Contributions in Merlin

Ambient assisted living (AAL) technologies are increasingly presented and sold as essential smart additions to daily life and home environments that will radically transform the healthcare and wellness markets of the future. An ethical approach and a thorough understanding of all ethics in surveillance/monitoring architectures are therefore pressing. AAL poses many ethical challenges raising questions that will affect immediate acceptance and long-term usage. Furthermore, ethical issues emerge from social inequalities and their potential exacerbation by AAL, accentuating the existing access gap between high-income countries (HIC) and low and middle-income countries (LMIC). Legal aspects mainly refer to the adherence to existing legal frameworks and cover issues related to product safety, data protection, cybersecurity, intellectual property, and access to data by public, private, and government bodies. Successful privacy-friendly AAL applications are needed, as the pressure to bring Internet of Things (IoT) devices and ones equipped with artificial intelligence (AI) quickly to market cannot overlook the fact that the environments in which AAL will operate are mostly private (e.g., the home). The social issues focus on the impact of AAL technologies before and after their adoption. Future AAL technologies need to consider all aspects of equality such as gender, race, age and social disadvantages and avoid increasing loneliness and isolation among, e.g. older and frail people. Finally, the current power asymmetries between the target and general populations should not be underestimated nor should the discrepant needs and motivations of the target group and those developing and deploying AAL systems. Whilst AAL technologies provide promising solutions for the health and social care challenges, they are not exempt from ethical, legal and social issues (ELSI). A set of ELSI guidelines is needed to integrate these factors at the research and development stage.

Multiple asset pricing theories predict that large price changes should be associated with abnormal trading volume, inducing investor rebalancing and possibly leading to flights. In contrast, consistent with market microstructure theories, this paper documents freezes, a reduction in trading volume (approximately 30% relative to the previous trading week) during market stress episodes in the European sovereign bond market. We trace the market freezes to increasing transaction costs driven by reduced risk bearing capacity of market makers.

This paper examines abnormal bank equity returns around the announcement and implementations of the largest central bank liquidity operations to date. Those were conducted by the European Central Bank (ECB) at the height of the sovereign debt crisis in 2011 and 2012. I find that banks in countries perceived as being relatively riskier at the time experienced larger positive abnormal equity returns. Relating country-level abnormal returns to country-level liquidity uptake shows that banks with higher liquidity uptake profit disproportionately more from larger returns over this period. This provides evidence that the ECB alleviates stress in the euro area through the provisioning of relatively more liquidity to banks in riskier countries.

We collected detailed personality traits data and preferred investment styles of participants from the German population. We find a significant relationship between personality traits and styles. Then we established a personality-based style profiler based on a least-distance algorithm. We tested its out-of-sample performance. The results of AB testing show that the style profiler provides significantly better fitting style recommendations than a random recommendation. Moreover, including socio-economic characteristics increases the fit. Our results suggest that further research on wealth management could benefit from including the personality of individual investors as a crucial factor, contributing to more satisfying recommendations helping people to invest consistently over time.

We study the role of competition in customers' reactions to litigation against firms, using anonymized mobile phone location data. A class action lawsuit filing results in a 4% average reduction in customer visits to target firms' outlets in the following months. The effect strongly depends on competition. Outlets facing more competition experience significantly larger negative effects. Closer competition matters more, both in terms of geographic and industry proximity. Announcement returns and quarterly accounting revenues around lawsuit filings also strongly depend on competition. Our results suggest that competition is an important component in customers' ability to discipline firms for misbehavior.

We analyze how individual investors respond to inflation. We introduce a unique dataset containing information on local inflation and security portfolios of more than 2,000 clients of a German bank between 1920 and 1924, covering the German hyperinflation. We find that individual investors buy less (sell more) stocks when facing higher local inflation. This effect is more pronounced for less sophisticated investors. Moreover, we document a positive relation between local inflation and forgone returns following stock sales. Our findings are consistent with individual investors suffering from money illusion. Alternative explanations such as consumption needs are unlikely to drive our results.

Answering multi-relation questions over knowledge graphs is a challenging task as it requires multi-step reasoning over a huge number of possible paths. Reasoning-based methods with complex reasoning mechanisms, such as reinforcement learning-based sequential decision making, have been regarded as the default pathway for this task. However, these mechanisms are difficult to implement and train, which hampers their reproducibility and transferability to new domains. In this paper, we propose QAGCN - a simple but effective and novel model that leverages attentional graph convolutional networks that can perform multi-step reasoning during the encoding of knowledge graphs. As a consequence, complex reasoning mechanisms are avoided. In addition, to improve efficiency, we retrieve answers using highly-efficient embedding computations and, for better interpretability, we extract interpretable paths for returned answers. On widely adopted benchmark datasets, the proposed model has been demonstrated competitive against state-of-the-art methods that rely on complex reasoning mechanisms. We also conducted extensive experiments to scrutinize the efficiency and contribution of each component of our model.

This paper presents a portfolio construction approach that combines the hierarchical clustering of a large asset universe with the stock price momentum. On the one hand, investing in high-momentum stocks stabilizes portfolio performance across economic regimes and enhances risk-adjusted returns. On the other hand, hierarchical clustering of a high-dimensional asset universe ensures sparse diversification and mitigates the problems of increased drawdowns and large turnovers typically present in momentum portfolios. Moreover, the proposed portfolio construction approach avoids the covariance matrix inversion. An out-of-sample backtest on a non-survivorship-biased dataset of international stocks shows that hierarchical-momentum portfolios achieve substantially improved cumulative and risk-adjusted portfolio returns as well as decreased portfolio drawdowns compared to the model-free benchmarks net of transaction costs. Furthermore, we demonstrate that the unique characteristics of the hierarchical-momentum portfolios arise due to both dimensionality reduction via clustering and momentum-based stock selection.

We study the impact of economic policy uncertainty on the term structure of nominal interest rates. In a general equilibrium model populated by an uncertainty-averse agent, we show that political uncertainty not only affects the yield curve and the corresponding volatility term structure but also bond risk premia carry a premium for political uncertainty. Our model simultaneously captures both the shape of the yield curve and the hump shape of yield volatilities, a stylized feature that is hard to match with a theoretical model. Our model gives rise to a set of testable predictions for which we find strong support in the data: Higher policy uncertainty leads to a significant decline in yield levels and increases bond yield volatilities. Moreover, policy uncertainty predicts future short rates and has an ambiguous effect on term premia. Finally, short (long) maturity bond risk premia respond negatively (positively) to increases in policy uncertainty.

Employing a quasi-experimental field setting, we examine whether people are outcome biased when self-evaluating their past decisions. Using data from Swiss driving license exams, we find that candidates who narrowly passed the theoretical driving exam are significantly less likely to pass the subsequent practical driving exam – which is taken several months after the theoretical exam – relative to those who failed narrowly. The candidates who passed the theoretical exam in their first attempt received more objections in momentary, on-the-spot kinds of decisions, consistent with the idea that worse preparation is the underlying behavioral difference.