Financial regulators: one, two, maybe a few?

In January 2023, a Jamaican brokerage, Stocks and Securities Ltd, told its regulators that almost $20 million had been stolen from client accounts, including around $13 million from the accounts of Usain Bolt, the Olympic gold-medallist sprinter.

Jamaica currently has separate regulators for the banking system (Bank of Jamaica) and the non-bank financial sector (the Financial Services Commission). The FSC regulated brokerages like SSL, but allegedly ignored governance problems at the firm.

After the scandal broke, Jamaican finance minister Nigel Clarke said the country would abolish the FSC and adopt a “twin peaks” model of regulation. The Bank of Jamaica would handle all financial sector regulation, while the rump FSC would handle consumer protection.

Darlene Morrison, financial secretary at the Jamaican ministry of finance, told Central Banking in July that the government will present legislation establishing the twin peaks system in 2025. “We are currently finalising the detailed proposals for the new regulatory arrangement,” she said. A white paper has not yet been published.

The twin peaks model – where there is one prudential regulator and one consumer protection regulator – is not new. The concept was first outlined in the 1990s. Several countries have adopted it since, including Australia, South Africa and the United Kingdom.

But what are the advantages and disadvantages of this model? Why has it not been adopted more widely? And should central banks really take direct charge of all financial regulation, given the scale of the challenge and potential conflicts of interest?

Two regulators to rule them all

In a 2018 law journal article, Ruth Plato-Shinar, a professor of banking law and financial regulation at Netanya Academic College in Israel, defined four major institutional systems of financial regulation:

• A sectoral model, where every sector within the financial services field has its own regulator.
• A functional model, where every activity (eg payments, derivatives trading) has its own regulator.
• A single regulator model, with one agency overseeing prudential and conduct matters.
• The twin peaks model.

Michael Taylor, chief credit officer for Asia-Pacific at Moody’s Investors Service who previously worked at both the Bank of England and Hong Kong Monetary Authority, outlined a twin peaks model of regulation in a 1995 paper, taking the name from the US television programme Twin Peaks. It envisions one “prudential regulator of all systemically significant financial firms – banks, building societies, insurance companies and investment banks”. The other peak is a consumer protection agency. In most cases, the central bank or a unit thereof is the prudential regulator.

Grant Spencer, deputy governor for financial stability at the Reserve Bank of New Zealand (RBNZ) between 2007 and 2017, tells Central Banking, “There is a saying: conduct regulation protects the customer from the bank, prudential regulation protects the bank from the customer.”

Andy Schmulow, an associate professor of law at the University of Wollongong, says prudential regulation and conduct regulation require different professional backgrounds: Prudential experts are often economists, while conduct regulators are often lawyers.

The main reason for turning to the twin peaks model is that the lines of demarcation between products, institutions and markets have become blurred

Ruth Plato-Shinar, Netanya Academic College

Schmulow is a strong advocate of the twin peaks model. He explains that sectoral and functional models became obsolete once financial enterprises began to operate in multiple sectors – when banks began offering insurance, for example. Sectoral and functional models fail because they are easy for institutions to game.

Since financial products are not material goods, “You can define that product or that service in as many different ways as you can think of to explain it and the number of different ways you can think of to explain it is infinite,” Schmulow says. “So, you then engender a state of regulatory arbitrage which is out of control.”

Plato-Shinar makes a similar point. She tells Central Banking: “The main reason for turning to the twin peaks model is that the lines of demarcation between products, institutions and markets have become blurred. So, keeping sectorial segregation becomes complicated.”

Dirk Schoenmaker, a banking and finance professor at Erasmus University Rotterdam and former official at the Netherlands ministry of finance, says Dutch regulators encountered this problem. “We had cross-sector products in the early 2000s with banking, securities and insurance features in the Netherlands. So, it was difficult to regulate them from a sectoral perspective,” he says.

On the other hand, Schmulow says, a single regulator cannot work because prudential regulators will tend to overrule conduct regulators. The prudential regulators can veto enforcing consumer protection rules for fear of alarming markets.

The problem with ignoring financial misconduct – aside from the injustices done to millions of consumers – is that scamming can itself produce a financial crisis. Referring to subprime lending, Schmulow explains: “Uh oh, turns out if you engage in misconduct in the market at a sufficient scale for long enough and you abuse enough consumers at a sufficient scale for long enough [it] turns out that can also be a financial crisis.”

