The ideal post-EU regulatory framework

Speech given by
Victoria Saporta, Executive Director of Prudential Policy

International Business & Diplomatic Exchange 2020 Annual Conference, London

10 March 2020

I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 1
A lot has changed since the last global financial crisis. Banks are much more resilient. The PRA has
authorised 18 new banks in the past five years, with further banks in the pipeline to be authorised.
The role of non-banks and market-based finance has been growing. And digitalisation is transforming
a number of financial services.

In spite of these changes, the fundamental market failures that justify the need for prudential
regulation remain. This is because financial firms can take actions that create “negative externalities”
– jargon for someone making bets that hurt innocent bystanders when they go wrong. And in the
case of “systemic” negative externalities they hurt the economy at large, as the Great Financial Crisis
reminded us. Equally, these market failures need to be dealt with in a manner that does not introduce
frictions that hurt innovation and long-term productivity and economic growth.

Prudential regulation that deals effectively with these market failures has to be underpinned by an
appropriate institutional structure: a set of clear responsibilities enshrined in law on who sets the
objective for prudential regulation and how, who regulates, and where the details of prudential
regulation should sit.

But what is the ideal institutional structure that would allow us to update, change and simplify these
rules?

This is a timely question. Less than a month and a half ago, the UK exited the EU. The
implementation period is due to last until the end of this year. While we are in the implementation
period, the Government and the regulators will continue to onshore any new financial regulations and
standards coming from the EU that take effect this year. And at the end of the implementation period,
the UK will be left with the same complex set of EU law and regulations onshored in a way that
replicates the rather unique way the EU – as a supranational institution – the area of financial
services. The technical detail of the prudential requirements will sit in a dispersed variety of places:
primary legislation, a range of statutory instruments, on-shored binding technical standards, and PRA
rules and guidance. This ‘patchwork’ naturally makes the framework difficult for firms to navigate –
particularly smaller firms with more limited resources to devote to compliance. It is partly for that
reason that the previous Chancellor Philip Hammond announced in the 2019 Spring Statement that
HM Treasury will be reviewing the post-exit regulatory framework. This review will include ‘the need
to ensure financial stability is delivered through an effective regulatory framework, with the
responsiveness necessary for a dynamic and open financial services sector and an appropriate level
of democratic accountability’.1 The first stage of that review, focused on regulatory coordination, was
launched in July 2019.2 Ultimately, the design of the post-EU regulatory framework will be for
Parliament to decide in due course.

1
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/785618/WMS_final_Commo
ns.pdf.
2
For more detail on the review, see https://www.gov.uk/government/consultations/financial-services-future-regulatory-
framework-review.
I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 2
Last year Sam Woods set out a set of principles to which any future UK regulatory framework should
adhere. This speech builds on these principles to propose an ideal framework from the perspective of
achieving the Bank of England’s objectives. 3 It then looks at whether the existing empirical evidence,
including some new research carried out recently by colleagues in the Bank, supports the framework.

Ideal features of the institutional structure of prudential regulation

Sam Woods’ six principles of regulation are shown in Figure 1. In building the framework, I start by
mapping these principles to three ideal features of regulation also depicted in the Figure.

The first feature is that the structure needs to have dynamism. By this, I mean adjustments made to
regulation over time to incorporate updated international standards in a timely manner, including to
adjust for any unintended consequences of regulation. For example if the international standards are
not designed with certain business models in mind, to reflect the advance of new technologies and to
reflect new risks and opportunities. I have spoken at length elsewhere about the benefits of
dynamism, so will not dwell on this more here.4

The second feature the institutional structure should have is time consistency.

Time consistency is perhaps best understood by describing its opposite, time inconsistency.

Public policy suffers from time inconsistency when there is a conflict between a policymaker’s short-
term incentives and their long-term incentives.

In the context of prudential regulation, over the long run, a policymaker wants to avoid financial crises.
Robust prudential standards are the means of achieving this.

In the short term, however, there may be benefits from weakening prudential standards. For instance,
a government might benefit electorally from the resulting short-term boost to credit supply. There is
empirical evidence for this happening in practice. 5

But the long-term consequence of always acting on those short-term incentives is an unstable
financial system.

The benefits of time consistency have been long recognised in monetary policy.

Finally, the structure will need to have legitimacy.

