Industry veteran Jamie Lloyd Evans discusses Basel III in light of concerns regarding capital adequacy and bank lending, and offers some thoughts on potential design improvements to the capital framework.
The Kunstmuseum in Basel is home to a sixteenth century work by Hans Holbein the Younger that is commonly known as the Dead Christ: a life-size painting of the body of Christ that depicts the marks of suffering, violence and putrefaction in unflinching detail. Dostoyevsky was deeply affected by the painting and included a discussion of the work in one of his novels, in which the protagonist declares that “a person looking at this painting could lose his faith”. The statement bears witness to the collision of the believer’s faith with the brutal realities of the material world (i.e. the metaphysical versus the physical) and provides the basis for a re-evaluation.
In a similar manner, the world is currently re-evaluating established norms and ‘truths’ in light of the Covid-19 health crisis. The Basel Committee on Banking Supervision (BCBS) and the BIS (Bank for International Settlements), which both sit a stone’s throw from Holbein’s painting, are likewise being compelled to re-assess aspects of the Basel III framework for bank capital and liquidity.
The first and most obvious question is whether the framework is sufficient to maintain bank solvency. The second question – one which has taken on increased importance in recent months – is whether banks can be encouraged to continue to extend credit to households and businesses in the current environment .
The legitimacy of this second question is predicated on the view that if banks choose to withhold credit in this environment, credit losses will crystallise, provisions will rise, and bank capital and liquidity will be reduced – potentially exacerbating an economic downturn. Or put more pithily by the Financial Stability Institute (FSI):
“It is entirely reasonable for a bank to pull back its horns in a downturn: the fragility of borrowers is more apparent and, by itself, the individual bank can have little impact on the economy at large. But if all banks retrench at once, they may amplify the downturn…”
There are nevertheless at least four reasons to think that the use of the framework to encourage bank lending will be no easy task.
Firstly, Basel III reflects the regulatory and political sentiment that was current in the last financial crisis more than a decade ago. The immediate concern at that time was to fix the inadequacies of the previous regime, but arguably the underlying animus of the framework was punitive in intent and effect. The bad actors at the centre of the 2008 crisis were seen as the banks, and under Basel III, they would be made to pay for government support by holding more capital and higher quality capital.
In marked contrast, the current pandemic has initially cast the banks in the role of government agents and distributors of government support, with the presumption that banks have been willing partners in executing public policy.
However, to use the language of psychology, a sudden switch in regulatory posture is more likely to result in a form of cognitive dissonance rather than a character reassessment of the framework. This could manifest itself in two types of reduction: (a) that the current constraints on risk and capital within the Basel III framework are – and always have been – excessive; or (b) that the encouragement to lend and operate at lower capital levels is bogus, and will lead to a capital poverty trap when regulators revert to type.
Both of these reductions present challenges for effective policy-making and both could have far-reaching implications for bank behaviour in light of any changes made to the framework.
Buffers and disincentives
Secondly, the design of Basel III is complex. The framework provides a degree of flexibility in the form of capital buffers that sit on top of the minimum requirements. Banks are permitted to draw down their buffers in times of distress, but if we were to turn the clock back to 2009, the BCBS conceived the draw-down as less an act of volition but as a consequence of incurred loss:
“Outside of periods of stress, banks should hold buffers of capital above the regulatory minimum. These buffers should be capable of being drawn down through losses…” (author’s emphasis)
Regardless of original intent, the apparent flexibility afforded by the buffer is largely negated by disincentives that are hard-wired into the framework: draw-downs result in restrictions on capital distributions (dividends, buy-backs, and bonuses). Unsurprisingly, most banks, supervisors, and other stakeholders have regarded the buffers as de facto requirements from the outset.
Blanket restrictions on banks’ ability to make certain types of distributions (principally dividend payments and share buy-backs) since the beginning of this year could be regarded as an effective removal of the disincentive. However, the indiscriminate nature of these restrictions penalises more prudent and stronger banks, and the negative impact on private sector appetite to invest in banks may prove costly in future.
Thirdly, it is too easy to presume a bivariate relationship between bank capital levels and regulation – i.e. that the Basel framework is the principal determinant of bank behaviour. In normal times, the binding constraints for bank capital managers are represented by market norms in the form of rating agency capital and shareholder expectations.
This is (or was) the underlying theory of market discipline embodied in Pillar 3 of the Basel framework, which, like an embarrassing relative, is acknowledged but largely ignored by the rest of the family. To encourage banks to operate at capital levels that are acceptable (albeit temporarily) to the regulator but unacceptable to the providers of capital is to risk acrimony.
