Don Mercer has covered our general aspirations on supervising the superannuation sector. I will direct my comments to several areas that are currently active in converting the Government’s policy announcement into reality. Let me first note that all my statements today are subject to amendment as the relevant legislation moves through Parliament. They are also subject to amendment as APRA consults publicly on the many issues associated with a new supervisory regime.
The key improvement in the new regime is universal licensing. APRA intends to relicense the current approved trustees, commence licensing all the other superannuation trustees and register all the superannuation funds for which we hold prudential responsibility.
We anticipate releasing our views on licensing requirements and processes by early 2003 with comments due back to APRA in March or April 2003. Trustee expressions of intent to be licensed will be due by 1 October 2003. From this point there will be a two-year period during which APRA will undertake the initial licensing exercise.
We are currently considering the pros and cons for creating more than one license tier based upon line of business. In particular, APRA may propose a basic tier for corporate funds with a single employer sponsor, where services are limited to relevant employees and retirees. Such a license may limit the scope for adventurous investments, and apply firmer controls to outsourcing. Let me note that APRA has no intent to use the licensing process to encourage or discourage any of the various sub-industries within superannuation. As a practical matter we expect that the shrinkage in corporate plans relative to total plans will continue, but this is neither a necessary nor intended outcome from licensing per se. It may, however, be an outcome from the increased risk management requirements attaching to licensing.
Another license tier might allow more flexibility, but would have correspondingly higher standards associated with the license. I expect that most ASFA members will apply for the highest tier superannuation license. As a general rule of thumb, a trustee license to offer superannuation services to the public will possess the same risk management expectations as a license to take deposits from the public, or to sell insurance to the public.
Among the many questions APRA will ask applicants in the licensing process, three are most important:
First, are all the responsible persons associated with the application fit and proper?
Second, is the Risk Management Plan sound, and likely to be carried out in practice?
Third, are all outsourcing arrangements sound?
A "no" to any of these questions will at best delay and at worst finish a license application.
On the fit and proper question, it is clearly important that trustees, key employees, and key contractors and servicers are honest and competent. Fitness and propriety standards help achieve this outcome. APRA will not seek to pre-vet every person or even every senior person associated with the superannuation industry. We are likely to develop sufficient information tools to catch obviously problematic cases, but trustees in particular will carry the main responsibility for policing fitness and propriety requirements.
A person may not be proper for a responsible position in the superannuation industry if he or she has been or is bankrupt, entangled in criminal, civil, or regulatory matters, associated with failed super funds or corporations, or (the general catchall) a disreputable person due to past business dealings. In addition, a person may be unfit if they have previously failed to perform competently in a senior role in the superannuation industry, or for that matter in a different industry.
Note that I say "may" be unfit. APRA is not looking for reasons to exclude people from the superannuation industry, unless the person in question has exhibited behaviour that materially reduces our confidence that the public’s life savings are safe in their hands. We are developing more explicit fitness and propriety standards, and procedures for assessing an individual’s fitness and propriety. APRA’s fitness and propriety assessments will be based on a person’s past and present business qualifications, track record, and expertise. We have no interest in assessing any person’s behaviour in their private life, absent any relevance to the question of fitness for access to superannuation funds.
A final point on fitness: APRA intends to take an encouraging approach to trustees in equal representation situations. We are prepared to tolerate a new trustee who lacks formal training in superannuation, but to become fit this person will need basic and continuing training in trusteeship. APRA will consult with industry and the public on the definition of "basic and continuing training"; the intent is to move super fund trusteeship from the potential realm of the well-meaning amateur to a more professional basis, without excluding equal representation models.
The Risk Management Plan or RMP is the core document in the license application. It will also be the core document in the ongoing supervision process. The fund trustees need to understand that the RMP is not simply a form to be filled out for licensing purposes. The document must be lived in fact; material adherence to the RMP will be an APRA requirement for ongoing licensing, and will be monitored by the external auditor.
The RMP is likely to start with the equivalent of a short business plan, outlining the fund objectives, the major risks to achieving these objectives, and the trustee’s key strategies in managing these risks in order to achieve the objectives.
