Andrew Tarrant was a personal advisor to Gregg McClymont in the previous parliamentary term. A Kiwi lawyer he has the distinction of having sung in New Zealand’s only serious punk band.

I have had the privilege of reading this substantial piece of work several times and am proud that Andrew has asked to have it published on this blog.

For anyone who is seriously interested in pension outcomes in the UK over the next thirty years, this should be an important document

Introduction

Automatic enrolment is bringing millions of new workers into defined contribution (DC) pension schemes. It is more important than ever that those schemes can deliver good outcomes for members. Many of these members will be unengaged, having been enrolled onto their scheme by default rather than as a result of an active decision to save. It is essential that there is a clear regulatory environment for DC schemes which promotes high standards in pension schemes and good outcomes for members.

Even before the introduction of automatic enrolment, the workplace pensions market had significant weaknesses on the buyer side. In its report into the DC workplace pensions market, the Office of Fair Trading (OFT) identified the buyer side of this market as one of the weakest that it has analysed.[1] It also concluded that competition alone could not be relied on to deliver value for money.

The automatic enrolment programme addresses the problem of low participation. It places a duty on employers to set up a qualifying scheme and automatically enrol their workers into it. The system of soft compulsion, where members are brought into the scheme unless they object, harnesses the inertia which until now had been a barrier preventing people from saving for their retirement. However, automatic enrolment does not address the deeper weaknesses in the buyer side of the market where the ultimate beneficiaries of pensions are not active consumers.

The current DC market is divided between contract-based schemes and trust-based schemes, and people can be automatically enrolled into either of these. While both have strengths and weaknesses, the key distinction is the presence of a fiduciary duty in trust-based schemes giving the trustee a responsibility to look out for members’ financial interests. In a market where the product is complex, will not produce returns for decades and where unengaged members have started saving by default, the trust model ensures that there is an alignment of interests between the scheme and the member.

When it comes to automatic enrolment, employers may choose either a contract-based or trust-based arrangement for their pension scheme or schemes. For employers, trust-based schemes have the advantage of providing arms-length governance with a fiduciary duty. If an employer chooses a contract-based scheme, it is down to the employer to ensure that there is strong governance in place. While some employers take this responsibility seriously, many do not have the time, inclination or expertise to guarantee a fair deal for their employees. Indeed, in the run up to the introduction of auto-enrolment DWP research found that 59% of employers with a contract-based scheme size of between 12 and 99 members incorrectly said that their employees did not pay any charges at all.[2] This facilitated a market environment where large numbers of employees were subject to excessive charges (and in so far as their savings are held in legacy schemes, may continue to be overcharged).

The future of DC provision looks likely to be increasingly dominated by master trusts. These are large, multi-employer schemes with trust-based governance arrangements; they have a fiduciary duty and in many cases will reach scale. It is possible that single employer schemes will decline as defined benefit (DB) schemes become increasingly rare and workplace pensions become commonplace. However, there are risks associated with master trusts. There are virtually no qualificatory hurdles for entry into a market that is in direct competition with regulated financial service providers. There are also potential conflicts of interest where a single insurer appoints the trustees and provides services to the master trust. There is an urgent need for a process to authorise or license master trusts.

The current regulatory regime

The current DC landscape is split between two types of pension scheme; contract-based and trust-based schemes. Contract-based schemes are insurance products, the providers of which are regulated by the Financial Conduct Authority (FCA) and Prudential Regulatory Authority (PRA). Trust-based schemes are established as a trust and are regulated by the Pensions Regulator (TPR).

The Pension Regulator

On the whole, TPR has a focus on scheme quality and outcomes for members. Its recent work on DC, such as its Code of Practice for trust-based DC schemes, has emphasised good governance, value for money, transparent charges and robust administration.

Yet TPR lacks the resources to enforce adequate regulation. It has few powers in comparison to the FCA. It has a welcome emphasis on educating and enabling before it begins enforcement action. However, this is not backed with an ability to enforce quality criteria on pension schemes when action is truly required. Conversely, it has stronger enforcement powers to oblige employers in relation to automatic enrolment, including the power to issue penalties and recover unpaid contributions. So, bizarrely, employers can be obliged to contribute to a scheme, without there being any reciprocal requirement that the scheme must be up to scratch and provide value for money.

