Assignment 1
Assignment 1
Assignment 1
Professor: Norman Schürhoff
Due date: 5 November 2023
f : R → R, f (x) = ex .
4. Use one of the proof methods seen in class to prove the following propositions:
(a) Let A, B ⊆ RN be convex sets. Then A ∩ B is convex. Prove that A ∪ B doesn’t have to be
convex.
(b) Prove the following statement for n ≥ 1 using mathematical induction:
n
X n(2n − 1)(2n + 1)
(2k − 1)2 =
3
k=1
1
2. Now drop the 11th asset and answer the following:
(a) What is the average daily return of an equally weighted portfolio?
(b) What is the variance-covariance matrix?
3. Now consider a factor analysis of the returns. Factor analysis allows for dimension reduction by
retaining only the top eigenvalues (factors) that capture the most significant variation in the data.
This reduces the dimensionality of the problem while preserving the essential information.
(a) Decompose the variance-covariance into its eigenvalues and eigenvectors.
(b) Show that each eigenvector represents a linear combination of the original variables, and each
eigenvalue represents the variance explained by its corresponding eigenvector.
(c) Perform factor analysis using 1, 2, 3, 4, 5, and 6 factors without rotation. How many factors
are actually required to perform factor analysis? Give reasons. (Hint: Check eigenvalues. The
number of factors is the number of eigenvalues exceeding 1. This is known as the Kaiser Rule.
The logic is that a factor with eigenvalue greater than 1 explains more variance than a single
observed variable.)
(d) Factor rotation helps to make the factors more interpretable. Varimax rotation is an orthogo-
nal rotation in which assumption is that there is no intercorrelations between components. In
contrast, oblique rotation is more complex, and can provide more superior results. See https://real-
statistics.com/linear-algebra-matrix-topics/varimax/. Perform factor analysis using the required
number of factors with varimax rotation. Report the factor loadings. Interpret the factor loadings.