Recognising the fact that superannuation funds, particularly non-retail funds, generally do not have sufficient shareholder reserves. The Australian Prudential Regulation Authority (APRA) requires funds to set aside an amount to meet the costs of errors and other losses that may arise as a consequence of investment or operational activities.
Prudential Practice Guide SPG 114 specifies that “Funds should have an Operational Risk Financial Requirement (ORFR) target amount that is equivalent to at least 0.25 per cent of funds under management (FUM). For the purposes of calculating the ORFR target amount, APRA views FUM as the total of asset balances of each RSE within the RSE licensee’s business operations.”
On the face of it, this seems like a reasonable approach. By requiring funds to set aside a ‘rainy day’ fund for potential future losses, it means that they will likely have sufficient reserves available to meet the cost of operational errors without risking default on the fund’s obligations to members or external parties. However, there are some flaws in this logic.
One size fits all and it’s XXXL
By imposing a standard threshold across all funds, there is no recognition of the fact that each fund’s operational risk profile may be different, often driven by a different investment strategy and operational model. In reality, funds have very different risk profiles and, consequently, this should drive a different operational risk reserve. This approach is common amongst other financial institutions such as banks that basically ‘self-assess’ their reserve requirements based on a review of their operational risk profile. This tailored approach means that the reserving policy is reflective of the true operational risks of the fund.
In addition, setting aside at least 25bps to fund operational losses is likely to be unreasonably large, especially given the quantum of historical operational losses incurred. If we review the aggregate across the industry, the results are telling.
According to APRA, the total value of superannuation assets was $2,046 billion in December 2015. Assuming all Funds adopted the 25bps reserving policy, the amount set aside to ‘fund’ operational losses is a whopping $5 billion dollars! Assuming growth of 15% pa (10% contributions and 5% market growth) this could grow to in excess of $20 billion by 2025.
It would be difficult to foresee any circumstance where costs to fund operational losses would approach anywhere near this magnitude, especially when appropriate insurance is in place. Of course, the response to this is ‘so what’? Isn’t it better to have a conservative reserving policy, even if this is excessive? Surely this is a better outcome than not providing for sufficient reserves? Whilst having sufficient funds to meet operational costs is, of course, a key requirement, we believe there are significant downsides from excessive reserving.
Inefficient use of capital
To prevent gearing, it is generally necessary for funds to ‘cash back’ liabilities and provisions. That is, to retain amounts in low risk assets to offset the value of these. Whilst a sensible investment approach, this means that capital is parked in low return assets. If the value of ORFRs were to grow as indicated above, $20 billion of assets would be subject to this constraint. Given the challenge of meeting the retirement needs of an ageing population this does not appear to be a wise use of capital and also serves to reduce Australia’s competitiveness.
Managing Member Equity
As the costs of funding the ORFR are effectively borne by members, there is an argument that the treatment is inequitable, especially in the current period when the reserve is being established. Rather than the Trustee meeting the cost to fund the ORFR, the creation of the reserve is typically achieved through the transfer of investment earnings to the reserve. If, as could very likely be the case, an assessment is made that the reserve is excessive, these funds will be transferred back to members. However, a sizeable number of members will have left the fund in the intervening period and new members will have joined. This causes an inequity issue insofar that exiting members will have been disadvantaged to the betterment of recently joined members.
Risk of Complacency
One might argue that having a significant reserve to meet operational errors might drive the opposite effect to that intended i.e. Trustees may become complacent about operational risks on the basis that there is a sizeable ‘bucket’ to Fund losses. Of course, there are other measures, such as the APRA reporting regime to ensure Trustees are attentive to risks. However, as there are limited economic consequences to Trustees, one might argue that these are less coercive than losses hitting the hip pocket of the Trustees. Experience in the corporate environment demonstrates that risk of economic pain is typically the most effective way of ensuring appropriate focus on operational risks.
So what can be done?
As a minimum we believe that, like the reserving approach initiated for banks, the ORFR level should be determined based on an assessment of the fund’s specific operational risks. This will prevent the risks of excessive reserving and will maintain the most efficient use of capital.
However, more broadly, why doesn’t the industry consider an alternate model where losses are funded by the Trustee? This approach has been adopted successfully for the non-superannuation investment market for many years. Under this model, members would not bear the costs of excessive reserving management and Trustees would bear the full operational costs and responsibility for inadequate internal controls.
The Funding of the Trustee would not require the transfer of Assets to the Trustee from the Fund provided an approved (APRA) indemnity could be put in place, as is currently the case for most Trustees. The appropriate Liquidity stress testing of the Fund would then be extended to include the Trustee indemnity for the ORFR, but in the meantime the assets could be appropriately invested.
Further, the Industry should ask the Regulator (APRA) to start collecting the history of errors and incidents through the breach reporting process so that an industry wide pattern can be determined. This would then allow a true risk weighted percentage to be developed, rather than one based on the Insurance Industry being imposed on the Super Industry in the absence of a better data point.
Finally, the Regulator could approve an appropriate OPFR model for each Fund (there is plenty of actuarial skills in this country to arrive at appropriate Models), and providing the Fund can demonstrate a five-year track record of incidents, they could then be allowed to transition (say five years) to their own ORFR. That way by the time we get to 2025 we will not have $20 billion sitting in Reserves!