The RMP will then move to investment management. APRA expects that every fund will possess an investment management plan that reflects modern professional practice. In particular, fund member time horizons and risk preferences need consideration, and an asset and liability model to manage these preferences should be created. This model need not contain zillions of Greek letters prepared by an external consultant, but it should produce a sensible investment strategy that is highly likely to achieve the fund objectives. The investment management plan should also be robust to changes in expectations, outlining the process by which trustees will review the asset liability model, and make changes to investments accordingly.
The investment management plan should not be careless in leaking value away from fund members. In particular, the intrinsic value attaching to voting equity should be captured, via direct voting policies or by instructions to fund managers. When using external fund managers, APRA will expect that the soft money issue will be considered and managed appropriately. Votes and soft dollars are negotiable assets that belong to the fund members, and should be treated accordingly when trustees set up their investment management plans.
APRA is not expecting a great deal of attention to individual security selection procedures in the investment management plan, particularly when fund management is outsourced. We do expect to see reasonable attention given to allocation among asset classes, among fund managers, and between active and passive investment strategies.
The RMP also needs to address liquidity management. This will become a larger issue if member choice is introduced.
Continuing with the RMP, APRA will look for considerable care and attention applied to anti-fraud measures and management of operational risk, including regulatory and tax compliance risk.
The third key licensing question concerns outsourcing. APRA is neutral as to whether a fund should outsource funds management, administration, and other functions. When outsourced, however, these functions need default management against the possibility that the outsource provider becomes unwilling or unable to provide the contracted service. APRA’s current ADI outsourcing standards are a good place to look if you would like a preview of our intended approach to superannuation outsourcing. One particular issue is APRA’s need to be able to meet with outsourcing providers, in order to assess their ability to provide appropriately low risk services to the superannuation industry. We have no desire to regulate these providers, but we may need to assess their suitability with some regularity.
In addition to the above areas, APRA will provide guidance on appropriate insurance for superannuation funds. The insurance industry has been through its own turmoil, and we recognise the fact that ideal insurance may not be on offer at an ideal price. APRA intends to take the most pragmatic approach possible in this area, consistent with fund member protection.
We have been asked whether we intend to make the licensing application a public document. This is not our intention, but we are open to the idea that the application or a synopsis thereof should be provided to fund members or potential fund members.
One natural question you may have, is in what order does APRA propose to license the industry? Our views are subject to change after consultation, but in general we plan:
First, to focus on public offer and larger funds, in order to quickly bring the bulk of Australia’s superannuation money into the new supervisory and licensing system;
Second, all new funds from 1 October 2003 will need to be licensed before they commence business,
Third, we may accelerate licensing, or in fact refusal of licenses, for problem funds or trustees identified through our ongoing supervision;
And finally, we plan to deal with our approximately 2,500 medium and smaller sized funds in sequence with their scheduled on-site inspections.
The licensing process is clearly a massive challenge for APRA and the industry. Again, subject to fund member protection we intend to take a pragmatic approach to licensing.
Once we license a fund or trustee, our ongoing issues are likely to surround honesty, competence, and agency risk.
Honesty and competence will be the focus of APRA’s work on fitness and propriety, and upon risk management, particularly investment management and outsourcing. In this sense ongoing supervision is simply an extension of the licensing process. Funds and trustees will need to regularly update their responsible person registers, observe and keep current the Risk Management Plan, and stay on top of outsourcing performance.
In addition to these areas, APRA is increasingly the sensible information source for agency risk assessment in superannuation. I have already touched upon voting and soft money, which are prudentially relevant because their non-management indicates less than full attention to fund member interests.
Other agency issues surround the risk that fund members may not receive the risk, return, and expense propositions that they could reasonably expect. This risk is particularly acute in accumulation funds, where a poor risk/return/expense proposition will materially reduce a fund member’s financial circumstances in retirement.
APRA has no intent or desire to regulate fees and expenses, absent the fortunately rare cases of excessive charging that we already deal with. We do intend, however, to use data gathered from the industry to help ASIC in their role as the conduct regulator. Over time we may also provide more data to the public, to help them deal with the information disadvantage most super fund members face when making retirement savings decisions.
