2015 0,092740646 -0,092572659 -1,001814653
9 2015-2016 0,309808226 0,034387133 9,009423096
10 2016-2017 0,15959253 -0,013172338 -12,11573288
11 2017-2018 0,062944162 -0,128159906 -0,491137708
12 2018-2019 0,239306814 -0,064850843 -3,690111067
13 2019-2020 0,228288305 0,034255599 6,664262446
14 2005-2020 1,02507892 -0,179344465 -5,715698666

Looking at the above table, we can say that the elasticity of demand for terms 1,2,11 is inelastic
since Ep is less than 1. And for the term 8 we can say that it is unitary. And, for the rest of the
terms we can say that the price was elastic. And if we consider the demand curve between 2006
– 2020, we find that the demand is elastic. By comparing the results from the regression analysis
and using -17,198.312 as regression coefficient when calculating price demand elasticity for the
onshore wind turbines as recommended, we get:

QorderintakeMW = 164345 – -17,198.312 *sellingpriceinmEURperMW +

0.041*totaladdedonshorecapacityMW + 0.785*shippedpreviousyear

Using the data for the year 2020, we get

Q = 164345 – -17,198.312 *0.74 + 0.041*70604 + 0.785*12618 = 164486.93

Price elasticity = -16102*(0.74/165229.414) = -0.0770

Totaladdedonshorecapacity elasticity = 0.041*(70604/165229.414) = 0.0176

Shippedpreviousyear elasticity = 0.785*(12618/165229.414) = 0.0602

The above result shows that the future demand for order intake is inelastic and will be driven by
Vestas’ total added onshore capacity.
Price Elasticity for Services
To analysis the price elasticity for services, we used the data from annal report from 2012 to
2020 (Vestas Wind Systems, 2021a). First, we calculated the changes of percentages in Quantity
and Price in each year, and then compared also the first and last year.

Looking at the above table, the elasticity of demand between 2013 and 2012 is 0.4, an amount
smaller than one, showing that the demand is inelastic in this interval. This means that, along the
demand curve between year 2012 and 2013, if the price changes by 1%, the quantity demanded
will change by 0.4%. And when we look at the year between 2013 and 2014, the elasticity of
demand is 1,2. The magnitude of the elasticity has increased as we moved up along the demand
curve from year 2012 to 2013. Therefore, from the calculation, we can see that the demand of
Service between 2012-2013, 2016-2017, the price elasticity for services was inelastic. While the
rest of years, the price for services was elastic. Compare with the price elasticity for offshore in
2012-2013 and 2016-2017, which are both larger than 1. However, considering the result from
using regression coefficient, and per Keat, Young and Erfle (2014, pp.149-150), that the
formular for computing point elasticity is ∈x= ∆Q/∆X* (X/Q), where Q is the Quantity
demanded and X= any variable that affects Q.

QserviceGW = 47 – 2.89*servicepriceperGWmE + 1.058*accumulatedVestasGWdel. Here we

used the data from the year 2020, the results show as below:

Q= 47 – 2.89*17.56+ 1.058*129= 234.259

Price elasticity: -2.89* (17.56/234.259) = -0,216921

Total accumulated V elasticity: 1.058* (129/234.259) =0,5824

The result shows that the future demand for service is inelastic and will be driven mostly by total
accumulated Vestas onshore wind power.
Sales Forecast for Onshore Wind Turbines
Visual Time Series Projection
Visual time series projection has been chosen to perform sales forecast for both Onshore and
Services.

By using the annual sales of MW from 2006 to 2020, the observations appear the quantity up and
down in the beginning years, however the observations appear a straight line from year 2012 to
2019. At the last year, the order intake slight decreases. Without consider other reasons of
change, the past data gives us a good prediction of trend of the future, we can say that the sales
of offshore of 2021 will stay positive trend. We are forecasting the onshore order intake of year
2021 could be around 17000 MW.

Exponential smoothing technique

As per Keat, Young and Erfle (2014, pp.150-152), exponential forecasting can be done using the
formula below:

Et + 1 = wXt + (1 - w)Et

where w is the weight(smoothing factor) assigned to an actual observation at the period t.