Randy Priem, an adjunct professor of finance at UBI Business School in Brussels, makes a similar point: “Some key decisions to make banks more financially stable might not always be in the best interest of financial consumers, and vice versa.

“Two authorities can focus on other goals and are likely thus to have a different mindset, making sure that the complete picture of financial markets is taken into consideration when decisions are made.”

Schmulow adds that prudential and conduct regulators may also have different ideas about the structure of the financial sector. Prudential regulators tend to want a small number of very large, well-capitalised banks. Conduct regulators will prefer greater competition at the risk of more instability.

Others, however, believe there are advantages to housing prudential and conduct regulation under one roof. In her book on banking regulation in Israel, Plato-Shinar uses Schmulow’s argument – that consumer abuse can become a prudential crisis – to argue in favour of keeping prudential and conduct regulation in the same agency. Israel has a sectoral model, with the supervisor of banks having both responsibilities. The supervisor is an official of the Bank of Israel.

“The supervisor of banks, as the one in charge of both these areas, has the tools and the ability to weigh up all the considerations, and to make a decision that balances between them in the best possible way,” she writes. Plato-Shinar tells Central Banking: “In my eyes, the two fields are connected, complement each other and create synergy.”

A prophet unloved at first at home

Taylor says he was initially motivated to propose the twin peaks model by “a series of regulatory failures by the Bank of England – BCCI [Bank of Credit and Commerce International] and Barings, most prominently”.

However, the UK did not initially follow Taylor’s advice. After Labour took power under prime minister Tony Blair in 1997, it instead created the Financial Services Authority, a single regulator model.

Australia would instead pioneer the twin peaks set-up. Schmulow recounts that the Wallis Inquiry into financial regulation took an interest in Taylor’s paper and staff arranged to meet with him in London. The inquiry published its final report in 1997.

As a result, Australia established its twin peaks agencies in 1998. The country has a single prudential authority, the Australian Prudential Regulatory Authority (Apra) and a separate conduct regulator, the Australian Securities and Investments Commission (Asic). Both are separate from the Reserve Bank of Australia.

The twin peaks model found adherents in Europe. The Netherlands adopted the model in the 2002. The Netherlands Bank took charge of prudential affairs, and the Authority for Financial Markets handled market conduct and consumer protection. Belgium adopted the model in 2011, with the National Bank of Belgium having prudential authority and a separate Financial Services and Market Authority being responsible for conduct.

A 2017 law established twin peaks in South Africa. C Charles Okehalam, a professor in the school of public policy at the London School of Economics, says the reform was partly a result of the 2008 financial crisis, when “there was a sharp increase in the level of international attention on the need to further buttress financial stability”.

“This led to discussions on ways to separate prudential regulation and economic regulation in the financial services sector,” Okehalam says.

In New Zealand, the twin peaks model did not emerge from a specific policy commitment or plan. According to the RBNZ’s Spencer, the “model was not planned but rather evolved through time”. During the market deregulation in the late 20^th century, securities and equities markets emerged separately from banks, and the state established separate regulators.

One jurisdiction that has not embraced twin peaks is the United States, which has multiple federal regulators for both prudential and consumer protection affairs. Taylor notes that Hank Paulson, the US treasury secretary (2006–09) when the global financial crisis erupted, did propose reforms along twin peaks lines. Paulson’s “blueprint”, published in 2008, proposed making the Federal Reserve responsible for macro-prudential affairs (financial stability), a Prudential Financial Regulatory Agency and a Conduct of Business Regulatory Agency.

The Treasury report criticised the sectoral model, contending that it created “duplication”, turf battles between regulators and delays. It also meant that no agency could take a system-wide view of the financial sector.

“However, the rapid evolution of the [global financial crisis] meant that the focus of the legislative response shifted away from institutional structure towards the range of issues which were ultimately incorporated in the Dodd-Frank Act,” Taylor says.

Fall of the FSA

During the 2007–08 banking crisis, the Bank of England struggled to respond appropriately because it had poor communications with the FSA, the micro-prudential regulator.

