Date: 19.07., 9:05 - 11:05
Format: open book, no camera
Description: 'I try not to have a too difficult exam', questions about interpretability, because of open book format problems are not gonna be 1:1 from the notes, 'grading your understanding', said the word likelihood more than anything else, 'multidimensional problem with linear model [...] CAPM [...] ARMA [...] (G)ARCH', 'I'd like to stay in the ARMA-GARCH world', download exam from ilias, scan our exams and upload them afterwards, 'go for the interpretation of the CAPM', 'challenging questions good way to go, since I don't post solutions for that'
ADD： A typical exam looks as follows: I describe a financial data science problem and ask you to solve it. The solution asks you to write down explicitly objective function (often likelihood based) and to explain how to learn the necessary parameters. You might need the finance intuition from above to translate the problem into math (stats) equations. Then, you will need the data science skills (mainly OLS/MLE/ARMA-GARCH) to solve the problem. Last, you need again some finance intuition to explain what you have done and found and how it helps to solve the problem. Roughly 80% is based on OLS/MLE/ARMA-GARCH, the rest on intuition like CAPM etc. Yet, if you do not know about basic concepts such as CAPM and the respective equations, you might face a problem right at the beginning to translate the financial data science problem into equations. Although this is a risk, it has not yet occurred in any of my exams.
Point scale:
Points for 1.0:
Points to pass:
Important themes:
Unimportant themes:
Constraints: