In remarks published in late March, Haldane argued for a reconfiguration of banks capital structures, which would bake-in the benefits of simplicity, robustness and timeliness. Integral to this, he suggests, is the requirement to issue cocos alongside equity. These coco instruments should have triggers based on market-based measures of solvency and the triggers should be graduated, stretching up the banks capital structure with the coco converting to equity on triggering.
The first of these points would mark a radical departure for this nascent market. Those banks that have issued contingent convertibles so far (Lloyds, Rabobank, Credit Suisse and Bank of Cyprus) have set trigger levels based on the banks Basel III common equity tier 1 ratios. The earlier Lloyds and Rabobank deals set single triggers at 5% and 7% respectively, while the Swiss finish encapsulates both a high-strike trigger and a low-strike trigger. Credit Suisses recent $2 billion public regulation S buffer capital note trade, following Sfr2.5 billion and $3.5 billion private placements, incorporates three tranches of cocos that convert to equity at a 7% CET1 high-strike trigger (to recapitalize the bank back up to its 10% regulatory minimum). The Swiss regulator expects further issues which convert at 5% CET1 essentially non-viability of the bank.
Haldane proposes that rather than using CET1 to trigger conversion the structures should use simpler market-based measures, which can be calculated on the back of an envelope by a competent clerk. These include the ratio of a banks market capitalization to its total assets; a leverage ratio of its market capitalization to total debt and the market value of a bank's equity to its book value (Tobin's Q).
A series of such triggers, he argues would alter potentially quite radically incentives, and thus market dynamics, ahead of banking stress becoming too acute. The system would incorporate a speculative disincentive by using measures over a 30-day period and rules preventing the covering of short positions with the proceeds of a coco conversion.
Haldane says that market-based measurements introduce much-needed simplicity and clarity to coco triggers, and that having a series of pops (trigger points) provides better transparency about the level of banking stress. Knowing that a trigger might be close to being pulled, and their claims converted, coco investors are likely to sit up and take notice, he says, adding that management are less likely to sail close to the wind or at least will be quicker to tack when fearing a squall if this is the case.
Many bankers disagree with his thesis, although they are not inclined to go on the record to say so. This creates more problems than it solves, warns one banker speaking not for attribution. Putting ourselves in the hands of market-based triggers is a tacit admission that the regulators cant keep up with reality. CET1 is, of course, a less dynamic measure but it mitigates volatility. Using market-based triggers can lead to unintended consequences one of which is more volatility, he adds. While Haldane suggests that using 30-day measures would dampen this down, critics additionally claim that CET1 provides a fuller picture. Market-based triggers wont even be reflective of the current state of affairs let alone be anticipatory, claims one. There is no need for banks to give up on their own metrics and hand over to market-based triggers, he states.
The evidence that Haldane produces to support his case could shine a light on why issuing banks are so reluctant to embrace it. He has examined bank performance from May 2002 to November 2008 to see which banks would have breached an 8% CET1 threshold and which would have breached a 5% market capitalization to book value of total assets threshold and finds a clear distinction between crisis banks (those that either failed, required government capital or were taken over in distressed circumstances) and non-crisis banks. The CET1 ratios for the two groups of banks varied very little over this period, whereas the book value thresholds clearly began to diverge in early 2006, with crisis banks hitting the 5% trigger by late 2007. If the market had been following Haldanes ratios, clear warning signals would have been flashing much earlier.
The distinction between crisis and non-crisis banks becomes a lot less clear, however, once the analysis is continued past November 2008. All 18 of the 'non-crisis' banks experienced a falling share price over [the following two to four months] and a couple were down as much as 60-odd percent from their November 2008 low point, recalls Gary Jenkins, head of fixed income at Evolution Securities in London. On my calculations the market capitalization to total assets of the non-crisis banks would have fallen below 4% in early 2009 and on an individual basis 13 of the non-crisis banks would have breached the 5% trigger level by March 2009. And this cohort comprises of some of what would generally be regarded as amongst the strongest banks in the world.
So while the book value trigger may result in an earlier pop for crisis banks, it also results in an eventual pop for otherwise healthy banks that would not have tripped a CET1 trigger. The reluctance of the issuing banks to adopt market-based triggers is very understandable if it raises the possibility that otherwise healthy banks could trip triggers on contingent capital that they were mandated by the regulator to issue and be forced to raise new capital amid periods of high volatility in bank shares.
It just goes to show how difficult it is to set parameters for coco triggers that are acceptable to bond investors, shareholders, bank management and regulators alike, muses Jenkins.
|Andrew Haldane, |
executive director of
financial stability, Bank