TPR’s limited financial resource is also a concern. Its total budget for 2014/15 is £75.1 million compared to £452 million for the FCA.[3] This means that TPR has a limited ability to monitor the schemes it regulates. Notably, its regulation of master trusts relies on independent assurance by the audit profession rather than direct monitoring and oversight by the regulator.

The Financial Conduct Authority (FCA)

By contrast, the FCA is a well-resourced regulator with substantial enforcement powers. However, it predominantly regulates retail financial products, such as general insurance, and therefore its approach tends to focus on disclosure and consumer choice. It emphasises treating customers fairly at the point of sale. Due to the weakness in the buyer side of the market, this approach is ill suited to workplace pensions. This has been entrenched by the introduction of automatic enrolment since the very notion of a point of sale is challenged in a system where people join a pension scheme by default.

It is important to note that the FCA’s “Treating the Customer Fairly” (“TCF”) principles and regulatory rules are significantly different from a trust-based regime. The FCA regime does not contain a requirement on pension providers to prioritise the best interests of scheme members[4]. An historical, and empirically-unjustified, conflation of the observation that there are a number of providers in the market with the concept of effective competition meant that the FSA/FCA never initiated action over excessive charges in workplace schemes. The FCA had a “faith-based” approach in the power of the disclosure of information as a solution to poor practice which was immune to the actual evidence of consumer outcomes. The FCA also allowed charging structures such as deferred member penalties which discriminated against people who change jobs and distort competition between market incumbents and new entrants. It has also not yet done nothing so far to remove the exit penalties charged by some old-style pension schemes.

Substantial risk in contract-based schemes arises for consumers because nobody is responsible for ensuring that the scheme prioritises the members’ interest. The FCA’s regulatory approach is ill-suited to the market because consumers cannot shop around actively for a workplace pension; they join the scheme chosen by their employer, and cannot normally change to a new provider. Even where they could do so, they run the risk of financial penalties, may have to expend significant time, and, risk losing contributions. The FCA has sought to meet the governance gap by requiring Independent Governance Committees for contract-based schemes. Gregg McClymont has pointed out in detail and at length precisely why these rules do not go far enough because IGC members will be appointed by and accountable to pension providers rather than members[5]. In short, they contain substantial and very obvious conflicts of interest.

Pensions are also a small and unglamorous part of the FCA portfolio. In the past year, the FCA has responded to political pressure and expanded what was a miniscule team devoted to workplace pensions. However, the FCA lack the specialised expertise available at TPR. In addition, except where Parliament has obliged the FCA to act, eg on the charge cap, the approach still remains one of trying to increase the information available to consumers rather than requiring that there are substantive quality criteria which apply to default schemes.

FCA officials have also expressed the view in the recent past that new competition law powers would allow it to tackle any problems in the market which it currently felt unable to tackle. This is a misplaced view. Absence the existence of a monopoly or a cartel, competition law is not able to deal with conflicts of interest in governance, failures to pursue economies of scale, failures to provide for effective information or excessive charging. Indeed, the weaknesses of competition law were highlighted by the outcome of the OFT inquiry into workplace pensions. The OFT conducted an excellent market analysis and identified the problems in the market. This was critical when pension providers and others were furiously denying that there were any serious problems of consumer detriment in the market. However, the OFT, despite the excellence of the analysis, were then constrained to accept a weak set of undertakings from the market participants in lieu of a reference – precisely because a reference would have left the Competition Commission, applying competition law, powerless to order any effective remedies.

Reforming the regulatory regime

In developing a blueprint for a new regulator, it is important to consider the wider impact on the economy and government spending. Government spending should not be increased by the creation of a single regulator for pensions. Any future regime should also not lead to a greater regulatory burden on employers; indeed, it would be preferable for a regulatory environment to reduce the burden on employers by providing a safe-harbour for those who make responsible choices on behalf of their employees.

Employers have a responsibility for automatically enrolling their employees into a pension scheme. Unless their scheme has effective governance, employers may also in practise have responsibility for looking after their members interests once enrolled since there may be no one else to take on this responsibility. Any future system of regulation should relieve this burden on employers by ensuring that all pension schemes are effectively governed. This could be achieved by guaranteeing safe harbour for employers who choose an automatic enrolment scheme that meets certain standards.

