Analyzing Data  [20.10.22]

Financial markets may be his specialty, but his methods of data analysis also open up a wide range of possibilities for interdisciplinary projects: Prof. Dr. Thomas Dimpfl has been Director of the Department of Business Mathematics and Data Sciences since last August.

 

But although - or precisely because - Big Data is his bread and butter, he also likes to take unexpected paths in teaching: The business mathematician is a self-confessed fan of the good old chalkboard.

 

Mr. Dimpfl, you have been at Hohenheim since August of last year. How was your start?

By and large, the start went quite smoothly. Due to the Covid restrictions, everything was certainly a bit different than normal, but I experienced a lot of support, which made it easier to settle in.

The start, if you will, is not quite complete yet, though. We are still in the process of finalizing the teaching program, and the last vacant staff position will also not be filled until the winter semester of 2022/23. But now the team is complete and we can get started in teaching and research.

 

Your department is Business Mathematics and Data Sciences. What exactly do you mean by this?

Strictly speaking, these are two fields, but they complement each other very well. Business mathematics refers to mathematical methods applied to problems in economics and business administration. This ranges, for example, from interest calculation for the building society contract to stochastic differential equations in large economic forecasting models of the central banks. Therefore, in the first semester, we try to cover a wide range of methods that will become relevant later on.

And this is where the bridge to data science begins, which itself is a conglomerate of mathematical statistics, econometrics, and computer science from the point of view of specific applications and problems. In the data sciences, we develop methods that respond to the current challenges in terms of data generation, data availability, and, in particular, data volume - the buzzword being "Big Data".

 

What fascinates you about this?

Both areas include methods that are used in all kinds of specializations. Personally, I find financial markets exciting, which is why I put a lot of emphasis on equity and crypto markets in my research. This is one of the reasons why the department is assigned to the Institute of Financial Management.

The nice thing, however, is that you can also do something different with the methods and dive into exciting new research fields in the process. For example, I conducted a study with colleagues from Tübingen and Leipzig on the mortality rate of COVID-19 in the first wave. We learned a lot about the virus, diagnostics, and testing strategies and benefited enormously from the medical knowledge of the Leipzig colleague. That was very interesting. And the connecting thread, the question of how to analyze data, makes such interdisciplinary projects possible.

I am currently working on two more projects in which I am primarily contributing my data science knowledge and benefiting from the specialist know-how of my colleagues. That is always exciting and, like everything new, thrilling.

 

What was your personal path to the professorship in Hohenheim?

In 2001, I started studying International Business Administration in Tübingen, with specialized language training in French and Spanish. I chose Tübingen at that time because, from what I could find out as a student, the course was supposed to be "math-heavy." There I graduated in 2006 with a degree in business administration and started a doctorate in econometrics.

In the middle of 2007, I left Tübingen with my supervisor for Erfurt, where I completed my doctorate in the middle of 2010. After that, I went back to Tübingen, initially as a post-doc with the goal of habilitation. In 2016, I received the the habilitation with the authorization to teach financial econometrics and from then on I was a Privatdozent. At the end of 2019, I was finally appointed as an associate professor in Tübingen before coming to Hohenheim in August 2021.

 

What research topics are you currently working on?

You could call it a colorful jumble connected by the common thread of data science. In the field of cryptocurrencies, for example, we are currently investigating the question of how strongly the herd behavior of investors in these markets influences the pricing of various coins. Not entirely surprisingly, Bitcoin again plays a central role in this as the largest cryptocurrency, which simply attracts the most attention and thus dominates the overall market.

Another research project revolves around "safe haven assets," i.e., investment assets that are supposed to represent a safe haven in the event of crises. Gold is usually the first to be mentioned here. This raises the question of how to find such an investment, what is suitable for it, and when would I need to buy it in the first place?

The students are already able to work on these projects. Last summer, for example, in the course "Basics of Computational Sciences," we assigned topics related to this for student research questions together with Professor Vogelgesang. The posters now hang in the hallway outside my office, and I'm delighted to see the results every time I walk past them.

A somewhat more methodological current research project looks at how important a particular market is for pricing a good when it is traded in multiple markets (e.g., secondary listing for stocks), or for otherwise closely related goods (e.g., a stock index, ETFs, and futures based on it). There is a famous measure for this, but it only works when both markets trade in parallel. We are working on an extension that removes this restriction so that markets can open and close when they want, which in reality they already do because of different time zones.

 

Let's move on to teaching: What does good teaching mean to you?

My teaching was good if the students still know what they learned from me several years later and can say in retrospect that it was relevant. Clearly, not everyone needs everything all the time. But especially in mathematics, you learn not only to calculate, but also to think in a structured way.

I think the central aspect of my teaching approach is to help students understand in depth the things we cover in lectures. To do that, I have to try to arouse their curiosity. The question of why, which was so popular as a child, is once again respectable at the university. This concept certainly comes at the expense of breadth. Especially in data science, it is important to me that students understand and are able to assess the methods and don't just leave the University of Hohenheim as a "plug and play data scientist."

Otherwise, those who know me from the first semester might have already noticed that I am a fan of the chalkboard. It's much better to develop concepts over 45 minutes than by scrolling back and forth on the screen. Of course, there are slide sets, and one or two visualizations come directly from the computer even in business math. But no readable-size image comes close to four blackboards when it comes to showing the solution conditions for systems of linear equations.

On the other hand, we are in the process of setting up a Jupyterhub for teaching, so that I can work with the students in the Data Science courses in such a way that they only need a laptop and browser. It just always depends on what fits at the moment.

 

Do you have any tips for successful studying?

Stay curious and open-minded so that you can leave the paths you have taken. After the first year of basic studies, I was actually quite happy that the statistics lectures were over. I could never have imagined doing a thesis at a department of statistics and econometrics. It was only when I stumbled across econometric applications again and again during my studies that my interest was aroused. And all of a sudden, statistics became fun. I think that's the most important thing: you have to find out for yourself what you like and enjoy. Sometimes you have to take detours, but they're worth it.

 

What do you do when you're not working?

When I'm not working, you'll find me either in the kitchen baking or jogging in the forest or climbing a tower. Depending on my mood, I also like to spend my time with friends, reading a good book, or in front of the Playstation.

 

Thank you very much, Mr. Dimpfl!

 

Interview: Elsner / Translation: Neudorfer