[Slide 1]

Regulators generally wish to avoid exciting times; where financial soundness is concerned excitement is less than welcome.

This is an exciting time for superannuation regulation in an entirely positive sense, however.  The Governments announced reforms to the prudential rules for superannuation will give APRA the tools necessary to improve industry safety.  At the same time APRAs maturation as an integrated regulator is beginning to produce risk analysis and response benefits across our four industry sectors, including superannuation.

[Slide 2]

APRAs responsibility in banking and insurance is more clear cut than in superannuation.  For authorised deposit takers, our job is to ensure that deposits are repaid in all reasonable circumstances.  Similarly, for life and general insurance companies our role is to ensure that valid claims can be paid.

The financial promise in superannuation, and particularly in accumulation funds, is far less straightforward.  The promise is not a minimum payment amount, or a stated rate of return.  Rather, the fundamental promise made under the superannuation industry supervision act is that trustees will do their best to maximise fund member interests, ahead of all other interests.

In addition to this central promise, under the law superannuation funds must observe a great many operating restrictions, for example limits on leverage, use of derivatives, in-house assets, and the like.  APRA certainly strives to ensure that super funds meet these specific requirements, but our overall focus is to ensure that trustees work first and foremost for the members.

[Slide 3]

What does APRA mean when it requires trustees to work for the member interest?  It is perhaps easier to start with what we dont mean.  APRA does not employ 20/20 hindsight when considering trustee performance, nor do we seek to direct trustees towards specific investment or operating decisions in the normal course of events.

We do require trustees to put the member interest first by acting honestly, diligently, and effectively in their duties.  diligence in this context includes not only sufficient time in trustee meetings to make the necessary decisions, but sufficient time and effort applied to trustee self-education, so that any decisions are made on a well informed basis.

Effective trustees recognise the expertise required in taking decisions that affect members, and either possess these skills or hire outside experts and monitor them closely.

[Slide 4]

Many trustees and other observers ask APRA how we respond when trustees make mistakes.  APRA accepts that even honest and diligent trustees can sometimes take decisions that look unattractive in hindsight.   APRA is advantaged in looking across the entire industry and the broader financial market place so we are reasonably well placed to understand emerging risks and appropriate mitigants.  As long as these decisions were taken in good faith and on a proper analysis, APRAs general response is to ensure that the trustees of the fund, and where relevant the industry more generally, learns from the negative outcome. 

To take a topical example, in 2002 and possibly in 2003 negative returns in many investment markets have led to poor superannuation returns at an industry level.  APRA does not regard these poor returns, at either the industry level or the individual fund level, as an indication that trustees are not doing their jobs well.

When mistakes are made due to ignorance or lack of diligence, even when no dishonesty is involved, we are considerably less understanding.  In such circumstances APRA will consider enforcement action including trustee replacement, and individual disqualification actions against trustees.

Losses arising from actions deliberately not taken in the members best interest are not due to mistakes:  they result from crimes.  In these circumstances APRA will certainly take robust enforcement action.  Where relevant this action includes referral to ASIC and the DPP for consideration of civil and criminal remedies.  APRAs experience has been that enforcement action at this level, as a proportion of individual superannuation trustees, is small.  Because there are many thousands of funds and over 10,000 individual trustees, however, the absolute level of enforcement in superannuation is likely to be substantial.

[Slide 5]

It is also worth clarifying the APRA definition of losses.  Losses do not include reductions in fund value due to price movements, where the underlying investment strategy and execution of the strategy are reasonable.  Losses also do not include properly disclosed fees and expenses; this area is typically an ASIC issue.

Regulatory losses most obviously include fraud and theft, which will always be an issue where several hundred billion dollars are concerned.  APRA is comfortable that the superannuation industry is well aware of fraud risk, and fraud losses as a proportion of assets are quite small.

Non arms-length dealing is a recurring theme in APRAs regulatory work.  We see trustees, for example, purchasing real property investments from relatives and from other sources where the transaction is less than fully robust from a fund member perspective.  We have also seen non arms-length dealing in third party contracts.  In this area losses are not only direct money costs, but loss generated by imposing undue and typically undisclosed risks upon a superannuation fund.  In some cases we have observed, and ultimately acted against, trustees who were not themselves dishonest but who were unable or unwilling to act when they observed inappropriate behaviour.

