Maximilian Weber, Visualization of Deep Features with Grad-CAM and LOTS, University of Zurich, Faculty of Business, Economics and Informatics, 2023. (Master's Thesis)
Deep learning models, particularly Convolutional Neural Networks (CNNs), have achieved remarkable success in image classification tasks. However, their lack of interpretability raises concerns about their trustworthiness, especially in high-risk domains like healthcare. To improve transparency, Explainable Artificial Intelligence (XAI) techniques have been developed.
This thesis has a primary focus on expanding the Layerwise Origin Target Synthesis (LOTS) method, which is originally designed as a technique for generating adversarial images, to incorporate visualization capabilities. The aim is to address the limitations observed in current CAM-based visualization techniques that only offer broad area visualizations. The research explores methods for evaluating and comparing visualization techniques in the absence of a standard evaluation metric framework. Additionally, it investigates the applicability of the extended LOTS visualization technique to classes not present in the training dataset.
Based on our findings, the LOTS visualization algorithm we propose, generates more focused visualizations that do not require explicit class specification, thereby also serving as a valuable tool for evaluating image quality within a training set. Furthermore, by adjusting the size of the Gaussian blur filter, it is possible to highlight fine locations in an image. Moreover, we demonstrate the potential for extending the LOTS algorithm to classes not included in the training dataset, although further research is required for validation. Lastly, we emphasize the importance of a standardized evaluation metrics framework. |
|
Jin Zhang, Enhancing COMFORT with Fractional Difference: An Empirical Study, University of Zurich, Faculty of Business, Economics and Informatics, 2023. (Master's Thesis)
We study the effect of using fractional returns instead of log returns on improving
portfolio optimization. The fractional return is defined as a fractional difference
of log prices. The rationale behind the substitution lies in the fractional
return’s similarity to log returns and its memory richness nature. We conduct
extensive experiments on 8 groups of stocks with Gaussian and COMFORT
models, with the optimization objective of maximizing the Sharpe ratio, with
or without short-selling. Our findings are the following. First, qualitatively,
portfolios using fractional returns behave generally more turbulent. Second,
the Sharpe ratio demonstrates a pattern of first increasing and then decreasing
with respect to degrees of fractional returns. Third, application of fractional
returns yields improved Sharpe ratios in most groups. Fourth, the improvement
effect is greater for portfolios with short selling. We also give an interpretation
of the findings and conclude with future directions. |
|
Noé Matumona, Stock Market Prediction: When Freely Available Data Meets Machine Learning, University of Zurich, Faculty of Business, Economics and Informatics, 2023. (Bachelor's Thesis)
This thesis aims to bring finance and machine learning together and examines whether a
model can be created that is used to evaluate if the US stock market can be outperformed
using solely freely available data excluding paid financial services of data providers entirely.
Said examination is done for two time dimensions yearly and quarterly with a
focus on both the complete market and a sector-based approach. Different machine learning
techniques such as logistic regression, random forest, gradient boosting, and support
vector machine are being used and applied for the task at hand. Although the results are
less than ideal, the model performance may improve over time when more data is added.
However, the sector-based results are more promising due to some sectors scoring higher
in the examined model performance metrics.
II |
|
Shana Meier, Nachhaltigkeitskriterien im Prozess der Unternehmenskreditvergabe durch Banken. Die Rolle der asymmetrischen Informationsverteilung., University of Zurich, Faculty of Business, Economics and Informatics, 2023. (Bachelor's Thesis)
Nachhaltige Firmenkredite sind ein aufkommendes Thema im Schweizer Kreditmarkt.
