Working Paper: NBER ID: w23999
Authors: Stefan Avdjiev; Bilyana Bogdanova; Patrick Bolton; Wei Jiang; Anastasia Kartasheva
Abstract: The promise of contingent convertible capital securities (CoCos) as a “bail-in” solution has been the subject of considerable theoretical analysis and debate, but little is known about their effects in practice. In this paper, we undertake the first comprehensive empirical analysis of bank CoCo issues, a market segment that comprises over 730 instruments totaling $521 billion. Four main findings emerge: 1) The propensity to issue a CoCo is higher for larger and better-capitalized banks; 2) CoCo issues result in statistically significant declines in issuers’ CDS spreads, indicating that they generate risk-reduction benefits and lower costs of debt. This is especially true for CoCos that: i) convert into equity, ii) have mechanical triggers, iii) are classified as Additional Tier 1 instruments; 3) CoCos with only discretionary triggers do not have a significant impact on CDS spreads; 4) CoCo issues have no statistically significant impact on stock prices, except for principal write-down CoCos with a high trigger level, which have a positive effect.
Keywords: Contingent Convertible Capital; Cocos; Bank Fragility; Financial Stability
JEL Codes: G01; G21; G28; G32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Larger banks (G21) | coco issuance (G33) |
| Better-capitalized banks (G28) | coco issuance (G33) |
| coco issuance (G33) | decline in CDS spreads (F44) |
| coco characteristics (mechanical triggers) (Y90) | decline in CDS spreads (F44) |
| Discretionary trigger cocos (C24) | no significant impact on CDS spreads (F69) |
| coco issuance (G33) | no significant impact on stock prices (G14) |
| Principal writedown cocos (high trigger levels) (G33) | positive impact on equity prices (G12) |
| coco issuance (G33) | reduced bank fragility (F65) |