Welcome to my personal website!
My research focuses on cross-sectional dependence in applied econometrics and, more specifically, on the role of social interactions and networks in economic outcomes.
I obtained my Doctorate in Economics from the Chair of Econometrics at Goethe University Frankfurt under the supervision of Professor Dr. Horst Entorf.
Since late 2021 I work only in settings that I can return to and test my econometric models. Why? Because I have been fortunate enough to realize that results obtained for the sake of academic research alone do not necessarily bear any predictive value once applied in the real-world. This type of research usually requires that the data remain proprietary - understandably so. Exactly because econometric findings have real consequences, I have concluded that the best methods are the simplest ones, e.g. linear regression, and never GMM or ML (there is greatness in simplicity).
Since July 2023 I work on econometric models that predict asset prices and related variables. The most pressing problems in the EU economy today are debt, pensions and demographics. Sooner or later we will have to rely on the stock market to solve all of the above. My solution is to identify high probability settings to reduce the risk of investing or trading in the stock markets as well as the time horizon needed to see any compounding effects. I aim at a 5% steady monthly return after the first year of investing in my models.
The models are proprietary because - to the best of my knowledge - I cannot patent them. They perform well because they use behavioral econometric techniques to explain economic outcomes as opposed to techniques mathematicians or physicists employ. Higher math is of no use as first and second derivatives can adequately describe economic behavior. Moreover, macroeconomics is of no use as what matters is certain shocks in certain macroeconomic variables.
I assume there exist different types of investors and settings in which they act. What is their incentive? To generate positive returns. How do they accomplish that? By selling an asset at a price higher than the one they bought it. Although theoretically there can be infinite such types, in practice the number depends on what multicollinearity allows for. Additionally, although there are infinite settings, only a finite number of those realizes. I know which type is more likely to dominate in each setting and assign weights accordingly to determine whether buying or selling dominates during a trading session, i.e. whether the signal is bullish or bearish. For trading, I consider signals with probability at least 75%.
Unlike academic research, these research models have real consequences, i.e. lose capital. Therefore, ideally I should experience a bear market before launching them commercially to see if a new type of investor emerges. If not, I would like at least five years of application which places launching at the end of 2028.
I work only with data that are publicly available, i.e. have no superior information. Thus, what is important here is not the data per se but what you do with them. As proprietary models cannot be replicated, I post predictions online anonymously* to keep a record of their performance. You are kindly invited to read more in my "Research" section (I will gradually upload full papers with results).
Curriculum Vitae
Research (New 2026, 2025)
Contact
Not available at the moment unless you already have my contact information (innovation requires isolation).
Cannot take new projects (23.04.2025).
*Please bear with me. I still have privacy and security issues to solve.