Medienanfragen richten Sie bitte an:
Telefon: +49 345 7753-720
Email: presse@iwh-halle.de
Team Kommunikation
So viel kostet FremdenfeindlichkeitReint GroppSüddeutsche Zeitung, 1. September 2026
In the wake of the recent financial crisis, significant regulatory actions have been taken aimed at limiting risks emanating from trading in bank business models. Prominent reform proposals are the Volcker Rule in the U.S., the Vickers Report in the UK, and, based on the Liikanen proposal, the Barnier proposal in the EU. A major element of these reforms is to separate “classical” commercial banking activities from securities trading activities, notably from proprietary trading. While the reforms are at different stages of implementation, there is a strong ongoing discussion on what possible economic consequences are to be expected. The goal of this paper is to look at the alternative approaches of these reform proposals and to assess their likely consequences for bank business models, risk-taking and financial stability. Our conclusions can be summarized as follows: First, the focus on a prohibition of only proprietary trading, as envisaged in the current EU proposal, is inadequate. It does not necessarily reduce risk-taking and it likely crowds out desired trading activities, thereby negatively affecting financial stability. Second, there is potentially a better solution to limit excessive trading risk at banks in terms of potential welfare consequences: Trading separation into legally distinct or ring-fenced entities within the existing banking organizations. This kind of separation limits cross-subsidies between banking and proprietary trading and diminishes contagion risk, while still allowing for synergies across banking, non-proprietary trading and proprietary trading.
Motivated by the variety of bank risk proxies, our analysis reveals that nonperforming assets are a well-suited complement to the Z-score in studies of bank risk.
We investigate four proxies for bank risk that are frequently used in the literature. Our analysis shows that non-performing assets are a good proxy for bank risk for two reasons. First, non-performing assets nest the alternative proxies as shown by the high share of variation in non-performing assets explained by the Z-score, loan loss reserves and loan loss provisions. Second, non-performing assets are well-suited to explain bank failures one year ahead. The latter point also holds for the Z-score whereby the information content of the Z-score seems to differ from the other variables. We conclude that non-performing assets are a well-suited complement to the Z-score, which may come with calculation issues regarding the volatility of profitability, in studies of bank risk.
In our analysis, we consider the distribution of decision power over financing and investment between MNEs’ headquarters and foreign subsidiaries and its influence on the foreign affiliates’ financial restrictions. Our research results show that headquarters of multinational enterprises have not (yet) moved much decision power to their foreign subsidiaries at all. We use data from the IWH FDI Micro Database which contains information on corporate governance structures and financial restrictions of 609 enterprises with a foreign investor in Hungary, Poland, the Czech Republic, Slovakia, Romania and East Germany. We match data from Bureau van Dijk’s AMADEUS database on financial characteristics. We find that a high concentration of decision power within the MNE’s headquarter implicates high financial restrictions within the subsidiary. Square term results show, however, that the effect of financial constraints within the subsidiary decreases and finally turns insignificant when decision power moves from headquarter to subsidiary. Thus, economic policy should encourage foreign investors in the case of foreign acquisition of local enterprises to leave decision power within the enterprise and in the case of Greenfield investment to provide the newly established subsidiaries with as much power over corporate governance structures as possible.