Finally, APRA observes more often than we would like that some trustees do not rigorously observe the need to put fund member interests first.  Instead, the employer sponsor, union, or retail sponsors interests are considered first or considered jointly. 

[Slide 6]

This slide outlines in brief APRAs prudential experience in superannuation.  Material fraud losses are very rare, but can be very severe.  Improper dealing is somewhat less rare, and generally manifests in higher than justifiable expenses and lower than reasonably predictable returns.  Undue risk exposure has been rare, but can materially reduce a funds assets.

Unfortunately APRA sees cases where trustees fail to put members interests first.  This does not mean that they altogether abandon the members interest, but that they are subordinated to another constituency.  The loss impact from this erroneous priority order ranges from the mild to the severe.

In trend terms, APRA sees no sign that prudential risks are increasing.  There are economic reasons to think that employer priority risk might increase in tougher economic times, but we see no evidence in our supervisory reviews that this is the case in fact.  We dont have a good handle on the trend in risk imposition generally, and the potential for priority dysfunction in retail funds; in both areas we are currently undertaking more work.

[Slide 7]

One learning from the HIH experience is that APRA cannot afford to be less than fully informed about risk issues at the industry and individual entity level, particularly for large and complex regulated entities.  We are taking a more proactive stance on risk analysis, and where appropriate a more proactive stance on enforcement actions designed to reduce prudential risk to superannuation fund members.

This risk analysis is to the extent possible data driven, a topic I will cover in more detail shortly.  In superannuation the key data item is the degree to which APRA can be confident that the trustees are honest and competent, and put fund member interests first.

[Slide 8]

APRA is currently in consultation on new statistical returns from the superannuation industry, to take effect from quarters and fiscal years ending 30 June 2004.  As shown in the chart, the number of maximum data items will increase substantially.  Not every super fund trustee will need to fill out all these items.  The chart compares the current position for authorised deposit takers and general insurance companies, where we see that the current data intensity is far greater than even the new superannuation forms.

APRA collects superannuation data for the Australian bureau of statistics on larger funds.  On smaller funds, we are trying to balance the need for extensive data with an aversion for creating cost and complication for funds and their trustees.

The new superannuation returns will for the first time give us proper asset allocations at the fund level, comprehensive financial statements, and good expense data.  All this material will usefully feed APRAs statistical early warning systems, which look for risk signals in the over 3,000 funds we directly supervise.

The new superannuation information we intend to collect is in nearly all cases information that we would expect the trustees to hold, or wish to hold, for their own purposes.  Indeed, much of this information is already reported to performance monitoring firms or other private sector groups.  The cost to providing this material to APRA should be, we trust, marginal in most cases.

[Slide 9]

It is clearly in APRAs interest and the fund member interest for us to understand superannuation industry investment performance.  APRA recently released a report undertaken by our researchers on returns, volatility, and estimated expenses in superannuation.  This report confirmed some of our qualitative observations, among which the most important are that other things being equal larger funds outperform smaller funds, and total expense ratios in superannuation are not decreasing over time.

This research also generated some new areas for investigation.  The difference between industry sectors is considerably greater than we had expected.  Most notably, retail funds outperform other funds on volatility, and underperform on net return.  The return underperformance is considerably greater than can be explained by differences in estimated expense ratios.

[Slide 10]

Retail underperformance on returns is a consistent and long running feature in superannuation.  Understanding and explaining this outcome is one of the key issues for APRAs ongoing research into the superannuation industry.  Other key issues include examining small APRA funds and possibly self managed funds in conjunction with the ATO, to determine their investment and risk characteristics.

We also hope to release, in the next few months, a study on expense drivers in superannuation.  This study should provide better insights into the relationship between industry expense levels and member dynamics, including average account balance, numbers of members, and other variables.

[Slide 11]

Moving on to licensing, the Government has announced major improvements to prudential supervision arrangements for superannuation.  APRA is looking forward to implementing these improvements to help create a safer superannuation sector.  All this work, however, is subject to legislation, so my comments today are necessarily somewhat speculative until parliament approves the eventual bill and associated regulations.