We use this technique to predict the estimate the order intake for the year 2021. The following
table shows the exponential averaging with smoothing factors 0.2, 0.4 and 0.8. Comparing the
least squared errors for the three smoothing factors chosen above, we see that the forecast with
exponential smoothing factor 0.8 has less squared error and hence we can say that the predicted
order intake for year 2021 could be around 17202 MW.
 Smoothing Factor = 0.2 Smoothing Factor = 0.4 Smoothing Factor = 0.8
Absolut Squared Absolut Squared Absolut Squared
Year Actual Forecast e Error Error Forecast e Error Error Forecast e Error Error
2006 5559                  
2007 5613 5559 54 2916 5559 54 2916 5559 2643 6985449
201780, 192194, 173722,
2008 6019 5569,8 449,2 6 5580,6 438,4 6 5602,2 416,8 2
2009 3072 5659,64 2587,64 6695881 5755,96 2683,96 7203641 5935,64 2863,64 8200434
5142,11 3530,88 1246717 4682,37 3990,62 1592508 3644,72 5028,27 2528351
2010 8673 2 8 0 6 4 0 8 2 9
6278,62 1118,37 7667,34 270,345 73086,7
2011 7397 5848,29 1548,71 2398504 6 4 1250761 6 6 4
6158,03 2420,03 6725,97 2987,97 7451,06 3713,06 1378688
2012 3738 2 2 5856553 5 5 8927997 9 9 2
5674,02 289,974 5530,78 433,214 4480,61 1483,38
2013 5964 5 7 84085,3 5 8 187675 4 6 2200435
811,979 659311, 5704,07 839,928 705480, 5667,32 876,677
2014 6544 5732,02 7 1 1 9 5 3 2 768563
5894,41 3048,58 6040,04 2902,95 6368,66 2574,33
2015 8943 6 4 9293863 3 7 8427161 5 5 6627203
6504,13 3989,86 1591903 7201,22 3292,77 1084236 8428,13 2065,86
2016 10494 3 7 9 6 4 3 3 7 4267807
7302,10 3873,89 1500705 8518,33 2657,66 10080,8 1095,17
2017 11176 6 4 2 5 5 7063181 3 3 1199405
8076,88 6137,11 3766417 9581,40 4632,59 2146097 10956,9 3257,03 1060827
2018 14214 5 5 9 1 9 1 7 5 5
9304,30 8572,69 7349104 11434,4 6442,55 4150657 13562,5 4314,40 1861410
2019 17877 8 2 7 4 9 0 9 7 7
11018,8 6230,15 3881481 14011,4 3237,53 1048163 17014,1 234,881 55169,2
2020 17249 5 4 3 6 6 7 2 4 7
12264,8 15306,4 17202,0
2021   8     8     2    
32954,2 1,91E+0 24439,2 1,01E+0 15901,7 4434096
Total     6 8   3 8   6 4
4119,28 2386667 3054,90 1258438
Mean     2 4   4 0   1987,72 5542620

Econometric model
Using the result of our regression analysis we can estimate order intake in MW for 2021.

Shipped onshore wind turbines in MW for year 2020 is taken from Vestas 2020 Annual Report
(Vestas Wind Systems, 2021, p.6). It equals to 17055 MW. As per Vestas 2020 Annual Report
(ibid., pp.14-15) the total added onshore capacity in MW is expected to be between 60000 and
70000 MW in 2021. Hence, we will use 65000 MW in our forecast. For the sales forecast we
estimate selling price in mEUR per MW to be unchanged at 2020 level, that is 0.736 mE per
MW.

orderintakeMW = 16435 – 16102*sellingpriceinmEURperMW +

0.041*totaladdedonshorecapacityMW + 0.785*shippedpreviousyear

orderintakeMW (2021) = 16435 – 16102*0.736 + 0.041*65000 + 0.785*17055 = 20637 MW

We forecast order intake in 2021 to increase 19.6% compared to order intake in 2020. The
increase will be driven by added capacity to Vestas production due to significant increase in
production capacity in 2020 (35% over 2019 level).

Sales Forecast for Services

Visual Time Series Projection

Services
140
120
100
Giga Watts

80
60
40
20
0
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
Years

By using the annual sales of GW from 2012 to 2020 as seen in the graph above, we can say that
the linear trendline is increasing over time. And hence we can say that it has an upward trend
based on the past data, we can predict the annual sales for services could be around 135 GW.

Moving averages smoothing technique

This technique was used with the historical data from nine years of Service in Giga Watts (GW).
For the three forecasts, we did a projection for one year forward of the final period we got. In
this case forecast for the year 2021. Also, we calculated the Root Mean Squared Error (RMSE),
the methodology that will help us to identify which of the moving average we should use to
minimize the deviations (Keat, Young, & Erfle, 2014).

Table #1:
Graph #1:

Analyzing the results, we can conclude that the best moving-average forecast to be used for
2021, will be the one from the 3 rd Year with a value of 99,67 GW. With the smallest value of
RMSE (42,241 indicated in red in table#1) done from the three moving average forecasts.