Improved statistical collection will be one aspect of the new superannuation supervision regime. We expect that better super data will be available from the June 2004 year-end.
To give an example of the data already available in APRA, this slide shows some preliminary results from a research paper to be released shortly. This slide outlines the risk and return propositions typically experienced by larger superannuation funds, split by type. The ellipses represent a one standard deviation spread from median for each fund group.
The first conclusion from this slide is that financial theory works; we see what looks like the northwest slope of an efficient frontier curve. From a supervisory perspective, we also see why corporate funds may prove a bit more problematic; they earn higher returns for higher risks.
We also see why APRA, which has competition and efficiency responsibilities as well as safety objectives, isn’t trying to move people from corporate superannuation to public offer funds. The public offer funds are less volatile, but earn materially lower returns on average.
Somewhat surprisingly, from this slide we see that fund size doesn’t predict returns or volatility, where in the previous slide fund type is highly predictive. From a supervisory perspective, at least in the investment risk vector, larger funds are not safer than smaller funds. Remember here that the population is limited to funds over about $60m in assets. This work does not examine the small end of the market.
I present these slides not to draw any conclusions, but to demonstrate that APRA data is about to become more useful in guiding supervisory attention, and also to assessing agency risk.
In recent months APRA’s Board and executive management has commenced a conscious shift in APRA’s supervisory culture. In a nutshell, in past years we started from the assumption that the depositors, policy holders, and super fund members for large regulated entities in particular would benefit from minimal supervisory intervention. After HIH and some other negative though less financially damaging experiences, we no longer hold this view.
APRA now supervises on the basis of proactive scepticism. That is, we are prepared to accept that a regulated entity is managed by honest and competent people, and that our beneficiaries are safe as a result. This acceptance, however, must be built on the entity’s continuous demonstration of sound position and management. The burden of proof is on management, and in the superannuation industry specifically upon trustees, to convince APRA’s supervisors that the fund is acceptably low risk. In terms of our internal risk modelling, we expect that every superannuation fund should be able to demonstrate a high probability, and as guidance at least a 97% probability, that it will pay beneficiaries as they would reasonably expect. 97% probability translates in human perception terms into: there are some circumstances where a negative outcome is possible, but these circumstances are highly unlikely, and the fund has strategic flexibility in responding to negative surprises.
Furthermore, APRA is changing its culture towards a more analytically driven approach. We are relying more on our statistical collections. This is one reason why we are working so hard to lift the super industry’s compliance with regulatory returns. In the past two years timely completion for annual returns has increased from 17%, a number that speaks for itself, to a bit over 80%. I expect that within three years the figure will be at around 95%.
In addition to statistical returns, APRA has greatly lifted its on site visit program to superannuation funds, and the information gained from these visits is feeding into more robust risk models. We are also looking to improve information feeds from auditors, actuaries, and other sources including whistleblowers.
This change to a more analytic approach is consistent with APRA’s move to a more proactive and effective approach to early intervention when we observe a regulated entity straying away from its expected risk profile.
To summarise, APRA looks forward to seeing government policy converted into legislation, so that we can improve superannuation supervision. This will allow APRA to better control super fund losses due to less than adequate risk management. Over time it will also allow Australia to better manage agency risk in the superannuation area, through better information on fund performance.
As with all our major change programs, this process has been and will remain consultative. We will seek input from the public and the superannuation industry on our proposed regulations and licensing approach, and any other material matters of interest.
The bottom lines in reforming superannuation are twofold.
First, by 2005 amateur hour will be over in superannuation management. For most ASFA members amateur management has long ceased to be an issue, but for many small funds we expect to see a distinct improvement in the quality of risk management, either directly or via outsourcing.
Second, superannuation members have long been severely disadvantaged in their ability to assess and respond to the risks attaching to their retirement savings. This reform gives APRA the tools we need to reduce the risks facing fund members, and to help fund members understand how these risks are managed in a given fund.
My colleagues at APRA and I look forward to working with the superannuation industry as we reform superannuation supervision. In the long run this reform will make the industry safer and more rewarding for Australia’s super fund members.
Don and I are also looking forward to your questions and comments.