APRA released a discussion paper considering capital requirements for life insurance companies on 13 May 2010. While much of the detail is still to come, the discussion paper sets out the structure of the future supervisory regime that APRA envisages. APRA’s stated aims are “to make its capital requirements more risk-sensitive and to improve the alignment of its capital standards across the industries it regulates”. Much of what APRA has presented represents a logical next step for capital requirements of Australian life insurers, particularly in light of international developments for insurance and banking oversight and regulation. However, there may be some unintended consequences and some aspects that are potentially at odds with APRA’s stated aims.
This article focuses on what may be seen as the more controversial aspects of the discussion paper, highlighting the potential implications of the proposals.
Summary of changes
APRA’s discussion paper proposes a single tier of regulatory capital to replace the existing two-tier solvency and capital adequacy requirements, with a regulatory focus on a comparison of an insurer’s required capital (termed the Prudential Capital Requirement) against its eligible capital base. Figure 1 illustrates the building blocks of the Prudential Capital Requirement.
Figure 1: Building blocks of the Prudential Capital Requirement
The Prudential Capital Requirement is intended to be consistent with a 99.5% probability over a one-year period of a company being able “to absorb unexpected shocks that may arise... and continue to be able to meet its obligations to policyholders at the end of that period”.
The eligible capital base will be determined as the excess of an insurer’s admissible assets over its liabilities. For the purposes of determining eligible capital, the life insurance liabilities will be measured at the greater of the Best Estimate Liability (“BEL”) and the Current Termination Value (“CTV”) at a product group level. The life insurance liabilities will also include funds that may only be distributed to particular policyholder groups, such as participating policyholder retained profits.
As is already the case for the banking industry, and to a lesser extent general insurance, a three pillar approach to capital regulation is being proposed, as summarised in Figure 2.
The key changes expected to arise out of APRA’s review of the life insurance capital standards are summarised in Table 1
Figure 2: Three pillars of capital regulation
First step of the process
APRA’s publication of the discussion paper is just the first step in the review of the capital standards. As outlined in the discussion paper, detailed technical specifications are due to be published in June 2010 and a Quantitative Impact Study (“QIS”) conducted later in 2010. Draft standards are expected late in 2010, with final standards expected in mid 2011 for implementation early in 2012. The process outlined by APRA will provide interested parties with a number of opportunities to comment on the proposals, including the opportunity to comment on the discussion paper (by 12 August 2010).
There are a number of areas where the discussion paper lacks detail, but these should become clearer as the technical papers and details of the QIS become available. A few examples of aspects that we hope will be clarified as more details come to light include the following:
Under the proposals, required capital is intended to be consistent with a 99.5% probability over a one-year period of a company being able “to absorb unexpected shocks that may arise over the one-year period and continue to be able to meet its obligations to policyholders at the end of that period”. However, the discussion paper does not go on to describe what is meant by meeting “obligations to policyholders”. For example, is this being able to only just meet benefit payments as they become due, or (similar to the existing standards) that the company can continue to satisfy the ongoing capital requirement over and above its benefit payment obligations, or some other meaning? A clear definition of the meaning of this term will be critical when assessing other aspects of APRA’s proposals, such as the liability measure to be used in the determination of the eligible capital base.
What definition of termination value is to be used as a minimum in determining the asset risk and liability risk capital charges, and what level of grouping is to be applied in testing for this minimum value?
Will it be possible to allow for management discretions in determining the asset risk capital charge for non-participating business?
Will asset concentration limits on amounts ceded by specialist reinsurers to overseas related parties be applied on an entity-by-entity or a group-wide basis?
What are APRA’s intentions with regard to supervising the industry based on a single tier of regulatory capital? For example, will target surplus need to be considered relative to avoiding a breach of the Prudential Capital Requirement, or relative to breaching a multiple (over 100%) of the Prudential Capital Requirement?
The process that APRA expects to follow, and in particular the opportunity to participate in the QIS, will no doubt give companies the opportunity to clarify such matters which, while possibly seeming trivial, could have a significant impact on the level of capital that will be required.
The bigger issues
While there is still more to follow, APRA has signalled very clear intentions for the direction it is taking, and has provided detail of a number of changes that may have a significant impact on the capital requirements for the industry as a whole and others that may have significant consequences for particular companies. We comment below on a number of the more important issues, in the interests of generating debate and discussion on these aspects of APRA’s proposals.
Risk-free discount rate
The discussion paper proposes that the risk-free discount rate generally be determined based on the yield of national government securities. Our main concerns with the use of yields of national government securities, and in particular Australian government securities, is the potential for market distortions in times of economic stress.