3
https://www.bankofengland.co.uk/speech/2019/sam-woods-ubs-20th-annual-financial-institutions-conference-lausanne.
4
See https://www.bankofengland.co.uk/speech/2018/victoria-saporta-speech-at-the-westminster-business-forum-keynote-
seminar and https://www.bankofengland.co.uk/speech/2019/victoria-saporta-speech-at-risk-net-liquidity-and-funding-risk-
europe .
5
See Antoniades and Calomiris (2018) and Dagher (2018).
I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 3
To achieve this, the regulatory authority needs to be given a mandate.

A mandate provides a regulator with legitimacy in two ways. First, it ensures there is democratic
control over the shape of the regulatory regime; e.g., the objective or objectives of the prudential
regulation, the types of firm in scope, the issues the regulator must have regard to in making policy.
Second, it makes possible to hold the regulator into account.

Figure 1: Links between the principles of ‘stylish’ regulation and ideal features of the
institutional structure of prudential regulation

Alternative models for the institutional structure

There are broadly four alternative models for the institutional structure for prudential regulation.

The first model is that prudential regulation is specified in primary legislation. The approach to
prudential regulation that the EU takes is an example of this model since for example regulatory
requirements are mostly specified in the Capital Requirements Regulation and Capital Requirements
Directive.

This means that many of the technical details that make up prudential regulation are included in
legislation. This would include, for instance, the mathematical formulae used to calculate risk
weights.

I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 4
This model provides legitimacy because the legislation in which the prudential regulation is set out
has been passed by a parliament. It is also time consistent, because a deviation from regulatory
standards would require a change to legislation.

But it is unlikely to be very dynamic because the only way prudential regulation could be adjusted in
response to an unintended consequence or new risk that has arisen is a change to the primary
legislation, which takes time to happen.

It typically takes nine months to issue a new PRA rule – with that time including policy development,
cost benefit analysis, process for public consultation, and approval by the Prudential Regulation
Committee. In contrast, the process for making changes to primary legislation in the UK – which
includes debates in committees and both houses of Parliament – requires finding Parliamentary time
for legislation, and then after Parliamentary time has been secured, takes longer than a year to
complete. Finding Parliamentary time is not an easy task, especially if the purpose of legislation is
carrying out technical amendments to financial services rules. Public officials need to compete for
attention with all the other important public-policy issues that Parliament and government must deal
with and if they do not find a Parliamentary slot or Parliament is dissolved for elections, the legislation
gets delayed for the next Queen’s speech. For example, over the past six months, such events have
delayed the passage of a Bill that is necessary to implement the final elements of Basel 3 (so-called
Basel 3.1). By contrast, the financial regulator’s task is solely focussed on delivering and maintaining
rules that are fit for purpose for their regulated sector only – there is no competition for other initiatives
and the process and timetable provides certainty to industry.

In a second model, prudential regulatory-setting powers could lie with government ministers.

This model would be dynamic because the government would be able to adjust prudential regulation.
It would be legitimate since the government is directly accountable to the parliament and ultimately
the electorate. But it may not be time consistent because it creates the possibility that prudential
regulation will be influenced unduly by the electoral cycle, as was deemed to have happened to
monetary policy prior to central bank independence.

Alternatively and in a third model, these powers could be given to an independent regulatory body.
This model is likely to ensure that prudential regulation is both dynamic and time-consistent. An
expert regulator would be able to identify and implement necessary dynamic adjustments, and, for
reputation reasons, it would have an incentive to maintain prudential standards provided that it
avoided industry capture. But, absent further accountability mechanisms to enhance it, it would lack
legitimacy.

In summary, each of these models lacks one of the ideal features for the institutional structure of
prudential regulation.

I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 5
The fourth model is an independent regulatory body with a clear mandate set out in primary
legislation and a clear set of accountability mechanisms to Parliament.

This model combines the benefits of an expert regulator that can adjust prudential regulation
dynamically, with an objective that incentivises it to maintain prudential standards, while using the
mandate and the accountability mechanisms to ensure legitimacy.

Table 1 summarises the degree to which these four models possess the three ideal features of an
institutional structure of prudential regulation.

In theory at least, this fourth model is preferable because, unlike the other models, it possesses all
three of the ideal features.

Table 1: Comparison of alternative models of the institutional structure of prudential

regulation
Model Dynamic? Time-consistent? Legitimate?
Primary legislation X √ √
Government √ X √
Regulator √ √ x
Regulator + clear mandate √ √ √
+accountability
mechanisms to Parliament

There are plenty of precedents for this institutional structure of prudential regulation, including in the
UK.

The Financial Services and Markets Act 2000, better known as FSMA, established a regulator in the
form of the FSA and specified the regulator’s objectives and responsibilities. The detailed regulatory
rules, however, were set out by the FSA in its handbook, not written into the Act.

