Scribd
Upload
7 views2 pages

Problem Set 2

The document presents a problem set focused on multiple linear regression in the context of financial analysis, specifically examining the relationship between news articles and stock returns. It includes tasks such as interpreting coefficients, conducting significance tests, constructing confidence intervals, and evaluating causality. Additionally, it discusses the impact of adding more variables to improve explanatory power and how to test for improvements in the regression model.

Uploaded by

Haroon Jamshed
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
7 views2 pages

Problem Set 2

The document presents a problem set focused on multiple linear regression in the context of financial analysis, specifically examining the relationship between news articles and stock returns. It includes tasks such as interpreting coefficients, conducting significance tests, constructing confidence intervals, and evaluating causality. Additionally, it discusses the impact of adding more variables to improve explanatory power and how to test for improvements in the regression model.

Uploaded by

Haroon Jamshed
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

MSc F&A: Applied Econometrics Problem Set 2

1 Multiple Linear Regression

A financial analyst is studying the relationship between the number of news articles about
a company (x2 ) and the daily (log) return on the company’s stock (y) using the below
simple linear regression model. Note that the log return has already been multiplied by
100.
yi = β1 + β2 x2i + εi (1)

The regression output gives the following results:

• β̂2 = 0.45

• SE(β̂2 ) = 0.12

• Sample size: N = 500

• RSS: 100

1.1 Interpreting Coefficients

How would you interpret the slope estimate specifically (make sure to specify the units)?

1.2 Significance Test

Test whether there is a statistically significant relationship between the number of news
articles and the stock’s daily return at the 5% significance level. Make sure to specify the
null hypothesis.

1.3 Confidence Interval

Construct a 95% confidence interval for β2

1.4 Causality

Is this causal? Why or why not?

1.5 Momentum

The analyst is told that momentum (x3 ) which is measured as the stock return over the
past month is really important for predicting stock returns. So the analyst updates the
regression model:

yi = β1 + β2 x2i + β3 x3i + εi (2)

1
MSc F&A: Applied Econometrics Problem Set 2

The regression output gives the following results:

• β̂2 = 0.045, SE(β̂2 ) = 0.025

• β̂3 = 0.35, SE(β̂2 ) = 0.1

• Sample size: N = 500

• RSS: 70

What is the confidence interval for β̂2 Explain your result.

1.6 Improving Explanatory Power

The financial analyst is now curious about whether they should many more variables in
case they missed something else. The updated regression is now:

yi = β1 + β2 x2i + β3 x3i + β4 x4i + β5 x5i + β6 x6i + β7 x7i + β8 x8i + εi (3)

where x2i is number of articles, x3i is momentum, x4i is volatility, x5i is earnings growth,
x6i is market sentiment, x7i is trading volume, and x8i is treasury yields.

The RSS from this regression is 40 and the sample size is still 500.

The analyst wants to test whether the addition of these 6 variables has improved upon
their original one variable model. How can they do this? Make sure to specify your null
hypothesis.

You might also like