In turbulent times, with increasing or volatile credit spreads, companies could be expected to look to de-risk their investment holdings, seeking to increase holdings in the risk-free asset class. At a time when credit spreads are deteriorating, the increased demand for the risk-free assets will push yields on the risk-free assets lower, leading to even higher credit spreads. The result is procyclical pressure on credit spreads. While the government could increase supply in such circumstances, it is arguable whether or not supply could be increased sufficiently to substantially negate procyclical pressure on credit spreads, keeping in mind there may be increasing demand from the banking and general insurance industries at the same time. While such effects will always be present if a benchmark asset is mandated (arguably the use of a swap benchmark for certain purposes in Europe may have contributed to swap rates at longer durations falling below government bond rates), it will be of particular relevance in Australia given the relatively limited supply of Commonwealth Government Securities (“CGSs”).
There are a number of other drawbacks of requiring use of the yields based on Australian government securities. For example, it will not be possible to hedge interest rate risks effectively due to the limited series of CGSs on issue and limited derivative contracts. This implies that other instruments (such as swaps) will still be used, but movements in the mandated risk-free yields may differ substantially from those associated with the instruments used to hedge interest rate movements. Further, apart from the limited range of instruments, the maximum terms available are short relative to the term of many life insurance liabilities. The fact that CGSs are only available with relatively short terms, relative to other potential candidates to be used as a basis for risk-free rates such as bank bill swap reference rates, will require more extensive use of extrapolation, and lead to potentially large divergences in the actual rates used amongst insurers.
We would encourage more debate of this important topic, considering not just the benefits or otherwise of the proposal put forward in the discussion paper, but also the relative advantages and disadvantages of alternative starting points such as swap rates, yields on semi-government instruments and yields on minimum cost replicating portfolios.
Liquidity premium
APRA’s discussion paper indicates that a liquidity premium adjustment to the risk-free rate is being considered for lifetime annuities with no provision for voluntary termination, subject to APRA identifying “a robust method for quantification of the liquidity premium”. Rather than limiting the possibility of applying a liquidity premium to the liability of a particular product type, we would suggest that a liquidity premium be available for any liability that meets a “certainty” requirement. Depending upon the definition of certainty, this would potentially extend the applicability of the liquidity premium to deferred annuities without voluntary termination provisions and to disability income claims in course of payment, amongst other possibilities.
Risk margins
To ensure the capital base is not overstated, adjustments are made to statutory liabilities. For life insurance, APRA’s proposal is that the liability for the purpose of determining the capital base be the greater of the BEL and the CTV, at a product group level. The discussion paper notes that “for many types of life insurance business, the termination value minimum brings the level of liabilities to a higher level than best estimate, thereby incorporating an implicit risk margin.” While this is certainly true, any implicit risk margin is not going to be truly reflective of the underlying risk.
It seems to us that the liability measure used to determine the capital base should be consistent with the requirement to continue to meet obligations to policyholders which underpins the definition of required capital. As discussed above, APRA has not yet defined what is meant by “obligations to policyholders” for this purpose. If the intention is to just meet the expected obligation to policyholders, an argument could be made for there being no requirement for a risk margin in the liability measurement. However, a risk margin would be appropriate to the extent the bar for meeting obligations to policyholders is higher than this.
Aggregation benefits
The proposed methodology for determining required capital includes allowance for aggregation benefits between the asset risk charge and the insurance risk charge. No aggregation benefits are to be allowed in respect of the asset concentration risk charge or the operational risk charge. Not allowing for any aggregation benefit between two areas of risk is equivalent to considering that the risks are perfectly correlated, and that if there is an extreme result in respect of one risk there will be an extreme result in respect of the other.
There is no obvious reason why asset concentration risk and operational risk should be assumed to be perfectly correlated with asset risk and insurance risk. The discussion paper states that “operational risk is related to both asset risk and insurance risk and these correlations become stronger in times of extreme stress.” While we agree with this statement, there will also be an element of independence between the risks. The most obvious recent examples of operational issues include a number of unit pricing issues. Such issues have arisen and will arise quite independently of any asset or insurance risk stress. There may be an element of correlation in the extremes (with extreme movements in markets highlighting issues that might fly under the radar in normal conditions), however, we would expect a reasonable portion of the allowance for operational risk to be in respect of more frequent events, with moderate or lower cost, that are not highly correlated with asset or insurance risk.
In relation to the asset and insurance risk concentration charges, the discussion paper notes these “are designed to address concentration of assets and liabilities and can be regarded as independent of the other risks.” The discussion paper does not discuss why zero aggregation benefit is proposed even though the risks can be regarded as independent of the other risks.
Use of participating policyholders’ retained profits
The current Capital Adequacy Standard permits participating policyholders’ retained profits (“PRP”) to be used to support the capital requirements of non-participating business to the extent that it is surplus to meeting policyholders’ reasonable benefit expectations and provided that the support is based on commercial terms. The discussion paper proposes that it will be no longer be possible to assume such support for capital purposes (aside from overseas PRP that APRA has approved to be transferred to the shareholders’ fund).
The management of PRP is an important task for a number of insurance companies. One consideration is how to best invest the assets backing the PRP taking into consideration the interests of participating policyholders.
APRA’s proposal to prohibit the utilisation of excess PRP to support the capital requirements of the non-participating business eliminates a currently available option to enhance returns to the participating fund.