Zusätzlich zur Kreditfähigkeit und -würdigkeit beinhaltet die Bonitätsprüfung
eine Nachhaltigkeitskomponente, wodurch die Informationengrundmenge steigt. Somit
ist die Forschungsfrage naheliegend, ob sich die Informationsasymmetrien zwischen
Kreditgeber und -nehmer vergrössern, sobald ein Kredit als nachhaltig deklariert
wird. Zur Beantwortung der Forschungsfrage wird eine qualitative Forschungsmethode
angewendet, wobei Experten aus vier verschiedenen Schweizer Banken interviewt
werden. Der Fokus liegt auf den Nachhaltigkeitskriterien und deren Einfluss
auf die Informationsasymmetrien. Die Ergebnisse zeigen, dass die Informationsasymmetrien
bei nachhaltigen Krediten tendenziell schwächer wirken, da das zu finanzierende
Unternehmen genauer analysiert wird und somit eine engere Kundenbeziehung
entsteht. |
|
John Ott, Auswirkungen von ESG-Exclusions auf die Performance eines Portfolios, University of Zurich, Faculty of Business, Economics and Informatics, 2023. (Master's Thesis)
Diese Arbeit untersucht die Auswirkungen von ESG-Exclusions auf die risikoadjustierte Performance
eines Portfolios. Dazu werden 22 Exclusions sowohl einzeln als auch in thematischen Gruppen
auf das Portfolio der Schweizer Krankenversicherung CSS angewendet. Die daraus entstehenden
Exclusion-Portfolios werden anhand der Rendite, Sharpe-Ratio, Tracking-Error, Information-Ratio
und in verschiedenen Faktor-Modellen ¨uber den Zeitraum von 2010 bis 2022 verglichen. Die Resultate
zeigen, dass die verschiedenen Exclusions unterschiedliche Auswirkungen auf die Portfolios haben.
Die Mehrzahl der Exclusions f¨uhrt zu einer besseren absoluten und risikoadjustierten Rendite bzw.
zu besseren Performancemassen. Jedoch sind diese Resultate abh¨angig vom gew¨ahlten Zeitraum, wie
rollierende Berechnungen zeigen. |
|
Fedor Doval, COMPARISON BETWEEN A DETERMINISTIC AND STOCHASTIC APPROACH IN MODELLING THE BEHAVIOURAL MATURITY OF NON-MATURING DEPOSITS (NMD) , University of Zurich, Faculty of Business, Economics and Informatics, 2023. (Master's Thesis)
|
|
Mo Yi, Gender differences in non-cognitive skills and outcomes in education and labor markets of Swiss students, University of Zurich, Faculty of Business, Economics and Informatics, 2023. (Master's Thesis)
|
|
Vibirthan Vijayarajah, Entstehung und Auswirkungen von Zombie-Unternehmen Eine umfassende Analyse der Eigenschaften, Risiken und Bekämpfungsmassnahmen von Zombie-Firmen, University of Zurich, Faculty of Business, Economics and Informatics, 2023. (Bachelor's Thesis)
Das Ziel dieser Arbeit ist es, dass die Leser*innen ein tieferes Verständnis bezüglich Zombie-Unternehmen erhalten und sich deren Gefahren in der Wirtschaft bewusst werden. Im
Rahmen dessen wird anfangs erklärt, welche Kennzahlen ein Zombie-Unternehmen charakterisieren, wobei im Allgemeinen gesagt werden kann, dass diese Unternehmen eine hohe
Verschuldung, sowie eine geringe Rentabilität aufweisen. Dabei begünstigen finanziell
schwache Banken, ein ineffizientes Insolvenzregime, sowie ein niedriges Zinsniveau und
Finanzkrisen die Entstehung und Aufrechthaltung von nicht lebensfähigen Unternehmen. In
den meisten Fällen betrifft dies kleine und mittlere Unternehmen mit einem Umsatz von
weniger als 500 Millionen US-Dollar. Im Vergleich zu gesunden Unternehmen weisen die
Zombie-Firmen eine geringere Investitionstätigkeit wie eine halb so hohe Arbeits- und Gesamtproduktivität auf. Zugleich binden sie begrenzte Ressourcen, wie Human- und Finanzkapital in der Branche, was zur Folge hat, dass gesunde Unternehmen ihr Potenzial nicht
ausschöpfen können. Durch die Verzerrung des Marktes gibt es ein Überangebot, daraus
resultiert eine geringere Gewinnmarge und höhere Barrieren für Markteinsteiger. Die Immobilienbranche weist mit rund acht Prozent den höchsten Anteil an Zombie-Unternehmen
in der Wirtschaft auf. Viele Wissenschaftler*innen sehen mit der Verbesserung der Insolvenzreglungen und der Stabilisierung der schwachen Banken essenzielle Massnahmen, um
die Entstehung beziehungsweise Aufrechthaltung der Zombie-Unternehmen zu hemmen. |
|
Winifred Huang, Philip Molyneux, Steven Ongena, Ru Xie, The new challenges of global banking and finance, The European journal of finance, Vol. 29 (7), 2023. (Journal Article)
The economic downturn caused by the Covid-19 pandemic has brought unprecedented uncertainty to the global banking system. Banks are facing critical market challenges driven by uncertain monetary policies, deterioration in credit quality, and regulation and compliance pressures. These challenges highlight the importance of better understanding the new role of financial intermediations in facilitating efficient capital allocations and economic development. This article reviews the related literature on monetary policy uncertainty, bank performance, digital finance, and introduces articles on these themes. Finally, we propose potential areas for future research. |
