Did recent Cocos issuances really reduce markets’ perception of financial institutions’ risk?

18 February 2015

Mathilde Fox & Stephan Van Lerberghe

While the Basel 2 agreements envisaged a more risk sensitive approach to define capital requirements, the recent financial crises highlighted the need for more capital of better quality. Indeed, financial regulators are now focusing on the capacity of equity and other subordinated liabilities to bear losses in case of bank failures. From this environment emerged an exotic named bond: the so-called “CoCo” bond. Although it sounds tropical, the first one was issued in Europe by Lloyds in 2009. Since then several banks have been joining the CoCos’ issuers rank. In Belgium, KBC was the pioneer by issuing a 1Bn USD CoCo in January 2013.

According to the BIS statistics, around USD 70bn CoCos have already been issued worldwide, of which 80 % were done by European Banks. The CoCo buyers mainly come from private banks and retail investors in Asia and Europe (52%); asset management firms mainly from the US (27%); hedge funds 9%; are banks 3% and insurers 3%.

A contingent convertible bond (CoCo) is a long-term subordinated debt which will be automatically converted into a predetermined amount of shares (equity tier 1 or 2) or written down once a pre-defined trigger is reached (e.g. common equity tier 1 falling below 7%). CoCos are part of the banking capital structure; somewhere between equity and subordinated debts depending upon the CoCo’s structure (triggers, conversion event and mechanism,…) and its level of loss absorption. That is why they are nicknamed “Hybrids”. The main feature is to make subordinated creditors participating in the recovery phase if any. According to Flaherty, the Canadian Finance Minister :”The use of contingent capital would create self-insurance, prefunded by investors, to protect the solvency of the bank and shift the cost of excessive risk-taking away from taxpayers towards shareholders and subordinated debt holders.”

A CoCo is a tool which helps in recapitalizing banks at a lower cost than an exclusively equity-based requirement. It appears to be a solution for to the “too big to fail” issue, avoiding further public sector injection of capital, distortion of market discipline and competitive advantages. Being a pillar of the real economy, banks have been rescued by government leading to increase public debts while any other commercial company would have first dug into the shareholders’ wallet. Unless they are written by governments (bail-out CoCos), CoCos will help in rebalancing risks from taxpayers to subordinated debt holders and shareholders (bail-in CoCos).

Today the risk seems remote since banks have been re-capitalised and are more closely regulated. Banks currently issuing CoCos are probably the least at risk of converting (capital being higher than before). This explains why those recent coco’s issues were overwritten and placed so quickly. Was it a chase for yield (currently issued between 7 and 9%) that drove investors to buy those riskier and innovative assets? So far, it seems that most of the coco bonds did not convert and had an excellent performance!

Explaining the risk perspective and Anthony’s study and result

Since those hybrid securities should reduce the likelihood of a confidence crisis by restoring the capital cushion in the event of adverse market conditions, we address the following question: Have the recently issued CoCos been considered as a risk reduction of the issuing bank? Globally, academics are split between two groups. On the one hand, those thinking that CoCos simply add complexity and innovation risk, are also convinced that CoCos will make the whole markets edifice even more complex and fragile (Koziol & Lawrenz, 2012). Some of them argue for example that investors can create risk and volatility by shorting banking shares to act as a hedge against the CoCo’s trigger (Maes & Schoutens, 2012). On the other hand, researchers think that CoCos can reduce issuers’ capital cost by improving corporate governance and discouraging excessive risk taking (Flannery & Perotti, 2011). Arguments pro and contra explain the intricacy of studying CoCos.

One could think about the moral hazard problem arguing that risk taking activities of a bank may change after having issued a CoCo. Actually CoCos may increase the likelihood of financial distress by perversely creating incentives that encourage the behavior that leads to financial difficulties. Conversely protagonists argue that CoCo issuance will reduce the bank default probability, consequently reducing systematic risk if banks do not hold other financial institutions’ CoCos in portfolio!

Many studies focus on the pricing complexity and conflicts of interest between CoCos stakeholders: regulators, existing shareholders (dilution problem) and investors. Some studies focus on the importance to develop a pricing model that is sensitive to conversion triggers. For example, premium could be adjustable according to changes in risk which will provide the issuing bank an incentive to act prudentially.Similarly, banks such as UBS and Barclays use CoCos as financial bonuses for top managers. Those bonds being amortizable in case of financial distress can discourage high level employees in taking too much risk and keep in line with the banks long term strategy.

