Scribd
Upload
60 views2 pages

Quantitative Methods in Finance Msc. in Finance Fall 2020: Instructions

This case study involves estimating market risk for a portfolio consisting of 40% S&P 500 and 60% NASDAQ over a 1-day period using various risk measurement procedures. Students will: 1) Calculate value at risk (VaR) using historical simulation and a parametric Gaussian approach to estimate 5% and 1% VaR over 2005-2020. 2) Perform back-testing of the VaR models by computing exceedances, removing NAs, and reporting the number of violations based on Kupiec's test at a 95% confidence level. 3) Compare the back-testing results when the portfolio allocation is changed to 60% S&P 500 and 40% NASDAQ.

Uploaded by

Alvaro Ramos Rios
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
60 views2 pages

Quantitative Methods in Finance Msc. in Finance Fall 2020: Instructions

This case study involves estimating market risk for a portfolio consisting of 40% S&P 500 and 60% NASDAQ over a 1-day period using various risk measurement procedures. Students will: 1) Calculate value at risk (VaR) using historical simulation and a parametric Gaussian approach to estimate 5% and 1% VaR over 2005-2020. 2) Perform back-testing of the VaR models by computing exceedances, removing NAs, and reporting the number of violations based on Kupiec's test at a 95% confidence level. 3) Compare the back-testing results when the portfolio allocation is changed to 60% S&P 500 and 40% NASDAQ.

Uploaded by

Alvaro Ramos Rios
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Quantitative Methods in Finance

MSc. in Finance
Fall 2020
Instructor: Romain Deguest

Case Study 3: Market Risk Measurement

In this assignment, you are the head of the risk unit of a small asset management company. The portfolio of
the AM company is invested in two equity indices as follows: 40% in the S&P 500 and 60% in the NASDAQ,
both denominated in US dollars. You are concerned with risk measure estimation over a 1-Day period and
computed with levels equal to 5% and 1%. you will have to implement several risk measurement procedures
to estimate the risk of your portfolio over a 1-Day period.

Instructions
This case study will be started during the online group work session following the lecture. Your group is
expected to work on this project during the session and at home before submission. You will be able to ask
any questions to help you complete the case study during the group online work sessions, the Q&A online
sessions, and on the Blackboard discussion board.
Assignments must be submitted using the R template provided (in both the .Rmd format and the .pdf
format obtained by using the Knit button in RStudio), otherwise assignment will receive a 0.
Late submissions will receive a 0.

1 VaR Computation (60 marks)


a. (15 marks) Download the S&P 500 and NASDAQ indexes from yahoo finance over the period between
2005-06-29 and 2020-07-01 and compute their daily arithmetic returns. Compute the arithmetic returns
of your portfolio and, from there, compute the evolution of $1 invested in that portfolio and plot it in
a graph together with the other indexes.
b. (10 marks) Compute the average returns and volatility (using arithmetic returns) of your portfolio
1
over the entire data-set (use ∆ = 252 for the annualization). Compare it to those of the indexes and
comment your results.
c. (10 marks) Compute the historical VaR of your portfolio returns based on a lag of size n = 100
observations at level 5% and 1%. Plot both historical VaR over the entire data-set on the same graph
and comment your results.
d. (15 marks) Consider now a parametric approach to estimate the VaR based on the Gaussian dis-
tribution. Compute the parametric VaR of your portfolio returns by assuming that the mean of the
log-returns is 0 (relative VaR) and its volatility is given by the GARCH(1,1) model. Plot both Gaussian
parametric VaR at level 5% and 1% over the entire data-set on the same graph.
e. (10 marks) Compare the two VaR at level 1% on the same plot and comment your results.

1
2 Back-testing VaR Models (40 marks)
a. (10 marks) Compute the time-series of exceedances for each VaR estimates computed previously. Then
remove all the NA values from your time series of exceedances, so that you have a data for each date
and each VaR. Report the number of violations that occurred over the common dates for each VaR
estimates computed previously.
b. (15 marks) Run Kupiec’s test to check if you can validate both models with a confidence interval of
95%. Justify your conclusion.
c. (15 marks) Compare the results of the Kupiec’s test to the case where your portfolio is invested as
follows: 60% in S&P 500 and 40% in NASDAQ.

You might also like