How to improve financial stability

CoCo issuances have risen strongly year-on-year since the first CoCos were issued in November 2009 (Figure 1). More than 20 European banks have so far issued almost 100 CoCos, mainly in the past two years. The majority of CoCos have been issued by British and Swiss banks (Figure 2). In the past year, banks from Denmark, Germany, France, Italy, and Spain issued CoCos to raise capital. Recently, CoCos with a temporary write-down mechanism (contingent on the bank having regained its financial health) have become more popular. The total volume of CoCo issuances reached a record level of €28 billion in the first half of 2014. The total outstanding amount of CoCos peaked in the first half of 2014, at €78 billion, with the number of CoCo issuances expected to grow further (BIS 2013, FT 2014).

Figure 1. Annual issuances and volume of CoCos in Europe (in € billion)

Figure 2. CoCos’ share per country (in Europe)

Main drivers of growing CoCo issuances in Europe

There are several explanations for the wave of new CoCo issuances in Europe.

• One explanation is the search for yield and investors’ readiness to take extra risks.

The effective yield on CoCos came out at 8.3% in end-2013, versus 6% on other high-yield bonds. The CoCo-only index initiated by Bank of America Merrill Lynch in end-2013 revealed an average yield on CoCos of approximately 7% in end-August 2014.

• Another important driver for the strong growth in CoCos issuances is the new set of rules and regulations on capital requirements for banks.

The Basel III Accord and the European Capital Requirements Regulation (CRR) specifically allow the use of CoCos as additional bank capital. Some countries moved even further beyond the minimum Basel III capital requirements (e.g., Switzerland).

• There are also indications that European banks have issued CoCos to strengthen their capital position in the run-up to the comprehensive assessment carried out by the ECB (Bloomberg 2014, FT 2014).

Given that CoCos count towards bank capital in the calculation of the leverage ratio, they can also be used to improve this ratio. However, this can also be achieved by issuing new common equity, but that tends to be more expensive; as the interest payments on CoCos are tax-deductible, CoCos are more cost-effective capital instruments than common equity.

• Another reason for issuing CoCos rather than new shares is that they are less restrictive; for example, CoCos do not result in capital dilution for incumbent shareholders.

Financial markets expect banks to issue more CoCos following the implementation of the new Bank Recovery and Resolution Directive requirement that a troubled bank must first write down a part of its liabilities before it qualifies for external support. In addition, the Directive requires a layer of bail-in instruments. CoCos can be regarded as such a bail-in instrument, because their holders will absorb losses in the event of a bank resolution.

• Generally speaking, CoCos qualify as instruments that can be bailed-in and provide an additional buffer between the other bail-in debt instruments and senior bond holders (and depositors).

CoCos are not a panacea

CoCos are complex instruments with a number of points requiring attention.

• Conversion clauses may create market distortions.

That is because holders of CoCos with such a clause can hedge their exposures only by short-selling the underlying shares, which puts pressure on the share price. If CoCos make out a significant part of a bank’s balance sheet, this may result in a death spiral, i.e. a downward price spiral caused by large-scale selling of shares (De Spiegeleer and Schoutens 2013). Academic literature points to the risk of fire sales if investors do not distinguish the CoCos of one bank from those of another (Goodhart 2010). To avoid a downward price spiral, banks could issue CoCos with book-value triggers only, which are not subject to price movements. Without room for supervisory discretion and higher triggers, the downsides of book-value triggers are that there is some delay in the publication of the relevant book values and that book values are subject to the interpretation of accounting standards.

• The preventive effect of CoCos comes into play only if CoCos are triggered well before a potential bank resolution takes place (by allowing the bank to operate as a going concern).

Hence, it is important that trigger levels are high enough. A trigger of 5.125% is relatively ‘low’, given that Basel III includes a 6% Tier 1 capital requirement in tandem with other buffers.[1] By the time the capital ratio has fallen below 6%, banks have already incurred so much losses that a bank resolution is imminent. Higher triggers imply that holders of CoCos are more likely to incur losses. However, given that investors want to be compensated for higher risk, it is more expensive for banks to issue high-trigger CoCos.