The key issues in the new regime are that every APRA regulated superannuation trustee will require a licence, and every super fund will require registration.  Every trustee will need a suitable risk management strategy, and every fund a risk management plan. 

Our current expectation is that there will be three licence tiers.  The unrestricted tier, known as tier one, is roughly equivalent to retail and larger non-retail approved trustee funds.  Tier two is focussed on multi-employer funds, and tier three on single employer funds.

Licensing expectations will vary somewhat by tier, with smaller funds subject to less demanding licensing arrangements and larger and more complicated funds subject to more demanding arrangements.  This position is not due to any APRA aversion to larger and more complicated funds.  Rather, we observe that equal representation provisions work best in funds that would fall into the third tier, work reasonably well in the second tier, and are generally not applicable in the first tier.  The balance between APRA licensing demands and equal representation protection is intended to lead to roughly equal regulatory safety for every Australian superannuation fund member, no matter the size or type of regulated fund that the member belongs to.

Choice of licence tier will rest with the trustee.  It is possible that APRA would approve an applicant for one tier but not for another, but this is not the expected outcome.  Generally we will assess applicants on the requirements associated with the tier they choose in their application.

APRA also notes that super funds are undertaking a lot of work to meet ASICs FSR licensing requirements.  To the extent possible APRA and ASIC will try to avoid making super funds and their trustees undertake the same effort twice.

[Slide 12]

The key themes in licensing will vary by tier.  For tier 3 entities APRA will focus on trustee competence.  Have the trustees on both sides of the table received sufficient initial and ongoing industry training to properly discharge their duties?  Practical issues such as time off and payment for training should feature in the risk management strategies for tier 3 entities.  In tier 3 we expect to see a lot of outsourcing, including to employer sponsors, and this outsourcing will need reasonably tight management.  Overall, APRAs expectations are that tier 3 licensees will have basic but solid business plans and risk management strategies.

Tier 2 licensees must also focus on trustee composition and training.  APRA encourages industry funds in particular to diversify their trustee membership away from relying wholly on HR directors on one side of the table, and union reps on the other.  A few more finance directors, operational division heads, or possibly external appointees nominated by the respective electorates.  Tier 2 entities will normally require reasonably good outsourcing management, and will enjoy a sophisticated investment management strategy in theory, carried out with appropriate controls in practice.

[Slide 13]

Tier 1 licensees will need to demonstrate that they can safely run a large and complicated business, with sophisticated investment management, operational risk, and other systems.  As a rough rule of thumb, a tier 1 superannuation license will carry the same risk management requirements as a license to operate an equivalently sized business in the deposit taking or insurance sectors.

In all tiers, APRA will not require every risk aspect to be covered by worlds best practice; this would impose undue cost for the marginal risk avoided.  We do however require that all material risks are covered to the extent reasonably expected by members, plus a little additional coverage to ensure a conservative approach.  Im sure this audience would expect no less, as we collectively look after Australias life savings.

[Slide 14]

APRA expects to devote 2003 to public consultation on the new rules, particularly the licensing rules.  In 2004 and 2005 we will move to the licensing process.  To avoid a late rush, we are currently considering how best to manage the licensing workload through the two year transition period.  APRA is currently scoping the large effort needed to manage the licensing project; we estimate that we need to create a thirty to forty person team to deal with the work involved.

It is too early to estimate the proportion of successful applicants, but given that I am often asked the question I will provide here a best guess.  This guess is based on APRAs experience with general insurance relicensing in 2002, and on our views of the superannuation industry today.

The great majority of licence applicants will receive a licence.  Among the approved applications, however, an unknown but substantial number will require amendment.  Some amendments are to the application itself, and should not present many concerns.  The more material effort will be APRAs desire to remediate issues with applicants who do not have sufficient risk management systems in place upon applying, but who could remedy the identified issues and therefore subsequently receive a license.

A small minority of applications will be refused.  Hopefully many marginal providers will recognise the wisdom of transferring their business to a stronger entity, in preference to pursuing an unlikely license approval.

At the end of the licensing process, every superannuation trustee will possess at least a reasonable risk management strategy, and every fund a reasonable risk management plan.  This will materially improve prudential outcomes in the superannuation industry.

