Date of Degree

6-2022

Document Type

Dissertation

Degree Name

Ph.D.

Program

Business

Advisor

Linda Allen

Committee Members

Karl Lang

Lin Peng

Youngmin Choi

Subject Categories

Finance and Financial Management

Keywords

Contingent Convertible Capital, CoCos, Financial Institutions, Regulatory Capital, Macroprudential Policy

Abstract

This dissertation consists of five chapters on contingent convertible capital securities, their macroprudential role within the regulatory capital stack of financial institutions, and the effects of the regulatory environments surrounding them on their design.

Chapter 1 This chapter briefly introduces this dissertation, including its motivation and structure.

Chapter 2 The first part of this chapter describes every design feature of contingent convertible capital instruments (CoCos) and provides a review of the literature on their role as macroprudential tools and on optimal CoCo design. The second part of the chapter offers a complete discussion of the role assigned to CoCos in the regulatory capital stack established by Basel III and European CRD/CRR frameworks. Finally, I reconstruct the evolution of the complete structure of capital requirements during the phasing-in of Basel III in 24 national jurisdictions, including buffer requirements and capital surcharges applied to individual financial institutions.

Chapter 3 Using a comprehensive sample of 720 contingent convertible capital instruments issued by financial institutions between 2009 and 2019, I document shifts in CoCo design that nullify their salutary macroprudential benefits. As regulators assign identical credit to CoCos irrespective of their specific loss absorption mechanisms, banks increasingly moved towards write-down instruments incapable of producing punitive wealth transfers from shareholders to bondholders, thereby removing incentives for bank managers to take preemptive action to avoid a trigger event. The issue extends to equity conversion CoCos, in principle the macro-prudentially superior designs; since regulations fail to impose any requirement in terms of conversion prices, issuing banks are free to set them for their own benefit. After extracting the terms of conversion from the prospectuses of a sample of 156 equity converting CoCos and conservatively estimating their projected wealth transfer at the trigger point, I find they are expected to produce mean and median positive wealth transfers in favor of shareholders comparable to those of mild partial principle write-down instruments. Banks achieve this by specifying fixed conversion prices or floor conversion prices set to cap the amount of equity that debt holders would receive in consequence of a trigger event. The CoCo markets appear aware of the individual instruments’ projected wealth transfers; in a sample of 615 CoCos I find that on the primary market yield spreads to a sovereign bond of same tenor are wider (tighter) for CoCos with projected wealth transfers at the trigger point more (less) favorable to shareholders. Ceteris paribus, a change in the directionality of a wealth transfer equal to 50% of the instrument’s notional value from positive for the shareholders to negative is associated with a tightening in excess of 200 basis points.

Chapter 4 The sparse empirical literature on drivers of CoCo issuance has until now focused on idiosyncratic bank characteristics such as capital position, profitability, asset composition and sources of funding, often relying on the assumption that banks faced rather homogeneous regulatory environments as a result of the Basel III framework they shared. In this chapter I will argue that the most important factors driving CoCo issuance are in fact regulatory in nature, and that they stem from important differences in the structure of capital requirements financial institutions faced both across-jurisdictions and within-jurisdictions. The Basel III framework introduced a Capital Conservation Constraint (CCC) threshold at which automatic mandatory curtailing of discretionary distributions (including dividend payments and discretionary compensation to employees) is imposed. The European CRD/CRR framework implements similar, but stricter, restrictions to discretionary distributions in the form of a Maximum Distributable Amount (MDA) to be computed whenever a bank’s CET1 ratio fails to cover a set of capital requirements including the entirety of the Combined Buffer Requirement. Banks can choose to meet all their capital requirements with common equity, if they so desire. However, both Basel III and the European CRD/CRR dictate that any equity used to meet requirements for which appropriately designed AT1 CoCo instruments could be used is to be subtracted for the purpose of establishing the CET1 threshold at which distribution constraints are applied. Effectively, this puts banks choosing to use equity rather than CoCos to meet their AT1 requirement at increased risk of facing distribution constraints compared to a bank with identical CET1 capital but filling the AT1 capital layer with contingent convertible debt. If a bank underutilizes CoCos in such manner, the market considers the bank as having an ”AT1 shortfall”, a condition describing the fact that by issuing sufficient CoCos to fill the capital layers for which they qualify the issuer could ”free” equity to be used in other capital layers. Since any common equity so released becomes a CET1 surplus against the CCC/MDA, the incentive to close the AT1 shortfall grows stronger the closer a bank is to the threshold. Building upon the work done in Chapter 2, I reconstructed for 141 global financial institutions (including both issuers and non-issuers of CoCo debt) the complete structure of capital requirements each was subject to during the 2009-2019 period; logit, probit and multinomial logit models were estimated to test the hypothesis that CoCos are primarily issued in response to incentives that are regulatory in nature. Across all model specifications, banks are more likely to issue CoCos when doing so to close an AT1 shortfall, particularly if their capital ratios put them close to their CCC/MDA thresholds. Banks with an AT1 shortfall have odds of issuing 1.7 to 2.7 times those of banks without an AT1 shortfall. Furthermore, for banks with (without) an AT1 shortfall the odds of issuing increase for decreasing (increasing) distances to their CCC/MDA thresholds. These results highly statistically significant and robust to the inclusion of country fixedeffects. Banks are also more likely to issue CoCos when their national regulators allow their use to meet a portion of Pillar 2 surcharge requirements, and less likely to issue if in their jurisdictions CoCos do not configure a tax shield benefit for their issuer. From a prudential perspective, these results are troubling. The observed interaction between AT1 shortfall condition and distance to the CCC/MDA threshold means that banks are more likely to issue CoCos when doing so allows them to avoid having to issue equity while still decreasing their risk of facing constraints to their discretionary distributions.

Chapter 5 This chapter summarizes the results, and concludes.

This work is embargoed and will be available for download on Sunday, June 09, 2024

Graduate Center users:
To read this work, log in to your GC ILL account and place a thesis request.

Non-GC Users:
See the GC’s lending policies to learn more.

Share

COinS