Following the financial crisis, the model for prudential regulation in the UK has been further developed
by the establishment of the Financial Policy Committee (FPC) and the Prudential Regulation Authority
(PRA). The mandates for both the FPC and PRA are set out in primary legislation: the FPC is
responsible for protecting and enhancing the overall stability of the UK financial system; and the
PRA’s primary objectives are to promote the safety and soundness of regulated firms and to
contribute to the securing of an appropriate degree of protection for policyholders.

But importantly, the mandates for the FPC and PRA ensure that other factors such as long-term
productivity and economic growth are taken into account when setting policy. The FPC must pursue
its financial stability objective without causing serious harm to the wider economy in the medium or

I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 6
long term. In other words, the FPC must avoid the ‘stability of the graveyard’. 6 And, the FPC also has
a secondary objective, subject to its pursuit of its primary objective: that of supporting the economic
policy of the government. The PRA also has an important secondary objective to facilitate effective
competition, and must have regard to a set of regulatory principles, including the desirability of
sustainable economic growth. Furthermore, the Chancellor writes regular letters to both the FPC and
the Prudential Regulation Committee, setting out the Government’s economic policy and making
recommendations on how they should discharge their functions.

Importantly, the PRA takes into account the FPC’s views on financial stability when setting
microprudential policy. To give an example of how this works in practice – in 2015 the FPC set out
the optimal overall level of UK capital requirements, balancing financial stability against economic
growth. The PRC then took into account this judgement when setting individual bank capital
requirements. And at the end of last year, the Financial Policy Committee (FPC) raised the structural
level of the UK countercyclical capital buffer (CCyB) rate - that is the rate it expects to set in a
standard risk environment - from in the region of 1% to in the region of 2%. Then, last month, the PRA
consulted on proposals to reduce variable microprudential minimum requirements (so-called variable
Pillar 2A capital requirements) to take account of the additional resilience associated with higher
macroprudential buffers in a standard risk environment.7

In addition to setting out a clear mandate in primary legislation, there are also a number of
accountability mechanisms in place for Parliament to scrutinise the work of the FPC and PRA against
that mandate. For example, PRA senior representatives are expected to appear before the Treasury
Select Committee (and other Parliamentary Committees), when requested. And the PRA Annual
Report is laid before Parliament.

Outside the UK, the model of independent regulators with mandates making rules is common practice
in most other jurisdictions. The need for operational independence is reflected in the Basel Core
Principles for banking supervision.8 All jurisdictions represented in the Basel Committee implemented
the Basel 3 reforms through regulators’ rules or guidelines rather than primary legislation, with the
exceptions of the EU, with its very specific supranational structure, and Switzerland, with its unique
federal structure.

6
Details of the FPC’s objectives and primary responsibilities are set out in section 9C of the Bank of England Act 1998. In
particular, the FPC is charged with ‘the identification of, monitoring of, and taking of action to remove or reduce, systemic risks
with a view to protecting and enhancing the resilience of the UK financial system’, but this does not ‘require or authorise the
Committee to exercise its functions in a way that would in its opinion be likely to have a significant adverse effect on the
capacity of the financial sector to contribute to the growth of the UK economy in the medium or long term’. For a more detailed
discussion of the FPC’s mandate, see my colleague Alex Brazier’s May 2019 speech, ‘Citizens in service, not people in power’,
available at https://www.bankofengland.co.uk/-/media/boe/files/speech/2019/citizens-in-service-not-people-in-power-speech-by-
alex-brazier.pdf. The Governor also spoke about the trade-off as part of his speech the “Grand Unifying Theory (and practice)
of Macroprudential Policy” available at https://www.bankofengland.co.uk/-/media/boe/files/speech/2020/the-grand-unifying-
theory-and-practice-of-macroprudential-policy-speech-by-mark-
carney.pdf?la=en&hash=53A55C800C6B04573DC04FD2F9579079A3503FED
7
See https://www.bankofengland.co.uk/-/media/boe/files/prudential-regulation/consultation-
paper/2020/cp220.pdf?la=en&hash=133449CBB3D0FDD2B03AC6242C99C3F36A150258
8
The second core principle states that ‘[t]he supervisor possesses operational independence, transparent processes, sound
governance, budgetary processes that do not undermine autonomy and adequate resources, and is accountable for the
discharge of its duties and use of its resources’ (https://www.bis.org/basel_framework/chapter/BCP/01.htm?inforce=20190101).
I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 7
Regulatory independence and financial stability in practice

Does the evidence support the theory that an independent regulatory body with a clear mandate and
strong accountability produces better outcomes?