In our current research, we focus on risk perception associated with CoCo as we studied the impact of the CoCo issuance on the risk perception of the market by means of the ‘event study’ methodology. Although investors seem in great demand, the number of issuance is still low probably due to legal uncertainty, pricing complexity and absence of standardized market. Nevertheless, we listed European non bail-out CoCos issuances until May 2013 and carried on a first event study to get an idea about markets’ perception of risk after a financial institution issued a CoCo.To measure these impacts, here are two different proxies used for the risk perception of the market: the 5 years CDS spreads on subordinated debts (iTraxx Europe Subordinated Financials 5-Years Total return Index) and the 5 years CDS spreads on senior debts (iTraxx Europe Senior Financials 5-Years Total Return Index). The 5 year bucket has been chosen because this is the most liquid CDS maturity.

Under the efficient market assumption, markets should react instantaneously to a new source of information. In our case, we assume that the CDS market is a semi strong efficient market. So as to check how a CoCo issuance will reduce the risk perception, we should observe a CDS spread reduction when an issuance is announced. Knowing that the issue announcement is generally made 7 days before the date of the issue, a 21- day event window was chosing to proceed with our event studies and analyzed abnormal CDS spread variations around the CoCos issuance (15 days before and 5 days after).

Our first results reveal abnormal variations on the 7th and 6th day before the date of the issue. There is a confirmation that the market incorporates the information in two ways. First, credit spreads seem to narrow when a CoCo issue is announced i.e. a risk reduction. Secondly, bank’s stock prices had positive abnormal returns between the 8th and 4th day before the CoCo date of issue. This highlights the positive effect on investors’ appetite.

As conclusion … Other CoCo topics frequently addressed

Being aware of the limits and caveats of these results, we will proceed with further investigations as long as the CoCos market develops and grows. Indeed, beside the risk perception, many questions and points of uncertainty remain about CoCos that would justify wealthy academic debates and researches on the subject.

A lot of questions remain about the CoCos’ marketability. The lack of standardized market and ratings are against marketability as most institutional investors (at least European) usually have permission to invest only in rated securities. Interesting work of academics (Pennacchi, Vermaelen, & Wolff, 2011) tends to prove that in moving from accounting triggers to market triggers could help solving those issues. Because of the lack of liquidity, investors seem uncertain about CoCos and may feel confused about the different structure (in terms of triggers, conversion ratios,…) that makes them too difficult to compare in a risk-return perspective.

Finally, the most tricky point comes from regulation uncertainty. Although the EU and the Basel Committee are still yet to opine on CoCos qualification to meet regulatory capital requirements, CRD IV text already argues that the banks’ liability side has to be loss absorbing. Our study confirms that the market considers CoCo as a risk reduction instrument. Furthermore, as CoCos are less expensive than equity, they could be a worthful alternative to fulfill the need of 1.5% additional Tier 1 Capital requested by Basel III. Experts quantify the future market amount to approximately EUR 240bn between 2014 and 2019 with an average issue of EUR 45bn per annum.

This article is based on the Thesis with Great Honors of Anthony Philippe, student at Ichec Brussels Management School.
(CURSIEF) About the authors : Mathilde Fox, is a professor and responsible of Research and Training in Finance at Ichec Brussels Management School. Stephan Van Lerberghe, is Owner at Finance and Risk Advisory.

References :
Flannery, M. (2010). Stabilizing large financial institutions with contingent capital certificates. CAREFIN Research Paper, (04). Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1798611
Flannery, M., & Perotti, E. (2011). CoCo design as a risk preventive tool. Retrieved from http://dare.uva.nl/document/461910
Koziol, C., & Lawrenz, J. (2012). Contingent convertibles. Solving or seeding the next banking crisis? Journal of Banking & Finance, 36(1), 90–104.
Maes, S., & Schoutens, W. (2012). Contingent Capital: An In-Depth Discussion*. Economic Notes, 41(1-2), 59–79.
Pennacchi, G., Vermaelen, T., & Wolff, C. (2011). Contingent capital: the case for COERCs. Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1656994

Leave a Reply

Your email address will not be published. Required fields are marked *