The post-COVID-19 recovery: What challenges and roadmap for the banking industry?

This speech is divided into two parts. First, I would like to share some thoughts on the regulatory response to the Covid-19 crisis. My comments will concentrate on the euro area policy reaction and its effects on the performance of the banking industry. Afterwards, I will focus on the impending challenge of how to handle the consequences of the economic crisis on the quality of the balance sheets of banks.

The regulatory response to the Covid-19 crisis

The economic effects of the restrictions on economic activity following the outbreak of the pandemic have been accompanied by an extraordinary response by policymakers.

We have seen unprecedented measures taken both in terms of fiscal and monetary policy. Fiscal policy has provided direct support to families and businesses as well as loan guarantees and debt moratoria implemented through the financial system. Monetary policy has provided liquidity support as well as enhanced credit facilities through interest rate policy and new exceptional measures of quantitative easing.

The extraordinary and coordinated policy response has encompassed banking regulators and supervisors.

In a context of extreme uncertainty, this policy reaction was essential to ensure that banks continued to fulfil their role of providing credit and liquidity to the economy.

The degree of coordination between authorities has been remarkable. National and international regulators and supervisors have acted with a common goal: to promote financial stability and ensure that funding reaches those who need it the most.

Broadly speaking we have seen decisions providing regulatory relief to financial intermediaries, as well as measures restricting the distribution of dividends, in order to increase the availability of capital to absorb losses and support lending in the face of increased uncertainty.

In terms of regulatory relief, the micro-prudential supervisor (the SSM) has adopted decisions based on a macro-prudential perspective, relaxing some of the capital requirements. The European Commission has also carried out a rapid review of the Capital Requirements Regulation,¹ anticipating the application of some elements that offered capital relief and facilitating the implementation of others over time (as with the introduction of the new accounting framework, IFRS 9).

Banks have also contributed, offering their own relief measures for debtors such as extending payment moratoria, granting pre-approved loans or advancing the payment of pensions and unemployment benefits.

All this has helped to preserve confidence and ensure credit and liquidity were extended to the economy at times of maximum need. Credit growth figures show that support has been unprecedented. For example, in Spain, France, Portugal and Italy, credit to non-financial corporations and individual entrepreneurs increased by rates above 7% year-on-year in October 2020, far above the pace registered before the pandemic.

I will review next both the regulatory relief decisions as well as the restrictions on dividends. The goal is to assess the extent to which these have been effective ways to help the banking system confront the challenges brought about by the pandemic.

Regulatory relief measures

The relief measures include those related to the recently introduced accounting framework (IFRS9), as well as the decisions regarding the capital ratios that must be satisfied by banks. In both cases these measures bring us back to the old debate regarding the procyclicality² of the banking regulatory framework and the need to counteract the undesired effects of this procyclicality.

IFRS 9

One clear example of procyclicality is IFRS9, the new accounting framework that has replaced the incurred loss model with the expected loss model as a way to estimate bank provisions. Under IFRS 9, bank provisions should be calculated based on an estimate of future losses rather than waiting until these losses have already materialised.

At the time this standard was being developed,³ some people already warned that its main drawback could be its procyclicality, as it substantially increases the volume of provisions during economic downturns.

Due to the outbreak of the pandemic and the sudden change in economic forecasts, this risk has materialized. This has forced the major accounting and prudential authorities to make the most of all the flexibility offered by IFRS9⁴ in its application, using alternative macro scenarios to smooth out the effect of the abrupt changes resulting from COVID until more information on the magnitude of the current shock becomes available. The implementation of IFRS9, carried out in phases over several years, has also been postponed by the European Commission⁵ to provide banks some temporary relief.

Ideally, we would like to have a system that builds a provision cushion in good times, not one that requires a huge and sudden effort in times of distress. This was emphasized years ago among others by economists at the Bank of Spain, which had a system of this kind (dynamic provisioning).

Capital ratios

The procyclicality of the regulatory framework was known also to be important in the capital regulation of banks.⁶

In theory, banks accumulate capital to be able to handle unexpected losses (and therefore not covered by provisions). Beyond the minimum capital requirements, which follow mainly a micro-prudential approach, banks also accumulate capital buffers,⁷ which are macro-prudential to some extent, to ensure the stability of the financial system as a whole.