At the Bank, Fraccaroli, Sowerbutts and Whitworth (2019) have been researching whether the
institutional structure for prudential regulation affects financial stability in practice. They have done
this by analysing whether there is a statistically significant association between the independence of
the prudential regulator and/or supervisor and financial stability outcomes.

The authors have constructed for different countries and over time an index of regulatory-supervisory
independence. The index combines three aspects of independence: institutional; regulatory; and
budgetary. The definitions of these aspects can be found in Table 2.

Table 2: Definitions of aspects of independence captured in the Regulatory-Supervisory

Independence index
Aspect of independence Definition
Institutional This captures the involvement of political bodies in the appointment
and dismissal of the head of the regulatory agency, as well as length
of their term. The more independent are the actors in charge of the
appointment and removal process, the more independent the
regulatory agency will be. Moreover, the longer the head’s term, the
more protected the agency is from the electoral cycle.
Regulatory This captures whether the agency needs government approval to
issue binding legislation. Higher regulatory independence allows the
regulator to adapt prudential rules quickly and flexibly and to better
identify with the rule-implementation and enforcement.(a)
Budgetary This is based on whether needs government approval on its budget,
which could affect the regulator’s ability to act independently.(b)

(a) See Quintyn and Taylor (2002).

(b) See page 79 of OECD (2016).

Data on these aspects are drawn from the responses given by supervisors to a World Bank survey
(Barth et al (2013)) and the statutes of the regulatory agencies and amendments to those statutes.

The authors then aggregated these aspects into a single index of Regulatory-Supervisory
Independence, covering 43 countries and the period 1999 to 2019.

I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 8
Chart 1 shows the average regulatory-supervisory index and the global average an index of central
bank independence with respect to monetary policy. Both have increased over time, but central bank
independence started growing earlier and initially as a faster pace.

Chart 1: Regulatory-Supervisory Independence and central bank independence, 1999-2019 (a)

(a) The data on central bank independence is drawn from Bodea and Hicks (2015) and Garriga
(2016). Weighted versions of the indices are used to aid comparison. The weighting is based on
Cukierman et al (1992).

The second step in the analysis was to examine whether there is an association between the
Regulatory-Supervisory Independence index and financial stability.

Defining an ex ante quantitative indicator of the success or otherwise of prudential regulation is hard,
unlike with monetary policy where such measures are easier to arrive at. But fortunately, this is less
of a problem for this analysis because we have ex post observations of distress in the banking
system. The authors used rates of non-performing loans at banks as a measure of financial
(in)stability.9

They analyse whether reforms that increased regulatory-supervisory independence were associated
with subsequent a reduction in the rates of non-performing loans. Panel data regression show a

9
That is, non-performing loans as a proportion of total gross loans.
I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 9
negative and statistically significant association between independence-increasing reforms and rates
of non-performing loans – see Table 3.

Table 3: Results for regressions of non-performing loans as % of total loans on the

Regulatory-Supervisory Independence Index
(1) (a) (2) (a)
Regulatory-Supervisory Independence Index -1.194 *** -1.300 **
(0.405) (0.495)
Regulatory-Supervisory Independence Index -0.970 **
lagged by one year (0.322)
Observations 29,782 28,002
Adjusted R2 0.227 0.245

(a) The regression equation includes bank fixed effects, country fixed effects, and year fixed effects. Standard errors, which
are clustered at the country level, are reported in parentheses. The results are robust under different specifications, including
using a set of bank-specific control variables (total assets, z-score, cost-to-income ratio, efficiency ratio, total debt, total
liabilities, total gross loans) and country-specific GDP growth.

This association is there even as the set of control variables in the regression is varied. The
association is also persistent: the coefficient on the reforms lagged by one year is similar in
magnitude and statistical significance.

The association is also economically significant. A reform that increases the index in a given year is
associated with a 1.3 percentage point reduction in the rate non-performing loans and a 0.97
percentage point reduction in the subsequent year, all else equal.

These results are consistent with what other papers have found. Dincer and Eichengreen (2012) find
that lower non-performing loans relative to GDP in countries in which the supervisor is independent of
the government. Klomp and de Haan (2009) find a negative association between a measure of
financial instability and the degree of central bank independence.