One of APRA’s concerns expressed in the discussion paper is that if PRP is made available to support non-participating capital requirements, on wind-up any losses relating to non-participating business may end up being charged to the participating fund rather than shareholders.
If one were to extend APRA’s case then one could conclude that investments of the participating fund should be restricted to risk-free investments. With any risky investment, such as corporate debt, there is the potential that the amount invested will be lost if there is a default on the investment. If this should occur, the effect is that the losses would be charged to the participating fund.
Provided the capital support for non-participating business is on commercial terms, then whether or not to enter into such an arrangement is effectively an investment decision for the participating fund. Hence, we would question the need to change the current requirements to prevent this practice.
Conclusion
In many aspects, the proposals that APRA has set out in the discussion paper represent a logical next step in the development capital standards for the Australian life insurance industry, addressing a number of gaps in the existing standards and achieving a closer alignment with approaches used in banking and general insurance. The result will be information that is more readily comparable to the equivalent requirements for the banking and general insurance industries, and standards that are moving in a similar direction to international developments. As discussed above, we consider that there are a number of areas where further discussion and debate would be beneficial in shaping the development of the capital requirements. It will be important for the Australian life insurance industry to get actively involved in this process, in the interests of achieving the most sensible and robust outcome possible.
Table 1: Key changes expected to result from APRA’s review of capital standards
Change
Probable impact1/comments
Focus on eligible capital versus single measure of required capital rather than total assets versus solvency and capital adequacy requirements
Unclear if this will have an impact in itself
Attention to be given to quality of capital
Neutral in relation to level of capital, but may influence the types of capital utilised
Three pillar regulatory approach to be followed:
Three pillar regulatory approach to be followed:
- Quantitative requirement
Impact will depend on relative impact of the component changes
- Regulatory review and adjustment
Negative impact if a supervisory adjustment is applied
- Disclosure
No impact on capital requirements
Risk-free discount rate generally to be government bond rate
Negative impact, as typically lower than mid-swap rates. Impact will be limited where CTV is greater than the BEL determined at the risk-free rate
- Liquidity premium adjustment may be possible for non-commutable lifetime annuities
Positive impact where applicable (although net effect of risk-free rate and liquidity allowance still likely to be negative)
Asset concentration risk charge to be separated from inadmissible assets
Neutral impact
- Intangible assets associated with all subsidiaries to be fully inadmissible
Negative impact, as captures all subsidiaries, not just financial service subsidiaries
- Other inadmissible assets to be along lines of current Solvency Standard
Negative impact relative to the Capital Adequacy Standard
- Admissibility of amounts ceded by specialist reinsurers to overseas related entities to be reduced (specialist reinsurer to be redefined as a statutory fund with no directly written business)
Negative impact (impact may be limited if limit is applied on an entity-by-entity basis rather than a group-wide basis)
- Higher admissibility for certain mortgages to be eliminated
Negative impact
- Higher admissibility to be available considering third party guarantees and collateralisation arrangements
Positive impact
Asset risk charge to separately consider inflation and real interest risk and volatility risk
Negative impact (likely)
- Impact of each risk to be separately quantified and the results aggregated using a correlation matrix to be specified by APRA
Expect negative impact as current basis assumes risks are independent, but overall impact is unclear as there is also a fundamental change in how test is to be applied (moving to a weighting of shock outcomes rather than a weighting of shocks)
- Asset risk charge is to consider all admissible assets (total assets less inadmissible assets and the asset concentration risk charge), not just assets necessary to support the required capital
Negative impact, but this should be offset by an equivalent reduction in target surplus (this is effectively an optical rather than substantive change)
Insurance risk to be separately identified
Neutral impact
- Detailed analysis required where risk margins are selected by the Appointed Actuary
Impact likely to be variable
- To include an allowance for pandemics and other risk concentrations
Negative impact
- Specific limitations are expected to apply to the application of management discretions
Negative impact
Allowance is to be made for operational risk
Negative impact as this is to allow for all operational risk, not just risk associated with investment linked business
Allowance will be made for partial independence of asset and insurance risks
Neutral for insurers with diversified risk exposures (but this will depend on how the new underlying parameters compare with current requirements); negative impact expected for insurers with undiversified risk exposures
- No allowance for partial independence of asset concentration or operational risk
Neutral impact
All participating policyholder retained profits are to be included as part of Prescribed Capital Amount and hence no longer available to support capital requirements of non-participating business
Negative impact for some providers
There will be no expense or new business reserves
Positive impact, but if support for new business is needed a supervisory adjustment may be imposed
Each entity to submit an annual Insurer’s Capital Adequacy Assessment Process report, which is to include capital projections covering a minimum of three years
No impact on capital requirements
1 We have indicated a positive impact where, in the context of the proposed approach, a change could be expected to either increase the eligible capital base or reduce the required capital, and conversely a negative impact if a change could be expected to either reduce the eligible capital base or increase the required capital.
Copyright 2012 Towers Watson. All rights reserved.