|
Steven Ongena, Manthos D Delis, Evangelos V Dioikitopoulos, Population diversity and financial risk-taking, Journal of Banking and Finance, Vol. 151, 2023. (Journal Article)
We hypothesize that financial risk-taking originates in preindustrial interpersonal population diversity. We use data on immigrants residing in the United States and show that controlling for all known determinants of portfolio decisions and more than 100 control variables, diversity in the country of immigrants’ origin positively affects stock market participation and the level of risky asset holdings. Our results remain robust when instrumenting diversity with plant variety. We also identify the channels through which the effect of diversity operates (mostly individualism and human capital), but also conclude that diversity exerts an independent effect. |
|
Tobias Schultheiss, Curdin Pfister, Ann-Sophie Gnehm, Uschi Backes-Gellner, Education expansion and high-skill job opportunities for workers: Does a rising tide lift all boats?, Labour Economics, Vol. 82, 2023. (Journal Article)
We examine how education expansions affect the job opportunities for workers with and without the new education. To identify causal effects, we exploit a quasi-random establishment of Universities of Applied Sciences (UASs), bachelor-granting three-year colleges that teach and conduct applied research. By applying machine-learning methods to job advertisement data, we analyze job content before and after the education expansion. We find that, in regions with the newly established UASs, not only job descriptions of the new UAS graduates but also job descriptions of workers without this degree (i.e., middle-skilled workers with vocational training) contain more high-skill job content. This upskilling in job content is driven by an increase in high-skill R&D-related tasks and linked to employment and wage gains. The task spillovers likely occur because UAS graduates with applied research skills build a bridge between middle-skilled workers and traditional university graduates, facilitating the integration of the former into R&D-related tasks. |
|
Alberto Huertas Celdran, Pedro Miguel Sánchez Sánchez, Miguel Azorín, Gérôme Bovet, Gregorio Martínez Pérez, Burkhard Stiller, Intelligent and behavioral-based detection of malware in IoT spectrum sensors, International Journal of Information Security, Vol. 22 (3), 2023. (Journal Article)
The number of Cyber-Physical Systems (CPS) available in industrial environments is growing mainly due to the evolution of the Internet-of-Things (IoT) paradigm. In such a context, radio frequency spectrum sensing in industrial scenarios is one of the most interesting applications of CPS due to the scarcity of the spectrum. Despite the benefits of operational platforms, IoT spectrum sensors are vulnerable to heterogeneous malware. The usage of behavioral fingerprinting and machine learning has shown merit in detecting cyberattacks. Still, there exist challenges in terms of (i) designing, deploying, and evaluating ML-based fingerprinting solutions able to detect malware attacks affecting real IoT spectrum sensors, (ii) analyzing the suitability of kernel events to create stable and precise fingerprints of spectrum sensors, and (iii) detecting recent malware samples affecting real IoT spectrum sensors of crowdsensing platforms. Thus, this work presents a detection framework that applies device behavioral fingerprinting and machine learning to detect anomalies and classify different botnets, rootkits, backdoors, ransomware and cryptojackers affecting real IoT spectrum sensors. Kernel events from CPU, memory, network,file system, scheduler, drivers, and random number generation have been analyzed, selected, and monitored to create device behavioral fingerprints. During testing, an IoT spectrum sensor of the ElectroSense platform has been infected with ten recent malware samples (two botnets, three rootkits, three backdoors, one ransomware, and one cryptojacker) to measure the detection performance of the framework in two different network configurations. Both supervised and semi-supervised approaches provided promising results when detecting and classifying malicious behaviors from the eight previous malware and seven normal behaviors. In particular, the framework obtained 0.88–0.90 true positive rate when detecting the previous malicious behaviors as unseen or zero-day attacks and 0.94–0.96 F1-score when classifying them |
|
Tommy Crépellière, Matthias Pelster, Stefan Zeisberger, Arbitrage in the market for cryptocurrencies, Journal of financial markets, Vol. 64, 2023. (Journal Article)