In response to the crisis and given the fear that uncertainty could reduce the provision of credit to the economy supervisory authorities such as the ECB have encouraged banks to make use of these capital buffers. However, they have been used by less than expected.

The supervisor estimated that capital relief measures would free up around 140 billion euros⁸ of CET1 capital to preserve the banks’ ability to absorb losses and to provide credit to the economy.⁹ However, according to the supervisor’s own data for the second quarter of 2020, the banking sector’s capital ratios remained at the same level as year-end 2019.¹⁰

Several factors have been proposed to explain the reluctance to use capital buffers.

• On the one hand, there are three regulatory aspects in the current framework that limit the buffer usability:
  • First, restrictions on the remuneration of equity instruments (shares and AT1 instruments) when operating below the combined buffer (MDA limits) and the small relative weight of counter-cyclical components. In fact, the countercyclical capital buffer, which is the only one specifically designed to be released at low points in the cycle, accounted for just 0.1% of risk-weighted assets in the euro area at the start of the pandemic and had not been activated in most EU jurisdictions.¹¹ The capital conservation buffer, on the other hand, is set by law at 2.5% and its use would automatically trigger MDA restrictions.
  • Second, uncertainty regarding the timeline for rebuilding buffers.
  • Finally, we should not forget that there are other regulatory/prudential requirements that are not as risk-based (such as the leverage ratio and resolution requirements) that could also be binding, at least temporarily.
• On the other hand, the role played by market pressure to avoid the erosion of capital buffers should not be underestimated: i) when information is incomplete, the payment of dividends is seen as a signal to the market that a business is profitable, with the result that this capacity to pay out dividends tends to be maintained; ii) banks’ preference for maintaining high levels of capital to be able to take advantage of potential M&A opportunities that may arise at times of crisis,¹² and (iii) the potential increase in funding costs (or discounts on equity prices relative to book value) resulting from a decrease in capital levels.¹³

It is estimated that out of the €140 bn of freed up capital due to the regulatory relief measures, possibly around 1/3 has been used by the industry during the first half of the year. This would correspond to those that can be employed more flexibly, and in particular the coverage of P2R with non-CET1 instruments, as well as the countercyclical buffer. Banks, for the reasons explained, have been on the whole reluctant to diminish the management buffer. It must be remembered, at the same time, that over this period some regulatory activities revising the quality of banks’ models has continued, often times involving higher levels of capital requirements (for example, the TRIM exercise).

Some regulatory and supervisory forums have already started to discuss the possibility of revising the current framework and its future design. One of the possible solutions is to recalibrate the relative weight of cyclical and structural buffers.¹⁴

These proposals seem to be heading in the right direction but there is a risk they will end up not being effective and will merely result in another generalized increase in capital requirements for the sector as a whole.

Let me explain. Irrespectively of the final design of the capital buffer structure, there is still a risk that the desired objectives will not be achieved unless the framework leads to a coordinated and similar use of buffers among banks.¹⁵ As banks interact in a global market, where they compete to raise funds, competitive pressure creates the incentive to stand out from the rest and maintain relatively high capital ratios that convey an image of greater solvency. Such competition also occurs in the eyes of supervisors, which often compare the ratios of different institutions. Consequently, while there may be flexible countercyclical buffer mechanisms that can be optimal at a collective level, competition creates incentives that undermine the coordinated use of those buffers.¹⁶

These factors must be considered when designing the prudential framework. Otherwise, we could end up increasing capital levels without solving the root cause of the problem. In fact, during the months of the pandemic, the problem has not been the lack of capital, as supervisors and regulators have acknowledged.¹⁷ The existing levels have been robust enough to cope with the shock whilst maintaining solvency and providing adequate credit to the economy.

It’s precisely for this reason that it would be better for the authority, at its discretion, to decide to reduce requirements for all banks on a temporary basis. It would thereby act as a coordinator, resolving the problem of collective action.

