PCA applied to european yield curves: a comparison between the positive and negative interest rate environments

• This is the repository of my code of my Bachelor Thesis from the Freie Universitat Berlin.

• I graduated in November 2020.

• This thesis won an award: Best Econometrics Thesis of the Year in 2020 Deutsche Bundesbank Award.

---

Above you will find 5 jupyter notebook files, you can simply click on them and read the code and the compiled code already all there.

There is a PDF file, that is the file to my Bachelor thesis titled "Principal Component Analysis applied to the Term Structure: an analysis on European interest rate yields between 2004 to 2020”.

What this bachelor research paper is about:

In this project I applied PCA to interest rate yield curves. There I computed PCA to the changes of average yields of AAA-rated bonds in the Euro Area in a positive (with sample of 2004 to 2007) and in a negative interest rates environment (with a sample of 2017
to 2020) to assess whether the amount of variability explained by the first 3 PCs would lower significantly due to the negativity of interest rates.

Contrarily to recent research (Lazarevic, M., 2019. Principal component analysis in negative interest rate environment. Acta Oeconomica, 69, 101–125.) that used a less negative sample for the negative interest rates environment (2014 to 2017), from my research the negativity of interest rates does not result to be impactful in the explanatory efficacy of PCA applied to the term structure.