[Slide 15]

APRA intends to provide considerable guidance in the licensing process.  We have for example drafted licence application outlines, possibly including pro forma risk management plans, and will doubtless provide more material as we move forward.

The risk management strategy and risk management plans, including the investment management plan, should constitute about 80% of the work in an application.  The business plan may also require work, and generally documenting processes and policies will take some time and effort where this work has not yet taken place.  I note that all this work will help fund safety even if no licensing process were impending, so it should not be viewed as regulatory drudgery, at least so APRA hopes, but a helpful creation or update of the risk management environment.

The most important criterion for licence approval is the trustees ability to convince APRA that it and they care just as much about the members as we do.  This is a high standard, as APRA sets great store by Australias super fund members right to receive the best possible service from their trustees, including robust and sensible risk management.

[Slide 16]

Let me comment briefly on liquidity.  APRA expects trustees to have good liquidity arrangements in place, based on the ability to forecast potential cash needs, and assets and other arrangements in place to meet these needs.

This does not mean that APRA believes cash is good, and more cash is better.  Cash is a low return and in some senses lazy asset, particularly for long term investment funds.  APRA will not penalise funds running low or even zero cash positions, as long as the ability to quickly liquify sufficient assets at short notice without a fire sale is sound.

Similarly, APRA has no expectation that all superannuation money should be invested in highly liquid assets such as bonds and large cap equity.  As a practical matter in a reasonably diversified portfolio most assets could be sold for a reasonable price in a few days to a few weeks, but all assets need not be in this category.

Liquidity is expensive, and excess liquidity is pointlessly expensive.  APRA will not penalise entities holding some proportion of illiquid assets as part of a sensible long term investment strategy, as long as these investments are an explicit and carefully considered part of the strategy.

APRA has had negative supervisory experience with funds that invested in illiquid assets, but not as part of a considered overall strategy to maximise fund member interests.  Investing the entire or nearly the entire fund in a single property, for example, is not consistent with liquidity or for that matter broader investment requirements.

Should we move to fund choice, APRAs experience with banks and other deposit takers indicates that fund liquidity requirements should not change much.  The key is to maintain public confidence, and trustees following controlled risk strategies that focus primarily on member benefit are well positioned to merit this confidence.

[Slide 17]

While reviewing issues, let me also touch on unit pricing.  This is a somewhat vexed issue, that has led in a few cases to substantial losses.  Due to unit pricings significance in calculating member entitlements, APRA will devote increased attention to this area.  APRA will shortly issue firmer guidelines on unit pricing, based on the current IFSA standards 8 and 9.  For now this will be a voluntary approach, but acceptable unit pricing practices will be a condition precedent to granting licences over 2004 and 2005.

APRA will also regularly review super fund unit pricing practices, as part of its ongoing supervision.

[Slide 18]

In prudential terms, the superannuation industry is headed in the right direction.  APRA is becoming a more experienced and knowledgeable regulator, as our institution matures.  The Governments announced reforms, once implemented, will materially improve the industrys position.  Most importantly, the trustees in superannuation continue to improve their ability to look first and foremost after the members interests.

[Slide 19]

The ideal relationship between APRA and trustees, as far as we are concerned, is that the trustees are honest and competent, and as a result there are no material prudential issues with the funds in question.  In such cases APRA will act to the extent necessary as a consultant and guide, and practice a minimum intervention strategy.

There will always be a few cases where the trustees are trying to do the right thing, but through inexperience or other reasons are unable to do so.  In these cases APRA will be a considerably firmer guide towards either improvement or transition out of the industry.

There will also always be a few cases, unfortunately, where the great mass of money in superannuation attracts people who do not put the fund members interest first.  With over 3,000 not for profit and over 100 public offer superannuation providers, there is no competitive reason to tolerate such people.  APRAs approach when we determine that a trustee or trustees, or other people associated with superannuation are less than honest is simple:  goodbye, but not good luck.

Superannuation management is not a matter of luck, but of competence and probity over the long haul.  APRA will strive to ensure that all responsible persons in the superannuation industry exhibit honesty and probity.  In these circumstances the industry will make its own luck as we continue to develop one of the worlds best superannuation systems.

Thank you.

Click here for the slide version