A single regulator should oversee a system of authorisation for pension schemes. Schemes with low charges, good governance and robust administration should be licensed by the regulator. Employers who choose an authorised scheme would not be expected to take responsibility for managing it since their scheme would have independently verified good governance. This reduces the burden on employers.

The pensions market is highly fragmented; there are around 38,700 DC trust-based schemes,[6] and over 100,000 contract-based schemes. Although there are signs that automatic enrolment has resulted in some consolidation, this has not gone far enough to promote the scale schemes which it is well known are best placed to deliver good outcomes for members.[7] A licensing regime will help to limit employers’ choice of schemes, and therefore drive scale, but the regulator also needs powers to close or merge schemes to create scale when this is in the members’ interest.

The regulator’s approach should focus on working closely with schemes, with officials visiting the largest schemes on a fairly regular basis. TPR currently has limited power to ban trustees, this should be extended to permit the removal of trustees where there is a repeated failure to pursue value for money. TPR should also have a bank of recommended trustees which can be parachuted into troubled schemes. It would also be useful for TPR’s powers of restitution to be expanded to cover DC. These can currently only be used in regard to DB schemes and employers’ automatic enrolment duties. The regulator should also have greater power to fine, including the power to issue recurrent fines where schemes consistently fall short of standards.

The regulator should be independent. It should be directly funded through a levy on its regulatory community. In order to ensure it is insulated from political pressures, it should also have some independence from government. The Department for Work and Pensions (DWP) should provide oversight of the regulator and its legislative framework.

Financing a single regulator

The Pensions Regulator is currently funded through a grant-in-aid from the DWP. These costs are then recuperated through a general levy, which is collected by TPR from occupational and personal pension schemes with two or more members. For the largest trust-based schemes, the levy is paid as an 86p per member fee. In 2014/15, the ‘levy budget’ (ie the grant-in-aid from the DWP) is £34.7 million. TPR receives an additional £40.4m from the DWP for its automatic enrolment duties.[8]

There is also a further levy to fund the Fraud Compensation Fund. This is not collected every year, and was last paid in 2012/13, at the rate of 25p per member.

By contrast, the FCA does not receive any public funding, though it is accountable to HM Treasury and Parliament. Instead, it is funded by direct fees payable by the organisations which it regulates. The FCA budget for 2014/15 is £452m.[9]

It is not clear what proportion of the FCA’s resource is spent on regulating workplace pensions, though even a small amount transferred to TPR could result in a considerable increase in the latter’s budget, and therefore also place pressure on the DWP’s departmental expenditure limits.

It is important that any solution is at least cost neutral; not adding to the Government’s total managed expenditure. Therefore, a single workplace pensions’ regulator should be financed in the same way that the FCA is currently financed. Its costs would be met by a levy payable by its regulatory community directly to the regulator.

The structuring of this levy is a crucial factor. The current structure of TPR’s general levy is unfair, since it is charged at a flat rate per member. This means that schemes face especially high costs for serving low-to-moderate earners or transient workers, whose pots are likely to be smaller than average. This part of the workforce has had historically low access to pensions, indeed; the National Employment Savings Trust (NEST) was created with a public service obligation precisely to address this market failure. One of the main driving forces behind the advent of automatic enrolment was that the industry had found it hard to provide pensions to people with small amounts of savings. Yet the current levy structure creates additional expenses for NEST and others who serve these workers.

The levy could easily be restructured to adopt a more progressive structure, with some proportionality in regard to members’ savings. This would be most logically and simply achieved through the introduction of a de minimis; where very small pots (below £500) are exempt from the fee. This would be broadly commensurate with TPR’s risk-based approach, since very small pots carry less risk. The rate of the levy would need to be recalculated to ensure that the same amount is raised overall. The regulator could also consider alternative forms of revenue, such as a registration fee for schemes gaining authorisation.

Decumulation

Decumulation – the process of converting a DC pot into a retirement income – is an integral part of automatic enrolment. Thus far, annuitisation has been the most common form of decumulation. This has historically been (very poorly) regulated by the FCA under the oversight of HM Treasury.

In the 2014 Budget, the Chancellor announced far-reaching reforms to the decumulation process. This amounts to a significant liberalisation where savers will have a wider range of choice at retirement, with more savers expected to choose income drawdown instead of purchasing an annuity. The Government has announced a guidance guarantee to supplement these reforms. There is no charge cap which applies to income drawdown products. Again, the regulatory approach consists of a reliance on information alone, when minimum quality requirements should be required for default products.

