Navigation and service

Deutsche Bundesbank (Link to homepage)

Frankfurt/Main | 27.03.2018

Monthly Report: Doubts about the effectiveness of CoCo bonds

In the aftermath of the 2007-08 financial crisis, one of the core aims of regulatory and supervisory authorities was to strengthen banks’ capital base so as to render them more crisis-proof. One of the instruments they discussed was contingent convertible bonds (CoCo bonds). CoCo bonds are subordinated bonds that pay a coupon, and which are either converted into common equity tier 1 (CET1) capital or written down when certain contractually specified events occur.

CoCos are a hybrid form of financing, combining the typical advantages of debt financing, such as fixed remuneration, with the loss absorbing capacity of equity. This means that this form of financing is designed to absorb losses when needed, allowing the bank to continue as a going concern. The Basel Committee on Banking Supervision therefore recognises CoCo bonds as regulatory capital under certain conditions. Moreover, coupon payments have been tax deductible in Germany since 2014. Data from financial news services show that between 2009 and the end of 2017 at least 285 CoCo bonds with a total volume of €193 billion were issued in the European Union (EU).

Doubts about loss absorbing capacity and high complexity

In the current Monthly Report, Bundesbank experts analyse the properties of these securities. They are sceptical about whether CoCo bonds do, in fact, possess the advantages ascribed to them: "The actual design of CoCo bonds gives rise to doubts about their effectiveness for banks as a loss absorbing instrument on a going-concern basis." One of the design elements they examine is the contractually defined trigger for conversion into equity or write-down. In practice, the CoCo bonds issued in the EU thus far have mainly been designed such that conversion is pegged to a CET1 threshold (the CET1 capital ratio). For over 40% of all bonds issued in the EU, this ratio is 5.125%, the statutory minimum for eligibility as additional tier 1 capital. Conversion of the bond is therefore triggered precisely when the issuing institution’s CET1 capital ratio as a percentage of its risk exposure falls below this level. The Bundesbank experts question whether a threshold of 5.125% is sufficient for CoCo bonds to actually perform their function. In the Monthly Report, they advocate examining "how much the mechanical CET1 threshold has to be raised in order to ensure that CoCo bonds have the effect regulators intend – to act as a loss absorbing instrument for banks on a going-concern basis".

The authors also establish that the complexity of CoCo bonds means they are not suited to all investors and refer to a warning to this effect issued for retail investors by the Federal Financial Supervisory Authority (BaFin). Banks and other institutional investors, such as insurers and investment funds, are the more feasible investor group for CoCo bonds, they argue; but even for institutional investors, CoCo pricing seems to present difficulties.

Perverse incentives to take on greater risk

The authors take a critical view of a further detail of CoCo design. In the case of bonds that convert into CET1 capital, both control rights and rights of participation in profits and losses are shifted, to varying degrees, from original equity holders to bondholders ("dilution"). However, they note that empirical studies have concluded that, in practice, CoCos almost exclusively entail weak dilution of the original equity holders’ stakes. The current design could thus tend to result in perverse incentives leading banks to take on greater risk. "From a regulatory perspective, therefore, the aim should be for substantial dilution, thus setting incentives for the original equity holders to ensure sustainable risk provisioning," the authors state.

In addition, they point to other cross-bank risks posed by CoCo bonds. The first crucial aspect is the interconnectedness of banks as well as of credit institutions and institutional investors, which could increase further as a result of CoCo bonds. Second, they argue that "information-based contagion effects" are a possibility: a trigger event occurring at one bank, for example, could produce a negative signalling effect on other banks, and beyond.

Overall, the authors are doubtful about what advantages such bonds have over CET1 capital. For instance, it is unclear whether their supposed cost advantage over CET1 capital actually exists, they write, if the high complexity and associated risks are factored in. From a regulatory perspective, CoCo bonds are primarily useful because of their capacity for early loss absorption, but in their current design this is unlikely to be assured, according to the Monthly Report. Even if they were better designed, the problem of their complexity and potential side effects would remain, which is why the authors determine that a "regulatory incentive" is not appropriate at present: "Instead, focusing on CET1 capital is likely to be the more expedient approach in the long term to safeguarding and improving the stability of banks".


Additional information