Econometric model
Using the result of our regression analysis we can estimate demand service in GW for 2021.

Accumulated Vestas onshore wind power in GW for year 2020 is taken from Vestas 2020
Annual Report (Vestas Wind Systems, 2021a, p.6). It equals to 129 GW. For the service sales
forecast we estimate service selling price in mEUR per MW to be unchanged at 2020 level, that
is 0.736 mE per MW.

serviceGW = 47 – 2.89*servicepriceperGWmE + 1.058*accumulatedVestasGWdel

serviceGW (2021) = 47 -2.89*17.56 + 1.058*129 = 133GW

We forecast service demand to be 133GW with unchanged service price. With service price
decreasing 5% to 16.70 mEUR per GW, the demand will slightly increase to 135GW. Hence, the
increase for service around 15% is expected in 2021 and will be driven primarily by increased
accumulated Vestas wind power throughout the world. This growth is higher than the CAGR of
8-10% for period 2021-2025 forecasted in Vestas 2020 Annual Report (ibid., p.15).

References
Business, R. f. (2021, 02 21). Reference for Business. Retrieved from Vestas Wind Systems A/S
- Company Profile, Information, Business Description, History, Background Information
on Vestas Wind Systems A/S: https://www.referenceforbusiness.com/history2/85/Vestas-
Wind-Systems-A-S.html#:~:text=Vestas%20is%20one%20of%20the,Sea%20off%20the
%20Danish%20coast.

Gupta, A., & Singh Bais, A. (2020, 09). Wind Energy Market Size By Rating (≤ 2 MW, >2≤ 5
MW, >5≤ 8 MW, >8≤10 MW, >10≤ 12 MW, > 12 MW), By Component (Turbine,
Support Structure, Electrical Infrastructure, Others), By Installation (Onshore {By
Rating [≤ 2 MW, >2≤ 5 MW, >5≤ 8 MW, >8≤10 MW, >10≤ 12. Retrieved 02 21, 2021,
from Global Market Insingths - Insignths to Innovations:
https://www.gminsights.com/industry-analysis/wind-energy-market

Keat, P., Young, P., & Erfle, S. (2014). Managerial Economics. Edinburgh: Pearson Education
Limited.

Pitteloud, J.-D. (2018). Global Wind Installations. Retrieved 2021, from Wind Energy
International: https://library.wwindea.org/global-statistics/

Research and Markets. (2020, August). Europe Wind Power Market - Growth, Trends, and
Forecasts (2020 - 2025). Retrieved 02 21, 2021, from Research and Markets - The
World’s Largest Market Research Store:
https://www.researchandmarkets.com/reports/5176446/europe-wind-power-market-
growth-trends-and

Statista. (n.d.). Global onshore wind energy capacity from 2009 to 2019. Retrieved from Statista:
https://www.statista.com/statistics/476306/global-capacity-of-onshore-wind-energy/

Vestas Product Portfolio. (2021). Vestas Product Portfolio. Retrieved from Vestas:
https://www.vestas.com/en/products

Vestas Wind Systems. (2020). Vestas Annual report 2019. Denmark: Vestas Wind Systems A/S.
Retrieved from https://www.vestas.com/~/media/vestas/investor/investor
%20pdf/financial%20reports/2019/q4/2019_annual_report.pdf
Vestas Wind Systems. (2021). Vestas Annual report 2020. Denmark: Vestas Wind Systems A/S.
Retrieved from https://www.vestas.com/~/media/vestas/investor/investor
%20pdf/financial%20reports/2020/q4/210210_01_annual_report_2020.pdf

Vestas Wind Systems. (2021a). Reports and presentations. Denmark: Vestas Wind Systems A/S.
Retrieved from https://www.vestas.com/en/investor/financial_reports#!pr
%C3%A6sentationer

Appendix: Meeting Minutes

Meeting Minutes - Group 14 (IY2593, V21, lp3)

Date: 13/12/2021 18:00-18:45
Attendance: Solamerica Diaz Eiserman, Naga Thejus Sambasivan Mruthyunjaya, Mladen
Toncev, Kun Xue
Location: Canvas Conferences
Index Topic Decision Responsibility Timetable

1 Chosen Company Vestas Approved by all. 13/02/202

1

2 List of Tasks A list with 15 tasks is Approved by all. 13/02/202

issued. 1

3 Task's assignment Tasks 1,2,8 (s),12 Solamerica

4 Task's assignment Tasks 3,4,5,9 Mladen

5 Task’s assignment Tasks 6 (s), 7(o), 11 Kun

6 Task’s assignment Tasks 6 (o), 7 (s), 8 (o), 10 Naga

7 Task 3 When Task 3 is ready, Mladen

inform the rest of the team
and start tasks 6,7,8.