Given these changes, it is a sensible point to reconsider the regulatory regime for decumulation. In many cases, decumulation providers will come from within the DC market. Income drawdown may be administered by pension schemes. Similarly, annuity providers will continue to offer their products, though trustees may remain involved in their members’ retirements, for example through the purchase of bulk annuities.

At the same time, the recent reforms mean that a far wider range of products which traditionally fall outside of the retirement market may become available to members approaching retirement. Nonetheless, it is reasonable to suppose that since the vast bulk of those saving for a pension do not   currently make an active choice and save into the default fund, many will likely also default into either annuities or drawdown products provided by the same pension provider with whom they saved.

It is important that the regulatory framework fits around this new market dynamic. Whereas a wider set of financial products into which employees could place their savings on retirement provided by retail product providers and advisors should continue to be regulated by the FCA, workplace pension schemes’ activities should be regulated by the single pensions’ regulator. In the case of income drawdown, where the pot remains within the pension scheme, the pensions’ regulator should regulate this form of decumulation.

Scheme wind up

As millions of workers are brought into pension saving, and many save into large schemes, it is essential that those schemes are robust and that there is a solid process for managing the failure of large DC schemes.

There is a process for DC scheme wind up, though this is lengthy and complex. The key stages in a wind up are the recovery of any unpaid employer contributions, the identification of beneficiaries, establishing fund values and then arranging annuities for members. This process is largely untested in the case of master trusts. Presumably, if the trustee felt that it was in the members’ best interests, it would also be possible to bulk transfer members’ pots into a new scheme.

If members have lost money in a trust-based pension scheme as a result of theft or fraud, then they may be able to recover some money through the Fraud Compensation Fund, which is administered by the Pension Protection Fund.

In the case of master trusts, the trust itself often has a close relationship to a particular service provider, who would generally be the founder of the scheme. That may be an administrator, fund manager or pension adviser. Master trusts could be at risk if a major service provider were to fail and the trustees were unable to find an alternative.

In some cases, trustees may be bound by the terms of the trust deed to use particular providers. In these cases, the trustees would be powerless to protect members’ pots if that provider failed. It is important, therefore, that master trusts are sufficiently independent to change service providers, or change agreements with their service providers. It is important to note that failure could take several forms, including a situation where the quality of a provider’s service consistently deteriorates. It is important that trustees are able to take action, including finding a new service provider, in all cases of failure.

Central to this issue is the regulator’s approach. The regulator should operate a close supervisory regime and have a good working relationship with the largest schemes, so that it can identify risks before they cause failure. In some circumstances, it should be willing to parachute its own trustees into a scheme to ensure that any necessary action is taken.

The process for dealing with provider failure will vary between trust-based and contract-based schemes. In trust-based schemes, the assets are held in trust and will have some protection if a service provider fails. This is less clear in insurance-based products, though members may be able to claim compensation from the Financial Services Compensation Scheme (FSCS).

The general points around governance apply here, members of contract-based schemes that fail would currently have to make their own personal arrangements to secure compensation. In trust-based schemes, the trustee should take responsibility to ensuring members’ interests are upheld.

The key point is that scheme failure in DC is untested. There should be a clear process to protect members’ savings.

[1] Office of Fair Trading, Defined contribution workplace pension market study, September 2013 (revised February 2014)

[2] DWP, Pension landscape and charging: Quantitative and qualitative research with employers and pension providers, 2012

[3] TPR’s 2014/15 budget figure excludes a further £7.4 million set aside for IT infrastructure development.

[4] For a breakdown of the substantial legal differences in the degree to which the two regimes require that savers’ interests are put first, see http://www.fabians.org.uk/wp-content/uploads/2013/05/PensionsAtWorkThatWork.pdf, p.25-29.

[5] Please contact me at andy_tarrant@yahoo.com if you would like a copy of these submissions.

[6] The Pensions Regulator, DC Trust, 2014

[7] For example, The Pensions Regulator, Defined contribution trust-based pension scheme features, January 2013

[8] The Pensions Regulator, Corporate Plan 2014-2017, May 2014

[9] Financial Conduct Authority, Business Plan 2014/15, March 2014