SAFE Policy Blog

Towards a Three-Part Structure of Banking Capital

The regulatory setting for bail-in does not ensure private liability and market stability at the same time

By Martin Götz, Jan Pieter Krahnen and Tobias Tröger

The current EU regulation is often thought to imply that a bail-in can and should be limitless as it, in principle, affects all the liabilities of a bank. The notion of a comprehensive private liability of debtholders corresponds to the desire that a bank should never again be bailed out by the taxpayer. At first, this desire seems plausible and consistent with regulatory objectives. After all, a comprehensive private liability also applies in the case of insolvencies in the manufacturing or service sectors.

The bright and dark side of the run risk

And yet, this demand misses its target as it causes additional problems. The error in reasoning lies in the disregard of a fundamental difference between (non-financial) companies and banks: Banks are exposed to the risk of a bank run by its investors – a risk which may threaten the existence of virtually any bank. Due to their financing structure, internationally operating major banks constantly find themselves under the sword of Damocles of a sudden withdrawal of investors’ money. Two factors in particular make a bank’s liability side fragile and expose a bank to the possibility of a run: First, its liabilities tend to consist to a significant portion of big corporations’ cash reserves and current assets. Second, major banks choose to use short-term refinancing on a large scale through the interbank market.

On the one hand, the resulting fragility acts as a disciplining device: The threat of sudden outflows curbs a bank’s risk taking. On the other hand, this positive effect can become a problem when the bank experiences difficulties to withstand a large outflow of their financing sources. When restructuring failing banks, the constant risk of withdrawal by short-term investors creates a lot of time pressure. It is specifically the bail-in threat which motivates investors to run for the door and withdraw short-term liabilities quickly and in an avalanche-like manner. It is important to balance these two opposing forces: the positive incentive effect of private liability through the introduction of subordinated long-term borrowed capital (bail-in bonds) and the negative incentive effect of the risk of a bank run due to the bail-in threat.

Differences in the liquidity between assets and liabilities contribute to the fragility of banks. Liabilities above the level of Total Loss Absorbing Capacity (TLAC) and Minimum Requirements for Own Funds and Eligible Liabilities (MREL) that are not secured savings deposits tend to be very liquid so that banks do not experience a difficulty when refinancing. A crisis period – or rumors of a possible crisis – may, however, lead to a drying up of this liquidity and drive a bank into insolvency. Emergency measures by its home country can then only be expected if the stability of the system as a whole is at stake. As a result, depositors remain insecure and add to the bank's uncertainty.

Overcoming the bail-in threat

The increased risk of a bank run due to the possibility of a bail-in may undermine bank stability. As a consequence, requiring credibly binding bail-in capital that is primarily governed by regulations can reinforce desired market discipline only to a certain extent. However, this was the reason why the equity and debt capital of TLAC / MREL was created in the first place. Therefore, it is necessary to consider the extent to which the disciplining, positive effect of the additional risk of a bank run creates a benefit that outweighs the negative effect on financial stability due to the inefficient liquidation of banks. To curb the negative effect, it is beneficial to restrict the threat of a bail-in to those investors who are involved in the area of the TLAC / MREL capital. Hence, an upper limit should be set in addition to a minimum limit for the amount of bail-inable capital.

We arrive at a division of the liabilities side of a bank balance sheet in three parts: (a) a liable (TLAC / MREL) part, (b) a non-liable (secured) part and (c) a conditionally liable segment between those two for which the amount needs to be determined. The latter part is a mezzanine-like intermediate tier because it is located in the risk-earnings structure between TLAC / MREL and (guaranteed) savings deposits. Depending on the cost-benefit ratio, this mezzanine-type loss absorption zone can also be set to zero and thus omitted.

Public backstop at the European level

The need of a bailout guarantee for run-prone liabilities beyond the aforementioned threshold leads to a follow-up question: How can the credibility of the bailout commitment be ensured? Only large industrialized countries have a sufficiently broad fiscal base to guarantee the secured savings deposits of large banks. In the case of small countries, the ability to pay (and probably also the willingness to pay) arises when their banks are very large. It is difficult to imagine a bailout guarantee above a sufficiently large liability core without fiscal coordination regarding a public backstop at the European level.

Our proposal of a trichotomy of liability corresponds more to the regulatory requirements than a complete bail-in by reducing or even eliminating the unintended consequences of an irrational run on bank liabilities. This does not imply a softening up of the bail-in rules, as these must be enforced with all consistency within the scope of a TLAC / MREL requirement.

Outside of the framework of TLAC / MREL, a consistent regulatory approach should aim to avoid externalities such as a bank run and its impact on other companies in the financial and non-financial sector. Achieving this regulatory balance requires creative effort towards the design of an incentive-based public protection umbrella for the European financial markets and economies.

Martin R. Götz, Jan Pieter Krahnen and Tobias H. Tröger are Professors at Goethe-University Frankfurt and SAFE.


A SAFE Policy White Paper on this topic will be available soon.