One particular instance of this problem is found in Additional Tier 1 (AT1) instruments that could be written down to zero or converted to equity if the bank’s CET1 ratio falls below 5.125%. Any encouragement to draw down the buffers into sub-7% levels would create precipitous trigger risk for institutional and domestic retail investors alike.
But even if the framework could – and ought to – bulldoze the private sector, we should also be aware of the role played by national authorities in implementing the framework. While these authorities are often supportive of greater policy flexibility at the international level, the domestic implementation and supervisory practice is often much stricter. There are sometimes good reasons for applying stricter local standards, but ultimately this exacerbates fragmentation and compounds home-host differences.
Using Australia as an example, the raising of minimum requirements to so-called ‘unquestionably strong’ levels while concurrently pursuing domestic market reform seems to have undermined the concept of a capital buffer and contributed to the crowding out of the countercyclical capital buffer (CCyB) in particular.
The point is that even if adjustments to the capital adequacy framework could be made to encourage banks to lend, the transmission mechanism is not particularly reliable.
Undermining risk management
Finally, the danger of encouraging banks to pursue risk-taking activities is that it could become self-defeating if it subtly undermines the risk management practices and processes that are foundational to the capital framework. A bank’s main function within the economy might be to allocate capital, but its basic skill-set is that of risk management – and good risk managers tend to protect their own balance sheet first and foremost.
Providing greater flexibility to banks in applying guidance on ECLs (expected credit losses) is arguably a reporting sleight of hand as the underlying economic position (whatever that may be) is largely unchanged, unless one believes that the act of reporting also has the capacity (through a series of market reactions) to affect the value of the underlying.
If banks use their professional judgement and simply choose not to maintain pre-Covid levels of credit extension or exercise forbearance in relation to defaulted loans, then any encouragement to act differently presents us with an indirect claim that the supervisor/government may know how to run a bank better than the current management.
Given these challenges what then can be done?
Political and particular lenses
One lens that may help us understand the mental discomfort caused by the potential conflict between capital adequacy and bank lending objectives is to recognise them as essentially different in nature: the former is a question of prudential supervision (in fairly mechanical terms), while the latter is ultimately a question of political determination.
The political aspect exists firstly in the sense that the expansion of bank balance sheets could facilitate a substitution of bank lending in the place of government expenditure. This may not necessarily be inappropriate, but there is reason to be cautious if banks are – intentionally or unintentionally – used as ‘off-balance sheet’ vehicles of government, given that taxpayers may well bear a higher cost in the final assessment.
Moreover, while the pandemic has led to demand for credit that is widespread throughout the economy, the allocative process will inevitably be political in nature as different constituencies jostle to be at the front of the queue. Most importantly, political posture has a bearing on the enforcement of property rights, which further complicates the relationship between governments and banks at this point in time.
To avoid misunderstanding, I believe that encouraging bank lending in the current environment is well-intentioned, but the catalyst to lend may well be extrinsic to the capital framework, particularly in some form of government-initiated risk-sharing or back-stop arrangements. The framework in this sense needs to be accommodative of these efforts, and to a large extent it already is – in terms of risk weights and ECLs to government-guaranteed exposures – but this may not necessarily provide the motivation for banks to lend.
A slightly different lens, to which the earlier quote from the FSI alluded, might help us re-frame the question of bank lending as a macro- or system-level concern, whereas capital adequacy is principally and practically an issue for individual institutions.
This may explain the difficulty in designing incentives for banks that exhibit a degree of heterogeneity within the same system. A general framework that encourages all banks to lend but only leads to weaker banks drawing down their buffers while other banks refuse to expand their balance sheets would be a suboptimal outcome.
These lenses may not lead to any solution, but the perspective of history would teach us that banks that choose not to cooperate with their governments eventually see their advantages eroded by new entrants that are friendlier to government policy. Should new life be breathed into the old concept of the state-controlled or state-owned bank, or has the time arrived for the entry of BigTech in banking?
Contours of a new design
This discussion has focused on the design of the risk-based capital framework, but there is vast scope to examine specific risk weights and the complex interaction of capital rules with liquidity and leverage ratios in more detail.
Some challenges to the effective modification of the existing framework have been outlined above, but what would the contours of a new design possibly look and feel like?