• CoCos provide extra protection to senior bond holders and to certain groups of junior bond holders and depositors.

Holders of CoCos absorb losses before other bond holders do. The possible occurrence of a trigger event raises the question of how unsecured depositors and subordinated bond holders would respond. They may react positively if a CoCo is triggered because the size of the CoCo buffer partly determines how much downsize they have to absorb. Also, a larger buffer may increase their willingness to provide funds to the bank. At the same time, there is a risk that these parties, who may be unfamiliar with CoCos, interpret conversion as a negative signal prompting them to sell bonds or withdraw deposits. In that case, conversion or write-downs would create instability rather than stability. In such an environment, banks may find it more difficult to attract new funding.

• Some triggers are discretionary in nature.

The European Capital Requirements Regulation allows the supervisor to intervene if warranted by a bank’s capital position. Decision-making by supervisors about triggering CoCos may result in more uncertainty, and the action taken may differ from supervisor to supervisor. That is why supervisors need a clear policy on the exercise of their discretion in triggering going-concern CoCos.

• It may be difficult for investors to determine their position in the capital hierarchy.

There are situations where the CoCo holder absorbs a loss, while the shareholder can maintain his claim on the bank’s assets. Nevertheless, the shareholder will usually be the first to absorb most of the losses. Setting the CoCo trigger at 5.125% means that the bank has already incurred significant losses which weakened its capital position by the time conversion or a write-down is triggered. This also has consequences for the role of (market) discipline, which shifts from the shareholder to the CoCo holder. This may result in more bank risk-taking (Koziol and Lawrenz 2012). Having said that, CoCos may also actually promote the bank’s discipline because the CoCo holders have a strong incentive to monitor the bank’s risky behaviour and the occurrence of a trigger event puts at risk the bank’s funding options (Maes and Schoutens 2012).

• There are no far-reaching restrictions on holding CoCos.

Financial institutions may hold CoCos issued by other financial institutions, and such cross-holdings may create systemic risk. As is the case with equity investments in financial institutions, a combination of sizeable CoCo holdings held by financial institutions and the occurrence of an event triggering conversion puts pressure on financial stability, with the CoCos spreading contagion (Zhouet al. 2013). This may have a negative impact on the system’s capacity to absorb losses.

Conclusions

The search for yield and the tightening of capital requirements have resulted in a wave of CoCo issuances. Banks have used CoCos to strengthen their capital positions, improve their leverage ratios, and protect senior bond holders against bail-ins. CoCos may contribute to financial stability provided that:

• Banks do not hold large CoCo cross-holdings; and
• CoCos have high triggers based on book value.

This is so because CoCos (and their conditional conversion or write-down mechanisms) strengthen the capital position in a banking crisis, which limits any bail-out costs in the event of a bank resolution.

At the time of writing, no CoCo has been converted or written down, which leaves unanswered the question of how financial markets will respond to the occurrence of a trigger event and what the actual effects of these instruments will be. While acceptance and standardisation of the instrument are growing, it remains difficult to price the risks associated with CoCos.

CoCos may act as a buffer that essentially makes banks more resilient in times of crisis. CoCos, therefore, are similar to share capital and can strengthen banks’ capital position when things go bad and issuing equity may be difficult. Coupon payments are tax-deductible, whereas dividend payments are not, which means that CoCos are a more cost-effective means of strengthening bank capital. While shareholders’ equity remains the main capital buffer, high-trigger CoCos – i.e. with a trigger level well over 5.125% – are also loss-absorbing and thus allow banks to operate as a going concern.

Published in collaboration with VoxEU

Author: Martijn Adriaan Boermans is a policy advisor at the Statistics Division of De Nederlandsche Bank (Dutch Central Bank).Sinziana Petrescu is a Policy Advisor at the Dutch Central Bank. Razvan Vlahu is an economist at the Economics and Research Division of De Nederlandsche Bank (Dutch Central Bank). 

Image: A sign for Bank Street and high rise offices are pictured in the financial district Canary Wharf in London in this October 21, 2010 file photo. REUTERS/Luke Macgregor