Promotions

Doctorate in the field of "Challenges in virtual leadership" at the University of Osnabrück

Doctoral period: since 2021

PhD in the field of "Algorithmic Decision-Making, Forecasting and Economic Behaviour in Financial Markets" at the Georg-August-University Göttingen

Doctoral period: since 2024

PhD in the field of "Biases and Heuristics in Portfolio Management - Determinants for non-optimal Portfolio Diversification" at the Georg-August-University Göttingen

Doctoral period: 2014 - 2019

Dissertation available here (external link, opens in a new window).

Short description

Recent crises, such as the financial and economic crisis or the debt and currency crisis, have shown that neoclassical economics has lost its credibility. The characteristics of homo economicus, who acts in a self-interested and rational way, maximises his own utility and has complete information, has obviously not appeared among capital market actors during the recent crises.

An empirical and experimental research approach will be used to shed light on the behaviour of capital market players. The focus is on the following research topics: 1) "Overconfidence - the effects of positive and negative affect". The research question here is whether the influence of emotions on self-assessment is robust to learning effects. It is examined whether repeated self-assessments lead to individuals being able to detach themselves from overconfidence. 2) "Diversification behaviour of capital market players - the effects of overconfidence". In this context, the effects of positive and negative emotions on the diversification behaviour of economic agents will be examined. 3) "Diversification behaviour of capital market players - the effects of the gambler's fallacy". The aim is to investigate the effects of the gambler's fallacy on the diversification behaviour of capital market players. The dissertation will be written cumulatively.

PhD in the field of "Algorithmic Decision-Making, Economic Behaviour and Predictability in Financial Markets" at the Georg-August-University Göttingen

Doctoral period: 2020 - 2023

Dissertation available here (external link, opens in a new window).

Short description

The digital transformation is producing more and more technical innovations that have an impact on our daily lives. In a variety of areas, economic actors increasingly have the opportunity to interact with algorithms, as the range of robo-advisors shows, for example, and thus also to influence events on the financial markets. The aim of this study is to investigate the behaviour of economic actors in dealing with algorithms and their willingness to use them in order to contribute to a better understanding of algorithm aversion. Algorithm aversion describes the negative attitude towards the use of algorithms that economic actors often develop as soon as they realise that algorithms are superior but not error-free. The first part of this paper comprises five experimental studies. The first contribution shows that algorithm aversion can be partially reduced in repeated tasks by increasing experience over time. The second article deals with the scope of a decision and shows that algorithms are often not used, especially in situations that can have serious consequences in the event of an error, even though their use has a higher chance of success. The third article shows that possible interventions by a user in the forecasting process reduce algorithm aversion more reliably if they are applied to the forecast result (output of the algorithm) rather than to the configuration (input of an algorithm). The fourth article analyses the influence of proxy decisions on algorithm aversion. However, making decisions for third parties does not lead to a reduction in the extent of algorithm aversion. The fifth article shows that the decision behaviour to use an algorithm varies with the previous usage rate of other economic agents and that previous high acceptance results in more frequent use of an algorithm than previous low acceptance. Overall, algorithm aversion proves to be extremely robust and can contribute to suboptimal decisions. Overcoming algorithm aversion is essential in order to utilise the great potential that technical innovations offer for forecasting. Two further studies form the second part of this paper, which are part of the literature on the quality of capital market forecasts. While the sixth article analyses the quality of interest rate forecasts in Latin America, the seventh article deals with stock market forecasts for three important indices. Overall, the capital market forecasts analysed are inadequate in most cases. While the forecasts in the Latin American region largely reflect current rather than future interest rate developments, the equity index forecasts show that the majority of equity market analysts underestimate the variability of reality and tend to be conservative. It is therefore crucial to improve forecasting models and react more flexibly to new developments.

PhD in the field of "Financial Market Actors: Cognitive Biases, Portfolio Diversification and Forecasting Abilitx" at the Georg-August-University Göttingen

Doctoral period: 2016 - 2019

Dissertation available here (external link, opens in a new window).

Short description

Portfolio theory (Markowitz, 1952) plays an important role for investments in the capital market and still forms an important basis for the structuring of investment portfolios. For risk-averse investors, diversification of portfolio components is a sensible strategy. This always applies when the future price trend is uncertain. Diversification only makes no sense if price movements are easily predictable. In this case, the optimal strategy would be to invest exclusively in the security with the highest expected increase in value (Markowitz, 1991). In practice, however, investors continue to hold under-diversified portfolios, which contradict the principle of portfolio theory. The first three studies of this cumulative dissertation deal with different aspects of portfolio management. In the first study, the suitability of fine wine as a diversification tool is analysed using a simulation calculation. The results for the specified periods are sobering. The inclusion of fine wine leads to a slight improvement in the annualised return on an index basis, but to a significant increase in risk. When looking at the real investments, the considerable costs of investing in fine wine come into play. The second article examines the influence of herd behaviour, the status quo bias and the gambler's fallacy on diversification behaviour. Neither herd behaviour nor the gambler's fallacy contribute significantly to a non-optimal portfolio decision. However, the gambler's fallacy plays an important role in these decisions. Many participants are eager to find patterns in a sequence of random events and to deduce the sequence of future events from these patterns. And in the third study, a method for improving the measurement of participants' risk preference is presented. Experimental research into diversification behaviour requires a clear distinction between risk-averse, risk-neutral and risk-taking economic subjects, as decisions that may make sense for a risk-taking economic subject are completely unthinkable for a risk-averse economic subject and vice versa. The methods used to date to determine risk preference have a number of weaknesses. Therefore, a new method for determining risk preference is proposed in the third article. The fourth study deals with the influence of sentiment on the tendency to herd behaviour in the context of share price forecasts. It is shown that sentiment does indeed have an influence on the tendency to herd behaviour. In particular, a neutral mood favours a tendency towards herd behaviour. Finally, the fifth study analyses real interest rate forecasts for the Asia-Pacific region in order to assess the forecasting ability of financial analysts. Overall, it can be said that, at least in some countries and for some forecast horizons, forecasts of future interest rate developments in the Asia-Pacific region are more successful than in other parts of the world.

Doctorate in the field of "Rationality and Quality of Economic Forecasts" at the Georg-August-University Göttingen

Doctoral period: 2009 - 2015

Dissertation available here (external link, opens in a new window).

Short description

Economic forecasts are intended to reduce uncertainty about future economic development and support the planning processes of governments and companies. However, empirical studies generally attest to their unsatisfactory quality. In the search for the causes, a central basic requirement for forecasters has emerged in the form of rationality. Obvious and systematic errors, such as regular overestimates, should be recognised and eliminated over time. The first study in the dissertation criticises the prevailing understanding of rationality. This is too far-reaching, which is why forecasters are prematurely denied rationality. Using a new empirical approach, it becomes clear that forecasts can certainly be regarded as rational from a different perspective. The second article shows that there is publicly available information in the form of survey results which, if used appropriately, would contribute to improving the quality of economic forecasts. The rationality of these forecasts is therefore severely limited. The third paper analyses forecast revisions and their causes. Surprisingly, it shows that there is no correlation between rationality and the quality of the forecast time series analysed. The fourth study presents the results of a forecasting game, which aimed to compare the forecasts of amateurs and experts. It turns out that the forecast errors show remarkable similarities.