At risk of being the fool that rushes in where angels fear to tread, I would suggest the following considerations:
Folding the CCB into the Minimum CET1 requirement: In any discussion of drawing down the buffers the elephant in the room is the fact that the minimum CET1 ratio is set at only 4.5%. This was a great improvement on Basel II, but in today’s terms many would regard this level as too close (to zero) for comfort. Folding the CCB into the minimum ratio requirement would produce a minimum of 7%, which would be more consistent with market practice and expectations. However, this minimum requirement should also represent the maximum requirement in implementation, with all additional Pillar 1 elements being included within a buffer that sit on top of the new requirement.
Increasing the variability and size of capital buffers: In contrast to the fixed CCB, the CCyB is variable at the discretion of national authorities and can be dialled up to the maximum (2.5%) or down to zero at the discretion of the national authorities. Despite the patchy uptake of the CCyB, I would suggest that a higher minimum requirement might encourage more proactive use of a variable buffer by regulators, and would envisage a buffer that combines elements of (a) national discretion to raise capital ratio levels due to the particularities of domestic circumstances, and (b) the existing CCyB. The range of this buffer would fall somewhere between zero and 5%, with the presumption that in normal economic conditions the buffer would be set close to the midpoint of 2.5%, and in a crisis be set at zero. National authorities would have ultimate discretion to set the buffer level within this range but would be expected to review and reset the buffer on a proactive and regular basis.
Removing restrictions on distributions from the buffer guidelines: The pandemic has taught us that there are occasions when restrictions on distributions ought to and can be applied regardless of reported capital levels, and national authorities in most cases have the powers to do so, regardless of a buffer draw-down. Removing automatic restrictions would not and should not preclude the imposition of restrictions by national authorities in any way, but may provide banks and regulators with improved design flexibility to encourage balance sheet expansion in exceptional circumstances.
Reclassifying AT1 capital: The predominance of CET1 as going concern capital and the uncertainty and complexity of how AT1 will perform in distressed scenarios may be a sufficient reason to reclassify AT1 and Tier 2 as ‘other’ forms of capital. This simplification may also afford an opportunity to recalibrate existing conversion or write-down triggers.
Reviewing Too-Big-to-Fail (TBTF) reform: Bank rescues in the form of acquisition are often undertaken by domestic peers, but for global banks, the additional G-SIB capital surcharges inhibit any significant cross-border bank M&A. National politics have also played a role in this development (to the benefit of the D-SIBs) but some form of leniency for global banks may be beneficial particularly in anticipated stress scenarios. This could be achieved by reducing the additional capital buffer corresponding to each of the G-SIB buckets by 0.5% points, i.e. G-SIBs in Bucket 2 would be expected to hold additional capital of 1% instead of the current 1.5%. This is a simple re-calibration rather than change in design, but if the real concern is that bank balance sheets are now ‘too small to save’ the economy, then TBTF needs to be reframed.
Unfortunately, all of these ideas would be better considered during peace time, and none of these suggestions represents a ‘quick fix’. The underlying premise of the above is that – somewhat counterintuitively – flexibility and behavioural change might be encouraged by raising (and fixing) the minimum requirement, and providing national authorities with the latitude to deal with the particularities of local market structure and local politics through an enlarged buffer.
There is a curious inconsistency in the accounts of Dostoyevsky’s reaction to Holbein’s Dead Christ. Forty years after the event, his wife, Anna Dostoyevsky, wrote in her memoir of the “devastating impression” of the painting on her husband. However, in her diary entry at the time, she described her husband as being in a kind of ecstasy, standing on a chair and proclaiming Holbein to be a great artist and poet, at odds with her own sense of revulsion for the painting in question.
The pandemic will invoke uncomfortable and contradictory responses. Whether these reactions undermine trust in the bank capital framework and its relation to the broader economy may depend on our ability to confront the realities of our current situation without presupposing the outcome.
Jamie Lloyd Evans is aseasoned financial services practitioner who has experience as a banker, regulator, and advisor in the fintech space. Most recently, he led the Financial Institutions Group at Citigroup in Singapore and Hong Kong, straddling both the investment banking and corporate banking businesses. Prior to this, Jamie held policy roles at both the MAS (Monetary Authority of Singapore) and the UK’s FSA (Financial Services Authority), where he advised on all aspects of prudential regulation for banks and insurers, particularly in the areas of capital adequacy and corporate treasury.
 Para 248, https://www.bis.org/publ/bcbs164.pdf, Strengthening the Resilience of the Banking Sector, December 2009
 Cf. Calomiris & Haber, Fragile By Design, on the role of US banks in relation to the Community Reinvestment Act and the subsequent subprime mortgage crisis