There is also evidence that supervisors that are probably more independent of the banks they
supervise tend to be more robust. For instance, Agarwal et al (2014) finds in the United States
federal regulators are twice as likely to downgrade banks’ supervisory ratings as state regulators.
Ben-David et al (2018) analyse the impact of supervisory responsibility for large banks moving from
national regulators to the ECB’s Single Supervisory Mechanism or SSM. They find banks for whom
the SSM became their supervisor sought to manipulate their balance sheets to fall below the
thresholds for being in scope of the SSM, suggesting banks anticipated that the SSM would be a
tougher supervisor.

I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 10
The results in Table 3 mean that when countries implemented reforms that increased the
independence of their prudential regulator and/or supervisor, non-performing loans as a proportion of
total lending decreased. Greater regulatory independence increased financial stability.

But did greater regulatory independence come at a price of less efficient banks or slower credit
growth, a version of what is sometime referred to “as the stability of the graveyard”?

It seems that it did not – see Table 4.

First, the authors replaced the non-performing loans with a measure of bank efficiency: the ratio
between a bank’s expenses and its income. The results show that the independence index does not
have a statistically significant effect on bank efficiency contemporaneously or lagged by one year.

Second, they used the loan growth rate as a dependent variable. The result is similar: the index does
not have a statistically similar effect on loan growth.

These results differ somewhat from others have found. Dincer and Eichengreen (2013) find when the
supervisor is independent some measures of bank credit are lower. In contrast, Barth et al (2013b)
find supervisory independence is associated with higher bank efficiency.

Table 4: Results for regressions of a bank efficiency ratio or loan growth on the Regulatory-
Supervisory Independence Index
Dependent variable
Bank efficiency ratio (a) Loan growth rate
Regulatory-Supervisory Independence 1.083 0.027
Index (0.992) (0.022)
Regulatory-Supervisory Independence 0.155 0.004
Index lagged by one year (0.367) (0.015)
Observations 30,698 30,698
Adjusted R2 0.737 0.105
(a) Defined as (non-interest expense before foreclosed property expense + amortisation of intangibles + goodwill impairments)
as a percentage of (net interest income + non-interest income).

Conclusion

In this speech, I have set out a framework for thinking about ideal post-EU institutional structure. The
ideal model involves independent regulators with mandates. The evidence backs that independence
of prudential regulation increases financial stability with no cost to efficiency.

I would like to thank Matthew Willison, Hugh Burns, Nicolo Fraccaroli and David Swallow for their assistance in preparing these
remarks 11
 References

Antoniades, A and Calomiris, C W (2018), ‘Mortgage market credit conditions and US presidential elections’,
NBER Working Paper 24459.

Agarwal, S, Lucca, D, Seru, A, and Trebbi, F (2014), ‘Inconsistent regulators: evidence from banking’,
Quarterly Journal of Economics, vol.129, pages 889-938.

Barth, J R, Lin, C, Ma, Y, Seade, J, and Song, F M (2013b), ‘Do bank regulation, supervision and monitoring
enhance or impede bank efficiency?’, Journal of Banking & Finance, Vol.37(8), pages 2879-2892.

Barth, J R, Caprio, G and Levine, R (2013), ‘Bank regulation and supervision in 180 countries from 1999 to
2011’, NBER Working Paper 18733.

Ben-David, I, Cerulli, G, Fiordelisi, F, and Marquez-Ibanez, D (2018), ‘Seeking my supervisor: evidence from
the centralization of banking supervision in Europe’, Mimeo.

Bodea, C and Hicks, R (2015), ‘Price stability and central bank independence: discipline, credibility, and
democratic institutions’, International Organization, Vol.69(1), pages 35-61.

Dagher, J (2018), ‘Regulatory cycles: revisiting the political economy of financial crises’, IMF Working Paper
WP/18/8

Dincer, N N and Eichengreen, B (2012), ‘The architecture and governance of financial supervision: sources
and implications’, International Finance, Vol.15(3), pages 309-25.

Fraccaroli, N, Sowerbutts, R, and Whitworth, A (2019), ‘Does regulatory and supervisory independence
affect financial stability’, Forthcoming.

Garriga, A C (2016), ‘Central bank independence in the world: a new data set’, International Interactions,
Vol.42(5), pages 849-68.

Klomp, J and de Haan, J (2009), ‘Political institutions and economic volatility’, Journal of Financial Stability,
Vol.5(4), pages 311-26.

12

All speeches are available online at www.bankofengland.co.uk/news/speeches

 OECD (2016), Being an independent regulator, OECD Publishing.

Quintyn, M, and Taylor, M W (2002), ‘Regulatory and supervisory independence and financial stability’, IMF
Working Paper WP/02/46.

13

All speeches are available online at www.bankofengland.co.uk/news/speeches 13