Arbitrage opportunities in markets for cryptocurrencies are well-documented. In this paper, we confirm that they exist; however, their magnitude decreased greatly from April 2018 onward. Analyzing various trading strategies, we show that it is barely possible to exploit existing price differences since then. We discuss and test several mechanisms that may be responsible for the increased market efficiency and find that informed trading is correlated with a reduction in arbitrage opportunities. |
|
Erdinc Akyildirim, Shaen Corbet, Murat Tiniç, Ahmet Sensoy, Adverse selection in cryptocurrency markets, The journal of financial research, Vol. 46 (2), 2023. (Journal Article)
In this article we investigate the influence that information asymmetry may have on future volatility, liquidity, market toxicity, and returns within cryptocurrency markets. We use the adverse-selection component of the effective spread as a proxy for overall information asymmetry. Using order and trade data from the Bitfinex exchange, we first document statistically significant adverse-selection costs for major cryptocurrencies. Also, our results suggest that adverse-selection costs, on average, correspond to 10% of the estimated effective spread, indicating an economically significant impact of adverse-selection risk on transaction costs in cryptocurrency markets. Finally, we document that adverse-selection costs are important predictors of intraday volatility, liquidity, market toxicity, and returns. |
|
Tilman Santarius, Jan Bieser, Vivian Frick, Mattias Höjer, Maike Gossen, Lorenz Hilty, Eva Kern, Johanna Pohl, Friederike Rohde, Steffen Lange, Digital sufficiency: conceptual considerations for ICTs on a finite planet, Annales des Telecommunications, Vol. 78 (5-6), 2023. (Journal Article)
ICT hold significant potential to increase resource and energy efficiencies and contribute to a circular economy. Yet unresolved is whether the aggregated net effect of ICT overall mitigates or aggravates environmental burdens. While the savings potentials have been explored, drivers that prevent these and possible counter measures have not been researched thoroughly. The concept digital sufficiency constitutes a basis to understand how ICT can become part of the essential environmental transformation. Digital sufficiency consists of four dimensions, each suggesting a set of strategies and policy proposals: (a) hardware sufficiency, which aims for fewer devices needing to be produced and their absolute energy demand being kept to the lowest level possible to perform the desired tasks; (b) software sufficiency, which covers ensuring that data traffic and hardware utilization during application are kept as low as possible; (c) user sufficiency, which strives for users applying digital devices frugally and using ICT in a way that promotes sustainable lifestyles; and (d) economic sufficiency, which aspires to digitalization supporting a transition to an economy characterized not by economic growth as the primary goal but by sufficient production and consumption within planetary boundaries. The policies for hardware and software sufficiency are relatively easily conceivable and executable. Policies for user and economic sufficiency are politically more difficult to implement and relate strongly to policies for environmental transformation in general. This article argues for comprehensive policies for digital sufficiency, which are indispensible if ICT are to play a beneficial role in overall environmental transformation. |
|
Christoph Grimpe, Wolfgang Sofka, Ulrich Kaiser, Competing for digital human capital: The retention effect of digital expertise in MNC subsidiaries, Journal of International Business Studies, Vol. 54 (4), 2023. (Journal Article)
Employees with relevant knowledge and skills for digitalization have become increasingly important for the competitiveness of MNCs. However, the shortage of such digital human capital in many host countries is putting pressure on MNC subsidiaries to prevent these employees from leaving. We theorize that the retention of digital human capital in MNC subsidiaries does not merely depend on salaries but crucially on the learning opportunities that subsidiaries offer. By integrating mechanisms from the literature on subsidiary-specific advantages into theoretical models explaining voluntary mobility constraints of employees, we reason that the opportunities for acquiring new skills in subsidiaries with advanced digital expertise will reduce the odds of losing these valuable employees. We test our theoretical predictions for 11,598 employees with digital human capital working for 866 foreign MNC subsidiaries in Denmark observed between 2002 and 2012. We find that digital expertise helps retaining digital human capital. The effect is stronger if subsidiaries have an internationally diverse workforce and when they possess patented technologies. Both factors provide distinct learning opportunities from digital expertise. The effect is weaker if the subsidiary is located in regional clusters of digital expertise since alternative employers may offer similar learning opportunities. |