8 Next meeting Thursday 18th 18:00 Approved by all.

Task 1: Describe the firm, its goals and product offerings. (Sol)
Task 2: Describe market demand for the firm’s products. (Sol)

Task 3: Collect data on price and quantity and enter it to excel. Independent variable
is number of MW delivered per year. (Mladen)

Task 4: Decide on independent variables (for example: price per delivered MW,
sentiment towards green electricity in US, in China and in EU, availability of offshore
wind turbines, etc.) and enter it in Excel. (Mladen)

Task 5: Estimate the demand function for number of MW delivered, using regression
analysis data on price and quantity from over the years as a guide. (Mladen)

Task 6: Calculate price elasticity. How elastic is the demand for the firm’s products?
(Need task 3) Onshore turbines (Naga), Services (Kun)

Task 7: Perform a sales forecast using projections (visual time series projection) and
explain briefly. (need task 3) Onshore turbines (Kun), Service (Naga)

Task 8: Perform a sales forecast using smoothing techniques (with moving averages)
and explain briefly. (need task 3) Onshore turbines (Naga), Services (Sol)

Task 9: Perform a sales forecast using econometric model (result of Task 5) and
explain briefly (need task 3) (Mladen)

Task 10: Executive Summary. (Naga)

Task 11: Meeting Minutes. (Kun)

Task 12: References. (Sol)

Meeting Minutes - Group 14 (IY2593, V21, lp3)

Date: 18/12/2021 18:00-18:30
Attendance: Solamerica Diaz Eiserman, Naga Thejus Sambasivan Mruthyunjaya, Mladen
Toncev, Kun Xue
Location: Canvas Conferences
Index Topic Decision Responsibility Timetable

1 Task's assignment Tasks 1,2 Solamerica 21/02/2021

2 Task's assignment Tasks 3,4 Mladen Done

 3 Task’s assignment Tasks 5 Mladen 21/02/2021

4 Task 6 add compare with using Naga & Kun 21/02/2021

regression coefficient.

5 Task’s assignment Task 7 Naga & Kun Done

6 Task’ assignment Task 8 Naga & Sol 21/02/2021

Next week

7 Task’s assignment Task 9 Mladen 21/02/2021

8 Task 10 Can be fully done after tasks Naga

1-9 are ready.

9 Task 12 Add references. Approved by

all.

10 Next meeting Thursday 22nd 18:00 Approved by

all.

Meeting Minutes - Group 14 (IY2593, V21, lp3)

Date: 22/12/2021 18:00-18:20
Attendance: Solamerica Diaz Eiserman, Naga Thejus Sambasivan Mruthyunjaya, Mladen
Toncev, Kun Xue
Location: Canvas Conferences
Index Topic Decision Responsibility Timetable

1 Task's assignment Tasks 1,2 Solamerica Done

2 Task's assignment Tasks 3,4 Mladen Done

3 Task’s assignment Tasks 5 Mladen Done

4 Task 6 add compare with using Naga & Kun 25/02/2021

 regression coefficient.

5 Task’s assignment Task 7 add conclusion for Naga & Kun 25/02/2021
coming year 2021.

6 Task’ assignment Task 8 Naga & Sol 25/02/2021

7 Task’s assignment Task 9 Mladen Done

8 Task 10 Can be fully done after tasks Naga

1-9 are ready. (25/02/2021)

9 Task 12 Add references. Approved by

all.

10 Next meeting Thursday 25th 18:00 Approved by

all.

Meeting Minutes - Group 14 (IY2593, V21, lp3)

Date: 25/12/2021 18:00-18:20
Attendance: Solamerica Diaz Eiserman, Naga Thejus Sambasivan Mruthyunjaya, Mladen
Toncev, Kun Xue
Location: Canvas Conferences
Index Topic Decision Responsibility Timetable

1 Task's assignment Tasks 1,2 Solamerica Done

2 Task's assignment Tasks 3,4 Mladen Done

3 Task’s assignment Tasks 5 Mladen Done

4 Task 6 add compare with using Naga & Kun 26/02/2021

regression coefficient.

5 Task’s assignment Task 7 add conclusion for Naga & Kun Done
coming years.
6 Task’ assignment Task 8 Naga & Sol 25/02/2021

7 Task’s assignment Task 9 Mladen Done

8 Task 10 Can be fully done after tasks Naga 27/02/2021

1-9 are ready. (25/02/2021)

9 Task 12 Add references. Approved by 26/02/2021

all.

10 Next meeting Thursday 28th 09:00 Approved by

all.