On the other hand, it should be noted that, unless countercyclical buffers are well designed, the mere expectation of greater capital requirements at low points in the economic cycle could have an adverse effect on lending. There is evidence in the economic literature that, while banks with higher levels of capital (stock) extend comparatively more credit,¹⁸ banks whose capital requirements are increased (flow) tend to reduce their lending until they’ve reached the required levels.¹⁹

It’s important to remember these effects at a time when the capital requirements resulting from the Basel III reform are expected to increase significantly. This could be particularly detrimental given the current economic environment. On December 10th, the EBA published its most recent assessment and it estimates an increase in minimum capital requirements of around 15% over the coming eight years. Not a negligible amount.

At the time of the introduction of the new Basel III capital requirements there was a lively debate between regulators and the industry. The industry was adamant that the new requirements were to be introduced very gradually, to make sure that the effect on credit extension was limited. This argument was quite often dismissed in official quarters as quantitatively irrelevant. Nevertheless, precisely the same argument has been used in 2020 for the regulatory relief of the capital buffers. In fact, according to research commissioned by the European Parliament²⁰ for each 1% of capital relief in CET1, it is estimated that the increase of credit to the real economy amounts to 1,2%-1,5% over the coming 12 months. Not a negligible effect.

Restrictions on dividend payments

In addition to these considerations we need to consider the costs for the industry of a blanket restriction on dividend distribution, which was approved by the supervisor in March and July this year and is subject to further review in the coming days.

The supervisory authorities argue that these restrictions have been imposed with two objectives:

• Ensure that banks retain enough capital to absorb losses and maintain their capacity to provide credit to the economy.
• Avoid that the capital released because of the relaxation of the capital requirements is used to remunerate shareholders.

This measure has always been labelled as extraordinary and temporary.²¹ At a time of maximum uncertainty, the generalised restriction could mitigate the stigma on those banks that may voluntarily opt for limits to shareholders payouts (see, for example, ECB²² and BIS-FSI²³).

However, as the ECB itself acknowledges, this measure is not without negative effects. And I agree.

In my opinion, the negative effects of these restrictions on the sector are relevant and may become very significant if the restrictions are protracted.

On the one hand, the restriction stigmatises the whole industry with respect to other industries which compete for funding in the stock market.²⁴ This makes access to capital markets even more costly, given the negative signal being sent to investors, an issue also recognised by the ESRB²⁵ and ECB²⁶.

On the other hand, the restriction discriminates among banks, particularly those based in different jurisdictions. Blanket restrictions that are not based on an individualised analysis penalise the most solvent banks, with a negative effect on the overall sector’s performance.

Both effects are particularly undesirable in a context of structural challenges such as consolidation and the need to access capital markets to comply with resolution requirements.

In fact, even if the measure is considered as exceptional, the mere fact that it has been established will make bank valuations and the cost of financing more expensive on a structural basis.

I believe that, going forward, this approach should be reviewed. The supervisory framework is sufficiently robust. Formulas can be found that confine these restrictions to the specific situation of each bank, in line with the risk profile and the supervisor’s assessment. Otherwise, not only are we penalising the industry across the board, but we are also weakening the supervisor’s role in its overseeing function.

Crisis exit management and non-performing loans (NPLs)

The second topic I would like to focus on is the challenge that lies ahead with regards to managing the economic crisis provoked by the pandemic and the expected increase in NPLs.

At a time of heightened uncertainty, public authorities have tried to strike a balance between strengthening the banking sector and providing the flexibility needed to stimulate and sustain lending.

As the crisis evolves and uncertainty decreases, there is a growing incentive for authorities to focus on their micro-prudential role, concentrating primarily on bank solvency and less on the macroeconomic effects of the regulatory measures, with the potential risk of discouraging the extension of credit to the economy and, thus, slowing down the economic recovery.

How can we strike a balance during this period of transition?

