Refine
Year of publication
- 2022 (99) (remove)
Document Type
- Working Paper (99) (remove)
Has Fulltext
- yes (99)
Is part of the Bibliography
- no (99)
Keywords
- ESG (3)
- AI borrower classification (2)
- AI enabled credit scoring (2)
- Artificial Intelligence (2)
- Banking Union (2)
- Big Data (2)
- COVID-19 (2)
- Experiment (2)
- FOMC (2)
- FinTech (2)
Institute
- Wirtschaftswissenschaften (99) (remove)
In more and more situations, artificially intelligent algorithms have to model humans’ (social) preferences on whose behalf they increasingly make decisions. They can learn these preferences through the repeated observation of human behavior in social encounters. In such a context, do individuals adjust the selfishness or prosociality of their behavior when it is common knowledge that their actions produce various externalities through the training of an algorithm? In an online experiment, we let participants’ choices in dictator games train an algorithm. Thereby, they create an externality on future decision making of an intelligent system that affects future participants. We show that individuals who are aware of the consequences of their training on the pay- offs of a future generation behave more prosocially, but only when they bear the risk of being harmed themselves by future algorithmic choices. In that case, the externality of artificially intelligence training induces a significantly higher share of egalitarian decisions in the present.
Since the 2008 financial crisis, European largest banks’ size and business models have largely remained unchallenged. Is that because of banks’ continued structural power over States? This paper challenges the view that States are sheer hostages of banks’ capacity to provide credit to the real economy – which is the conventional definition of structural power. Instead, it sheds light on the geo-economic dimension of banks’ power: key public officials conceive the position of “their own” market-based banks in global financial markets as a crucial dimension of State power. State priority towards banking thus result from political choices over what structurally matters the most for the State. Based on a discourse analysis of parliamentary debates in France, Germany and Spain between 2010 and 2020 as well as on a comparative analysis of the implementation of a special tax on banks in the early 2010s, this paper shows that State’s Finance ministries tend to prioritize geo-economic considerations over credit to firms. By contrast, Parliaments tend to prioritize investment. Power dynamics within the State thus largely shape political priorities towards banking at the domestic and international levels.
Biased auctioneers
(2022)
We construct a neural network algorithm that generates price predictions for art at auction, relying on both visual and non-visual object characteristics. We find that higher automated valuations relative to auction house pre-sale estimates are associated with substantially higher price-to-estimate ratios and lower buy-in rates, pointing to estimates’ informational inefficiency. The relative contribution of machine learning is higher for artists with less dispersed and lower average prices. Furthermore, we show that auctioneers’ prediction errors are persistent both at the artist and at the auction house level, and hence directly predictable themselves using information on past errors.
Lack of privacy due to surveillance of personal data, which is becoming ubiquitous around the world, induces persistent conformity to the norms prevalent under the surveillance regime. We document this channel in a unique laboratory---the widespread surveillance of private citizens in East Germany. Exploiting localized variation in the intensity of surveillance before the fall of the Berlin Wall, we show that, at the present day, individuals who lived in high-surveillance counties are more likely to recall they were spied upon, display more conformist beliefs about society and individual interactions, and are hesitant about institutional and social change. Social conformity is accompanied by conformist economic choices: individuals in high-surveillance counties save more and are less likely to take out credit, consistent with norms of frugality. The lack of differences in risk aversion and binding financial constraints by exposure to surveillance helps to support a beliefs channel.
Colocation services offered by stock exchanges enable market participants to achieve execution costs for large orders that are substantially lower and less sensitive to transacting against high-frequency traders. However, these benefits manifest only for orders executed on the colocated brokers' own behalf, whereas customers' order execution costs are substantially higher. Analyses of individual order executions indicate that customer orders originating from colocated brokers are less actively monitored and achieve inferior execution quality. This suggests that brokers do not make effective use of their technology, possibly due to agency frictions or poor algorithm selection and parameter choice by customers.