|
Giampiero Marra, Matteo Fasiolo, Rosalba Radice, Rainer Winkelmann, A flexible copula regression model with Bernoulli and Tweedie margins for estimating the effect of spending on mental health, Health Economics, Vol. 32 (6), 2023. (Journal Article)
We develop a flexible two-equation copula model to address endogeneity of medical expenditures in a distribution regression for health. The expenditure margin uses the compound gamma distribution, a special case of the Tweedie family of distributions, to account for a spike at zero and a highly skewed continuous part. An efficient estimation algorithm offers flexible choices of copulae and link functions, including logit, probit and cloglog for the health margin. Our empirical application revisits data from the Rand Health Insurance Experiment. In the joint model, using random insurance plan assignment as instrument for spending, a $1000 increase is estimated to reduce the probability of a low post-program mental health index by 1.9 percentage points. The effect is not statistically significant. Ignoring endogeneity leads to a spurious positive effect estimate. |
|
Douglas G Lee, Todd Anthony Hare, Value certainty and choice confidence are multidimensional constructs that guide decision-making, Cognitive, Affective, & Behavioral Neuroscience, Vol. 23 (3), 2023. (Journal Article)
The degree of certainty that decision-makers have about their evaluations of available choice alternatives and their confidence about selecting the subjectively best alternative are important factors that affect current and future value-based choices. Assessments of the alternatives in a given choice set are rarely unidimensional; their values are usually derived from a combination of multiple distinct attributes. For example, the taste, texture, quantity, and nutritional content of a snack food may all be considered when determining whether to consume it. We examined how certainty about the levels of individual attributes of an option relates to certainty about the overall value of that option as a whole and/or to confidence in having chosen the subjectively best available option. We found that certainty and confidence are derived from unequally weighted combinations of attribute certainties rather than simple, equal combinations of all sources of uncertainty. Attributes that matter more in determining choice outcomes also are weighted more in metacognitive evaluations of certainty or confidence. Moreover, we found that the process of deciding between two alternatives leads to refinements in both attribute estimations and the degree of certainty in those estimates. Attributes that are more important in determining choice outcomes are refined more during the decision process in terms of both estimates and certainty. Although certainty and confidence are typically treated as unidimensional, our results indicate that they, like value estimates, are subjective, multidimensional constructs. |
|
Zacharias Sautner, Laurence Van Lent, Grigory Vilkov, Ruishen Zhang, Firm‐level climate change exposure, Journal of Finance, Vol. 78 (3), 2023. (Journal Article)
We develop a method that identifies the attention paid by earnings call participants to firms' climate change exposures. The method adapts a machine learning keyword discovery algorithm and captures exposures related to opportunity, physical, and regulatory shocks associated with climate change. The measures are available for more than 10,000 firms from 34 countries between 2002 and 2020. We show that the measures are useful in predicting important real outcomes related to the net-zero transition, in particular, job creation in disruptive green technologies and green patenting, and that they contain information that is priced in options and equity markets. |
|
Mascia Bedendo, Emilia Garcia, Linus Siming, Managers' cultural origin and corporate response to an economic shock, Journal of Corporate Finance, Vol. 80, 2023. (Journal Article)
We exploit the exogenous Covid-19 shock in a bicultural area of Italy to identify cultural differences in the way companies respond to economic shocks. Firms with managers of diverse cultural backgrounds resort to different forms of government aid, diverge in their investment decisions, and have different growth rates. These findings are consistent with cultural differences in time preferences and debt aversion. Specifically, we find that the response of managers belonging to a more long-term oriented culture is characterized by a lower recourse to debt, more investments and higher growth rates. Overall, our results show that the cultural origin of managers significantly affects firms' reaction to economic shocks and real economic outcomes. |
|