Three factors seem relevant to me: i) how quickly the health-related uncertainty decreases and economic activity picks up, and the extent to which government support to households and businesses is maintained throughout; ii) the ability of banks to manage credit risk and adequately increase provisions; and iii) and, finally, the effect of the regulatory framework and market institutions in tackling the NPLs accumulating on bank balance sheets.

i) The reduction in uncertainty and the pace of recovery from the crisis will depend on how the pandemic evolves and the government’s response to further waves of contagion. Should restrictions on economic activity be maintained, it will be necessary to prolong the measures that provide support to those sectors most affected by lockdowns.

• This support must be withdrawn gradually so that viable businesses and households can generate liquidity as their activity returns to normal. When appropriate, funding should not be limited to new loan facilities but should include direct equity support, to prevent excessive leverage in those industries most affected.
• The gradual withdrawal of support will maintain viable businesses afloat and prevent a sudden worsening of the quality of banks’ balance sheets. The withdrawal of support needs to be synchronised with the economic recovery.

ii) The right amount of supervisory pressure will also depend on the extent to which banks apply prudent risk management and are able to provision adequately.

• In fact, despite government support, a significant number of companies are likely to be unable to meet their financial obligations. It should be possible to restructure and/or liquidate these companies.
• This might happen to companies with financial difficulties or others that simply have no future due to structural changes (for example, reorientation towards a more sustainable economy).
• It is therefore essential to ensure that banks apply prudential principles when classifying borrowers and recognise and provision impaired asset portfolios.
• The problem lies in doing this efficiently. It is necessary, first, to differentiate between viable and non-viable companies. Second, viable projects must have access to financing or be able to restructure their financial obligations.
• And that’s the challenge. The current regulatory framework for classifying NPLs, which is largely based on guidelines from the authorities, leaves little room for a differentiated response; exposures that require refinancing/restructuring will ultimately be classified most of the time as “non-performing”.
• This substantially reduces the incentive for banks to offer forbearance to borrowers that are considered viable in the medium term. In fact, the existing framework seems to be designed primarily to encourage the disposal of such exposures.

iii) Which brings me to the last point to consider: the extent to which market conditions facilitate the sale of NPLs by banks.

• The imminent review of the EC Action Plan to tackle the possible increase in NPLs is significant in this respect. The Plan includes initiatives that aim to encourage securitisation and improve transparency, as well as liquidation and insolvency procedures. It also proposes a network of asset management companies (AMCs or “bad banks”) to help tackle NPLs.
• The problem is that this framework was set up in response to the previous financial crisis. But there are significant differences between the previous crisis and the current one that must be considered. The assets that are expected to be most damaged by this crisis have very different characteristics from those in the previous crisis. In this crisis, NPLs are likely to be concentrated among SMEs, extremely heterogeneous and mostly without much valuable collateral.
• These types of assets may not fit very well with the bad banks used in the previous crisis. Nor it is clear what management advantages are offered by those institutions compared with the banks themselves, which are usually much better prepared to manage this type of asset. Relationship banking has a comparative advantage when dealing with this type of assets.

In summary, we are faced with a problem of asymmetric information in a world of uncertainty. In an ideal world, with complete information and clarity regarding the developments in the pandemic and the economy, the actions to be taken would be clear.

The supervisory authorities would force banks to book provisions for their exposure to non-viable companies and, in turn, they would encourage the refinancing of viable projects, not penalizing them in terms of provisions. In this way, credit provision would support the economic recovery and the viability of sustainable businesses, which in turn would benefit the financial health of the banks.

But uncertainty and asymmetric information can create perverse incentives: i) businesses that are less likely to stay afloat will try to convince banks of their creditworthiness to access further financing; ii) and there will not be sufficient incentives for banks with excessive exposure to non-viable businesses or weak pre-provision profits to recognise the potential deterioration of their assets.

We must be able to design efficient risk management and prudential mechanisms, based on reliable financial indicators, that can help us navigate such periods of uncertainty but without encouraging the maintenance of non-viable companies. I am afraid that our current framework for identifying and treating NPLs is way too rigid.

As I said at the beginning, it’s a question of striking a balance. An equilibrium that ensures that viable companies can still access the financing they need and, at the same time, guarantees the sustainability of the banking industry, thereby producing a virtuous circle that encourages economic recovery and reduces the risk of generating a new financial crisis.