We employ a proprietary transaction-level dataset in Germany to examine how capital requirements affect the liquidity of corporate bonds. Using the 2011 European Banking Authority capital exercise that mandated certain banks to increase regulatory capital, we find that affected banks reduce their inventory holdings, pre-arrange more trades, and have smaller average trade size. While non-bank affiliated dealers increase their market-making activity, they are unable to bridge this gap - aggregate liquidity declines. Our results are stronger for banks with a higher capital shortfall, for non-investment grade bonds, and for bonds where the affected banks were the dominant market-maker.
We develop a two-sector incomplete markets integrated assessment model to analyze the effectiveness of green quantitative easing (QE) in complementing fiscal policies for climate change mitigation. We model green QE through an outstanding stock of private assets held by a monetary authority and its portfolio allocation between a clean and a dirty sector of production. Green QE leads to a partial crowding out of private capital in the green sector and to a modest reduction of the global temperature by 0.04 degrees of Celsius until 2100. A moderate global carbon tax of 50 USD per tonne of carbon is 4 times more effective.
The great financial crisis and the euro area crisis led to a substantial reform of financial safety nets across Europe and – critically – to the introduction of supranational elements. Specifically, a supranational supervisor was established for the euro area, with discrete arrangements for supervisory competences and tasks depending on the systemic relevance of supervised credit institutions. A resolution mechanism was created to allow the frictionless resolution of large financial institutions. This resolution mechanism has been now complemented with a funding instrument.
While much more progress has been achieved than most observers could imagine 12 years ago, the banking union remains unfinished with important gaps and deficiencies. The experience over the past years, especially in the area of crisis management and resolution, has provided impetus for reform discussions, as reflected most lately in the Eurogroup statement of 16 June 2022.
This Policy Insight looks primarily at the current and the desired state of the banking union project. The key underlying question, and the focus here, is the level of ambition and how it is matched with effective legal and regulatory tools. Specifically, two questions will structure the discussions:
What would be a reasonable definition and rationale for a ‘complete’ banking union? And what legal reforms would be required to achieve it?
Banking union is a case of a new remit of EU-level policy that so far has been established on the basis of long pre-existing treaty stipulations, namely, Article 127(6) TFEU (for banking supervision) and Article 114 TFEU (for crisis management and deposit insurance). Could its completion be similarly carried out through secondary law? Or would a more comprehensive overhaul of the legal architecture be required to ensure legal certainty and legitimacy?
Search costs for lenders when evaluating potential borrowers are driven by the quality of the underwriting model and by access to data. Both have undergone radical change over the last years, due to the advent of big data and machine learning. For some, this holds the promise of inclusion and better access to finance. Invisible prime applicants perform better under AI than under traditional metrics. Broader data and more refined models help to detect them without triggering prohibitive costs. However, not all applicants profit to the same extent. Historic training data shape algorithms, biases distort results, and data as well as model quality are not always assured. Against this background, an intense debate over algorithmic discrimination has developed. This paper takes a first step towards developing principles of fair lending in the age of AI. It submits that there are fundamental difficulties in fitting algorithmic discrimination into the traditional regime of anti-discrimination laws. Received doctrine with its focus on causation is in many cases ill-equipped to deal with algorithmic decision-making under both, disparate treatment, and disparate impact doctrine. The paper concludes with a suggestion to reorient the discussion and with the attempt to outline contours of fair lending law in the age of AI.
Search costs for lenders when evaluating potential borrowers are driven by the quality of the underwriting model and by access to data. Both have undergone radical change over the last years, due to the advent of big data and machine learning. For some, this holds the promise of inclusion and better access to finance. Invisible prime applicants perform better under AI than under traditional metrics. Broader data and more refined models help to detect them without triggering prohibitive costs. However, not all applicants profit to the same extent. Historic training data shape algorithms, biases distort results, and data as well as model quality are not always assured. Against this background, an intense debate over algorithmic discrimination has developed. This paper takes a first step towards developing principles of fair lending in the age of AI. It submits that there are fundamental difficulties in fitting algorithmic discrimination into the traditional regime of anti-discrimination laws. Received doctrine with its focus on causation is in many cases ill-equipped to deal with algorithmic decision-making under both, disparate treatment, and disparate impact doctrine. The paper concludes with a suggestion to reorient the discussion and with the attempt to outline contours of fair lending law in the age of AI.