Sylvester Eijffinger, emeritus professor of financial economics at Tilburg University and visiting professor at Harvard University, traces the FSA’s failure to its separation from the Bank of England. It may be the case that prudential and conduct regulation should be separate, but Eijffinger stresses that prudential regulation cannot be separated from the lender-of-last-resort function.

What I observe is what seems like a quite incredible amount of regulation related to the banking system – we have not just ‘twin peaks’ but a veritable mountain range of peaks

Don Brash, formerly of the Reserve Bank of New Zealand

“The central bank can only be the lender of last resort,” he says. To do that effectively, the central bank must also be well informed about the health of individual banks. The separation between the Bank of England and the FSA meant “it was very hard for them [the Bank of England] to have the full information about the problems with [Northern Rock] and Royal Bank of Scotland”, Eijffinger said.

The coalition government that took power in 2010 under prime minister David Cameron agreed. It abolished the FSA and adopted something like a twin peaks model. The Financial Conduct Authority (FCA) became a consumer protection and standards-policing agency, and the Prudential Regulation Authority (PRA), a Bank of England subsidiary, became the financial regulator. The BoE later absorbed the PRA entirely.

Anywhere between one-and-a-half and a half-dozen peaks

Some believe the term ‘twin peaks’ implies two “separate, but equal” prudential and conduct regulators, as Schmulow says. In most cases, the prudential regulator will either be the central bank or a subsidiary agency under the central bank. However, the tectonic plates of public policy rarely produce two identical mountains. The consumer protection authority is often the much smaller of the two agencies. Eijffinger notes that the Netherlands Bank has four times as many personnel as the Dutch financial markets agency.

The twin peaks model also assumes all financial regulation fits under the categories of “prudential” or “consumer protection”. Don Brash, a former governor of the RBNZ, tells Central Banking that even if the central bank is the formal overseer of banking, there are other government agencies that play a role in monitoring financial services, including the police. “What I observe is what seems like a quite incredible amount of regulation related to the banking system – we have not just ‘twin peaks’ but a veritable mountain range of peaks,” says Brash.

There is also the question of whether the macro-prudential function – a concept that was not commonly deployed in advanced economies in the 1990s – is a ‘third peak’. Eijffinger sees this as always being a central bank responsibility, and is tied to the lender-of-last-resort function.

Schmulow wrote in a 2016 blog post that Australia really has three peak regulators, because the central bank is the macro-prudential regulator but has no financial oversight authority.

In Central Banking’s Financial Stability Benchmark 2023, 83% of 30 responding central banks said they could put macro-prudential restrictions in place. The most popular tools were reserve requirements, counter-cyclical requirements and caps on loan-to-value ratios.

In the Financial Stability Benchmarks 2022, 52% of the 31 responding central banks said they had “sole responsibility” for bank regulation. Only 19% (of 31) were the sole insurance regulator, and only 19% (also of 31) had exclusive responsibility for non-bank financial institutions.

Too small for the both of us?

Setting up a twin peaks model may mean increased spending. The South African treasury estimated that setting up the new regulatory framework in that country would require a 42% increase in budgetary resources by 2020, an increase of more than 400 million rand ($21 million). In the UK, the introduction of the separate PRA and FCA led to a 15% increase in regulatory budgets between 2012/13 and 2013/14, or about £84 million ($107 million).

In a 2011 speech, Taylor argued the twin peaks model was suited to a certain kind of financial system. He said he had argued the model was right for the UK because the country had a large non-bank financial sector, and a system of consumer protection laws and regulations.

Not all countries, even wealthy ones, have well-developed frameworks for consumer protection. Plato-Shinar noted the Israeli supervisor of banks traditionally neglected this aspect of its mission.

Christopher Malcolm, deputy dean of graduate studies at the University of West Indies law faculty at the Mona, Jamaica campus, notes that Jamaica lacks an institutional history of rigorous consumer protection in the banking sector. The regulatory focus of the Bank of Jamaica has been prudential and not conduct of business in a significant manner, if at all, he says.

Small countries should not pile every financial regulatory function into the [central bank] in order to save costs. That is a recipe for disaster if it raises conflicts

Grant Spencer, formerly of the Reserve Bank of New Zealand

Malcolm adds that the proper protection of consumers requires close engagement with bank customers and the monitoring of their complaints and needs, something he does not believe has happened, so far, to the extent that should be required in Jamaica, he says.