Conclusion

The banking industry has been able to provide the necessary response in the first phase of the crisis. It has been part of the solution, rather than a source of problems.

The regulatory response has been quick and, to a large degree, quite effective. Nevertheless, it has also brought into the open some of the shortcomings of the new regulatory framework. Specifically, the accounting framework is excessively procyclical and the capital requirements framework lacks the flexibility to be used as a countercyclical tool in times of stress.

Moreover, the use of blunt policy instruments such as industry-wide restrictions on dividend distribution, while understandable in times of extreme uncertainty, has important adverse effects, both for the industry performance as well as for the credibility of the overall regulatory and supervisory framework. As such, it should be an instrument left for truly exceptional circumstances and lifted as soon as possible.

Going forward, it is of the utmost importance that the authorities strike the right balance between ensuring that the banking industry is adequately prepared to tackle the expected increase in non-performing exposures and their continued role in the provision of financing to household and non-financial corporations. Whilst it is understandable that prudential regulators may wish to ensure that individual banks are sound and solvent, it is important to consider also the aggregate, systemic perspective. Erring on the side of excessive caution and provisioning may thwart the extension of credit to the economy and have a negative impact on the perception of the industry by capital markets. Ultimately it could undermine the role of the financial sector as a positive contributing factor to the economic recovery.

Authors note: This blog is based on a speech at the SUERF-EBF Conference on “Banks’ funding & revenue prospects in the low for long era“, on 11 December 2020. I would like to thank the team at CaixaBank for their support when preparing this talk. Enric Fernández, Montserrat Martínez, Sandra Jódar-Rosell and Tatiana M. Reyes provided very significant input.

Endnotes

1   Commonly known as the “quick fix” to the Capital Requirements Regulation (CRR).

2   Gual, J. and S. Jódar-Rosell (2014) “La prociclicidad del sistema financiero tras las reformas“, Documento de Economía “la Caixa” no. 27, January 2014. See also Repullo, R. and J. Saurina (2012) “The countercyclical capital buffer of Basel III: A critical Assessment” In CEPR: The Crisis Aftermath: New Regulatory Paradigms, London.

3   Ibid.

4   “ECB Banking Supervision provides further flexibility to banks in reaction to coronavirus”. ECB Press Release, March 20, 2020; “Accounting implications of the COVID-19 outbreak on the calculation of expected credit losses in accordance with IFRS 9”. ESMA, March 25, 2020; “Statement on the application of the prudential framework regarding Default, Forbearance and IFRS9 in light of COVID-19 measures”. EBA, March 25, 2020; “IFRS 9 and covid-19”. IFRS, March 27, 2020; “IFRS 9 in the context of the coronavirus (COVID-19) pandemic”. Andrea Enria, letter to the significant banks, April 1, 2020; “FAQs on ECB supervisory measures in reaction to the coronavirus”. ECB, July 28, 2020.

5   Agreement to review the CRR (quick fix), European Parliament, June 19, 2020.

6   Gual and Jódar-Rosell, ibid.

7   The capital conservation buffer (CCB), countercyclical capital buffer (CCyB) and systemic buffers. The CCB was introduced to ensure that banks have an additional layer of usable capital that can be drawn down when losses are incurred. The CCB was implemented in full as of 2019 and is set at 2.5% of total risk-weighted assets. The CCyB aims to protect the banking sector from periods of excess aggregate credit growth that have often been associated with the build-up of system-wide risks. Both the CCyB and the systemic buffer are determined at the country level.

8   “ECB Banking Supervision provides further flexibility to banks in reaction to coronavirus”. ECB Press release, March 20, 2020.

9   The total amount would correspond to P2G (90bn€), redistribution within P2R (30bn€) and the countercyclical buffer (20bn€).

10   “ECB publishes supervisory banking statistics for the second quarter of 2020”, ECB, October 6, 2020; “Introductory Statement“, Andrea Enria speech at the virtual meeting of the European CFO Network organised by the UniCredit Group, June 12, 2020.

11   Behn, M., Rancoita, E., Rodriguez d’Acri, C. (2020), “Macroprudential capital buffers – objectives and usability”, ECB Macroprudential Bulletin, Issue 11, October 2020.

