(This interview appeared in SAFE Newsletter Q4 2016)
Martin Götz is Professor for Regulation and Stability of Financial Institutions at the Research Center SAFE at Goethe University Frankfurt. His research interests lie in financial economics, banking and applied microeconomics. From 2010 to 2013, Götz worked as a Financial Economist in the Risk and Policy Analysis Unit at the Federal Reserve Bank of Boston. He earned a Ph.D. in Economics from Brown University, Providence/Rhode Island, in May 2010, and an M.Sc. in Financial and Industrial Economics from Royal Holloway, University of London.
SAFE has recently launched the “Bail-In Tracker”, a project of SAFE Law Professor Tobias Tröger, Stephan Lorz (Börsen-Zeitung) and you. What is the project about?
The idea of the Bail-In Tracker is to provide information about the applicability of the new European bail-in regulation laid out in the Bank Recovery and Resolution Directive (BRRD). The BRRD constitutes a legal framework that allows resolution authorities to force banks’ creditors to share the burden in case of financial distress by converting parts of their debt into equity (bail-in). A waterfall principle specifies the order in which specific liabilities of banks are subject to a bail-in. A bank’s subordinated debt is the first debt instrument to be bailed in if the bail-in of the bank’s Tier 2 capital is not sufficient. To investigate how effective the new bail-in regulation might be, we collected publicly available information on the amount of outstanding subordinated debt of 36 large European banks headquartered in 15 European countries. Based on these data, the Bail-In Tracker provides regularly updated information on the magnitude of a potential bail-in at one of these banks with respect to one of the most critical balance-sheet positions. Furthermore, we make the data and methodology available on our website.
What are your findings after reviewing the data?
We find that the aggregated amount of outstanding subordinated debt at large European banks has increased overall in our sample period which goes back to July 2011. On the one hand, this indicates that there is a larger portion of these financial securities in the market that could be subject to a bail-in. On the other hand, banks in our sample have also grown in recent years so that the share of subordinated debt in the percentage of total liabilities has remained fairly stable over the sample period, at a little less than 3 percent. In a separate study (Götz and Tröger, 2016), we look at individual institutions and find that banks that hold less equity capital tend to have a higher level of subordinated debt. This is interesting because banks with smaller equity ratios have less of an equity buffer and hence the outstanding subordinated debt of these banks may be more likely to be subject to a bail-in.
Have banks adjusted their balance sheets since the adoption of BRRD?
We are very interested in getting answers to this question. One could imagine that, for example, banks have moved out of the segment of subordinated debt as it has now become riskier for investors. These might prefer to invest in banks by using other instruments that are less likely to be bailed in when things get tough. Unfortunately, as there is a variety of possibilities for banks to adjust their balance sheets as a reaction to the new regulation, more granular data is needed to look into this particular question in more detail. These data are collected by resolution authorities, but they are not available to the public. So I hope that we will see some studies from these institutions in the future that examine this question.
Are there differences across countries with respect to the level of subordinated bank debt?
We found some striking differences across banks headquartered in different European countries (see Figure 1). For example, banks in the United Kingdom tend to rely more on subordinated debt in their financing than banks from other countries: About 6 percent in 2011 and now more than 7.5 percent of their liabilities is publicly traded subordinated debt. This is due to a different culture of bank financing in the UK where subordinated debt has always been a more common form of bank financing. At the other end of the spectrum are financial institutions in Greece, Denmark and Spain. Banks in these countries have on average a very small share of their liabilities in subordinated debt – only about 1 percent. In all three countries the share has also considerably decreased during the last five years. This is not surprising when you consider Greek banks and all the trouble they have gone through in recent years. Their capital structure has changed substantially and it will take more time for them to go back to capital markets and issue subordinated debt as a means of financing.
You also looked into the currency of subordinated bonds.
Yes, it was interesting for us to see that banks headquartered in the eurozone issue about 15 percent of their subordinated debt in a currency other than the euro. Most of these bonds are issued in US dollars, the second most common currency are British pounds. Over the sample period, the amount of subordinated debt issued in a currency other than the euro slightly increased (see Figure 2). Looking into this subject in detail, it seems that the larger the bank, the larger is also the share of subordinated debt in a currency different than the euro. The increase in bonds issued in another currency is most probably due to better market conditions and the opportunity to address different investors. We are interested in this pattern as it may influence the applicability of BRRD – particularly as issuances in foreign currencies involve a currency risk. This has to be kept in mind when evaluating the total amount of subordinated debt and the question of whether this would be sufficient in times of distress or not.
A further finding is about the company level that issued the debt instrument…
We analyzed whether banks issued subordinated debt at the parent level or at a subsidiary or affiliate level. On average, about a quarter of outstanding subordinated debt is issued at a subsidiary level with a slight decrease over the sample period (see Figure 3). With respect to BRRD, this issuance behavior adds a further level of complexity. Debt issued by a subsidiary might, for example, be subject to a different jurisdiction, different market reactions and so on.
With respect to all our results, however, more research is needed to further examine the motivation that drives banks to issue their debt in this way or the other. We plan to look into all these questions in more detail and we hope that our data collection encourages other researchers to conduct their own studies on the topic.
The Bail-In Tracker (www.bail-in-tracker.eu) is an interdisciplinary project by the Research Center SAFE in cooperation with Boersen-Zeitung, funded by the VolkswagenStiftung in the context of the project “Science and Data Journalism”.
Götz, M. and T. Tröger (2016)
"Should the Marketing of Subordinated Debt be Restricted/Different in one way or the other? What to do in the Case of Mis-Selling?",
SAFE Policy White Paper No. 35.