Taylor tells Central Banking that small countries may lack the resources to sustain multiple regulators. “One alternative is to centralise all regulation within the central bank, which in many smaller economies (especially in emerging and frontier markets) is usually the best-funded government agency,” he says.

Spencer, the former RBNZ deputy governor, says his country “considered taking out prudential from the RBNZ when monetary policy independence was established in 1989”, creating a separate prudential agency like Australia.

“[There] was [also] an argument to keep prudential oversight in the [central bank], but certainly not to put securities/conduct regulation alongside prudential regulation,” Spencer adds. “Small countries should not pile every financial regulatory function into the [central bank] in order to save costs. That is a recipe for disaster if it raises conflicts.”

Schmulow says a country that cannot staff two regulatory agencies will not be able to adequately provide personnel for a single regulator, since conduct and prudential regulation are effectively different professions.

Does it work? Will it work?

Although Schmulow believes that twin peaks is the best model for financial regulation so far devised, he also thinks it has failed in Australia.

Schmulow says the prudential regulator, Apra, has allowed an “oligopoly” to consolidate in Australian banking, where the ‘big four’ banks have 90% market share. “In its neurotic attempt to stave off financial crises, [Apra] has so diligently favoured big banks and so diligently promoted big banks that we have very few small banks in Australia anymore.”

Schmulow is also critical of the conduct regulator, Asic, saying it has too many responsibilities to be effective.

He points to New Zealand and South Africa as having done a better job of constructing a twin peaks system. In the latter case, Schmulow believes this is partly because the South African financial sector is not a major political actor.

It is too soon to say how Jamaica’s twin peaks model will function, as it appears that the finance ministry is still preparing the necessary proposals.

The point about twin peaks, as with any regulatory structure, is that it won’t guarantee high-quality regulation. Getting the structure right is a necessary but not sufficient condition

Michael Taylor, Moody’s Investors Service

The University of West Indies’ Malcolm, who has had significant interest and involvement in the field of financial market regulation, is concerned about the lack of detail so far. He is also concerned Jamaican officials may have turned to, or hastened their intention to consider, twin peaks as a “knee-jerk” reaction to the SSL scandal.

Okehalam, the LSE professor, argues that institutional change will not neutralise the danger of novel regulatory challenges, like those from digital assets. “So viewed through those lenses, banking regulation and supervision can’t be fixed, it just has to ensure that it does not stay too far behind financial sector innovation. And where it can, it should try to go ahead,” he adds.

But Spencer, formerly of the RBNZ, says: “In my view twin peaks is the best model and I believe it has generally worked well in NZ and Australia.” However, he adds the model did sometimes result in a lack of conduct regulation over banks. Following the Hayne review into financial misconduct in Australia, which published its final report in 2019, Australian and New Zealand authorities have increased oversight of bank conduct, Spencer says.

Priem, at UBI in Brussels, maintains that the structure is insufficient without power and money: “If a regulator does not have a sufficient legal mandate to supervise certain financial actors, it cannot do much to reduce risks. Also, regulators should get a sufficient budget to have sufficient resources to do their work properly.” Priem believes “the twin peaks model is better”, given “a lot of additional conditions” that allow two truly effective bodies to function.

Although twin peaks arrangements can facilitate more nimble regulation, Taylor says that structural design alone cannot produce good outcomes. “The point about twin peaks, as with any regulatory structure, is that it won’t guarantee high-quality regulation. Getting the structure right is a necessary but not sufficient condition,” he says. Ultimately, you also need “a whole culture and environment, as well as strong institutional arrangements to underpin independence and accountability”.

Schoenmaker, the Erasmus University professor, agrees on the importance of independence. “Number one is an independent supervisor (independent from government and the regulated),” he says. This can militate towards giving the central bank a pivotal role: “In some countries with weaker structures the central bank is often the most independent institution.”

As Arnoud Boot, a former chair of the DNB’s bank council and a professor of corporate finance at the University of Amsterdam, says: “There is not one answer.” It is “just a matter of doing supervision well”.

This analysis article contains references to data provided by the Central Banking Institute’s Benchmarking Service. The Central Banking Institute is a members’ club open only to central banks and is part of Central Banking, the publisher of this article. If you have a query about the data or would like to take part in a future Benchmark, please email benchmarks@centralbanking.com.