12   Berger, A.N. and C.H.S. Bouwman (2013), “How does capital affect bank performance during financial crises?”, Journal of Financial Economics, Vol. 109 (1), July 2013.

13   Andreeva,D., Bochmann, P., Couaillier, C. (2020), “Financial market pressure as an impediment to the usability of regulatory capital buffers”, ECB Macroprudential Bulletin, Issue 11, October 2020.

14  “If so, we should consider whether countercyclical, releasable buffers are superior at least in terms of usability on rainy days. This can also call for a recalibration of the buffer structure, with a greater role for buffers that can be switched off by the authorities”. Campa, J. M. (2020) “The regulatory response to the Covid-19 crisis: a test for post GFC reforms”. Speech at the Italian Banking Association (ABI), September 21, 2020.

15   Behn, M., Rancoita, E., Rodriguez d’Acri, C. (2020), ibid.

16   “Importantly, for this benefit to materialise, it is necessary that the banking system collectively takes this positive externality into account in its lending behaviour. This may not necessarily be the case, as collective action problems may prevent individual banks from factoring the social benefits of the continued provision of key economic services into their private lending decisions.” Behn, M., Rancoita, E., Rodriguez d’Acri, C. (2020), ibid.

17   “The regulatory response to the Covid-19 crisis: a test for post GFC reforms”. Speech by JM Campa at the Italian Banking Association (ABI) on 21 September 2020.

18   “In a bank-level study with time and firm fixed effects, we have found that higher bank capital is associated with greater lending, and that the mechanism involved in this channel is the lower funding costs associated with better capitalised banks.” Gambacorta, L., Song Shin, H. (2016), “Why bank capital matters for monetary policy”, BIS Working Papers No. 558, 2016.

19   “Second, in the year following an increase in capital requirements, banks, on average, cut (in descending order based on point estimates) loan growth for commercial real estate, for other corporates and for household secured lending.” Bridges, J., et al (2015), “The impact of capital requirements on bank lending”, Bank of England Working Paper No. 486, 2015.

20   See Matyunina, A. and S. Ongena (2020), “Has the relaxation of capital and liquidity buffers worked in practice?”, In-Depth Analysis requested by the ECON committe, European Parliament, October 29.

21   “ECB asks banks not to pay dividends until at least October 2020”, ECB, March 27, 2020; “EBA provides additional clarity on measures to mitigate the impact of COVID-19 on the EU banking sector”, EBA, March 31,2020; “Recommendation of the ESRB of 27 May 2020 on restriction of distributions during the COVID-19 pandemic (ESRB/2020/7)”, ESRB, May 27, 2020.

22   ECB comments on the proposed amendment to CRR2

23   Borio, C. and Restoy, F. (2020), “Reflections on regulatory responses to the Covid-19 pandemic”, FSI Briefs nº 1, BIS, April 2020.

24   While the cost of capital for the banking industry in the euro area has decreased between March and October 2020, the industry specific risk premium (beta) has increased by around 70 bp as a whole (see “Recent developments in the cost of bank equity in Europe”, Boletín Económico, 4/2020, Bank of Spain, September 23, 2020).

25   “Banning dividend payments may undermine the relationship between a bank and its investors. This could potentially restrict the bank’s future access to market funding. Capital instruments subject to the restrictions might become less attractive to investors, particularly compared with instruments of entities in jurisdictions that do not impose such restrictions. In turn, this could reduce the ability of the affected banks to raise additional capital, or it could increase their cost of capital“ ESRB (2020) “System-wide restraints on dividend payments, share buybacks and other pay-outs” , June 2020.

26   “I am well aware that healthy banks need to be attractive to potential investors, and I am also aware that a regular flow of dividends at euro area banks has been an important factor for equity investors, as profitability remains persistently low” (…) “I want to reiterate that this is an exceptional and temporary measure to deal with an exceptional and temporary situation”. Enria, A. (2020) “The current crisis is a wake-up call”. Supervision Newsletter, May 2020.31

This blog post is part of https://voxeu.org/