Professional Documents
Culture Documents
Introduction
The aim of this project is to have you work in a group on a subject that requires both
the knowledge and application of digital tools.
The project must be done in the groups assigned by administration. Each group will
have a project manager who shall supervise the distribution of tasks and the
progress of the work. To do this you will use Trello, Google Drive and any other
collaborative tool that contributes to the good progress of the project.
The work will take the form of a report (one per group) where the quality of the
presentation will be particularly evaluated (as you will be using Word). Citations
should be in Harvard or MLA style.
Project Presentation
The focus of the project will be to study the performance of two CAC 40 companies.
Question 2.
P t−P {t−1 }
The formula for calculating the daily return on day t is made by Rt = in which
P {t−1 }
{t −1 } represents the day prior to day t .
Thus, in the database, one can only calculate yields from 2 January 2019 onwards
(as we have the quotation price starting from the previous day, i.e. 1 January 2019).
We are unable to calculate the yield from January 1, 2019 since we do not have the
price from December 31, 2018.
For each share and each corresponding index calculate the daily returns.
Comment on the range of variation. Express the values as percentages (and state
between which maximum and minimum percentages the returns may vary).
Question 3: Make a table for each share and for the corresponding reconstituted
index.
For the 3-month period show:
The number of observations (quotations) each month.
The monthly average of the share's returns and the monthly average of the
corresponding index.
The minimum and maximum returns (share and corresponding index) for each
month.
The standard deviation of the returns (share and corresponding index) for each
month.
The coefficient of variation (share and corresponding index) for each month. The
standard deviation
coefficient of variation is equal to .
mean
Question 4.
Create a graph comparing the average monthly return of the share and the average
monthly return of the corresponding index, which you have calculated in the previous
question for each of the three months.
Comment on this.
In this section we will introduce the market model. This model consists of explaining
the return on a company's share based on the performance of the CAC 40 index.
For example, for some companies their return increases significantly when the return
of the CAC 40 index increases slightly. For others, their return decreases when the
return of the CAC 40 index increases.
R A =α^ + ^β R M
^
Where R M is the exact return of the A-CAC39 index, which is referred to as the
market return; and ^
R A is the return estimated by our A-share model.
RA=^
R A + ε in which ε is an error term.
notary fees, your estimated price for a flat with a surface area of S would be:
^
P=20,000+ 3,000× S
Applying the numeric value of S=100m2 would give an estimated price of 320,000
euros.
However, the real price for this flat is 170,000 euros, as it is located in a less-
desirable area of the city. So, there is an error term (which is not related to the
surface area) which will explain a variation between reality and the model.
Of course, there are different error terms from one flat to the next. That is why each
error term (independent of the model) is specific to a flat and for this reason it is
called an idiosyncratic error.
Problem: How to find the model? We will do this for Kering and then you will repeat it
for other companies.
Answer by using the following procedure:
To select a range of estimation here, we will take the month of January. So, in the
example for the Kering company (the returns calculated for that month) in column D
we have the daily returns for Kering; in column E the market returns (return of the
CAC 40 index reconfigured without Kering).
Figure 1
Then copy and paste (pay special attention when pasting the values!) the Kering
yields which are on the left to put them on the right of Kering-CAC39 in order to insert
the distribution in a scatter chart (so that XY will have X as Kering-CAC39 and Y will
show Kering, which must be on the right).
Then select the 22 recorded Kering-CAC39 yields and the 22 Kering yields
(remembering to select the headers as well) to insert them into the chart as follows:
Figure 2
Finally, once you click the first suggestion for the scatter chart in the top left corner
and the chart appears, then press the plus to then select the trend curve then other
option to then display the equation in the graph. Thus, we obtain Figure 3:
Figure 3
In this way the yield model for Kering is created (using the month of January or what
is estimated from January data). The model is given by the equation y=1.378x-
0.0014
One then recognizes α^ =−0.0014 and ^β=1.378 and the explanatory model of the yield
performance is:
R Kering =α^ + ^β R M
^
In fact, EXCEL traces the line that passes closest to the observed data (the blue
dots). Reality is shown by the blue dots. The model is the line. Since this line is the
least distant from the blue points, the distance or error between the model and reality
is minimized.
Each blue point has the following coordinates:
On the X-axis (the horizontal line) is the market yield
Y-axis (the vertical line) shows the return on a share of Kering
Thus, the blue point connected by the red line has the following coordinates
(2.08%;3.56%) and corresponds to the returns actually observed on 08/01/2019, i.e.
for the Kering-CAC39 return of 2.08% and for the Kering share return of 3.56%.
Our model explains the return on the Kering share by looking only at the return on
the CAC 40 index, in this case the Kering-CAC39 return.
Thus, on 08/01/2019 we see that R M =2.08 % from which we deduce an estimate
(using our model) of the return on 08/01/2019 of the Kering share, i.e.:
^
R Kering =−00014 +1.378× 2.08 %=2.73 %
The circumflex above the R means that this is what our estimate, our model, gives
us. The black point on the right has the coordinates:
( R M =2.08 % , ^
R Kering =2.73 %)
Similarly, in the example of the flat for the same 100m 2 surface area of the flat we
wanted to buy there is the price we had estimated of 320,000 euros and the real
price 170,000 euros.
Question 5. Estimate the model of the returns on Sanofi and Société Générale
shares according to the market return for the month of January using the same
approach as for Kering.
Question 6. Create two new columns for each share. The first column will be the
returns estimated by the model and the second will be the error terms.
We ask you to show for each share present the mean and the standard deviation of
the errors for the month of January.
In the graph below we have in column D the actual recorded Kering returns and in
column F the returns estimated by the model. In column G are the error terms
Figure 5
Question 7: We now will consider the month of March. For each share you have a
template for explaining the share performance (you created this template earlier).
Each day in March you have the market return so you can estimate the return of
each share (respectively Sanofi and Société Générale) for each day with your model.
For each day of March, you can therefore calculate the difference between the
observed share performance and the estimated share performance. This difference
is called an abnormal return. Finally, you divide this difference by the standard
deviation (corresponding to the share) calculated in question 7.
So, each day and for each share you have a T defined by:
^
R Kerring − R Kerring
T=
σ erreurKerring
in which σ erreurKerring is the standard deviation of the error terms for Kering calculated in
question 6.
Make a chart for each share showing all the dates when the absolute value of T is
greater than or equal to 2