1 January 2013 saw the commencement of a suite of new prudential standards which significantly changed how capital adequacy is determined for general insurers, Level 2 insurance groups, life insurers and friendly societies (‘institutions’).
An important component of the new prudential standards is the requirement for institutions to have an Internal Capital Adequacy Assessment Process (ICAAP). This includes a requirement for institutions to prepare an ICAAP Summary Statement (Summary Statement) as well as an annual ICAAP Report (Report).
In 2013 APRA undertook a comprehensive review of Summary Statements. Feedback from this review was provided to institutions in a letter dated 9 December 2013.
During the first six months of 2014, APRA undertook a detailed review of ICAAP Reports submitted to APRA. The review included peer comparisons to identify examples of better practice and potential areas for improvement.
The attachment to this letter sets out the findings from this peer comparison which APRA consider warrant consideration by boards. Findings are classified under four headings – General Observations, Looking Back, Capital Projections and Stress Testing.
The board of each institution is responsible for ensuring that the ICAAP is appropriate for that institution. The attached findings are intended to assist institutions in improving and refining their ICAAPs and Reports and to give guidance on how APRA supervisors approach the assessment process.
Institutions should also refer to the previous feedback on Summary Statements1 for additional guidance.
APRA expects ICAAPs, Summary Statements and Reports to continue to evolve and improve over time.
Keith Chapman, Executive General Manager
Diversified Institutions Division
Brandon Khoo, Executive General Manager
Specialised Institutions Division
Overall, the majority of ICAAP Reports reviewed were of a reasonable standard and represented an adequate initial attempt at compiling such a document. However there were some areas where institutions fell short of APRA’s expectations with regard to the content and quality of Reports. These areas included the comparison of planned versus actual ICAAP outcomes, description of changes in risk profile, commentary on factors driving future capital needs and the utilisation of the outcomes of stress testing in decision making processes.
In addition, for some Reports reviewed it appeared that the document had been approached as a compliance exercise to meet prudential requirements rather than a document for the benefit of the institution and the board. In a small number of cases this was clearly evident from the structure of the Report, with individual requirements transcribed verbatim from the prudential standard into the Report. APRA expects this ‘compliance focus’ to reduce in the future.
Linkage between ICAAP Summary Statement and ICAAP Report
APRA considers that a practical way to approach the Summary Statement and the Report is to have the Summary Statement explain the institution’s capital management processes, with the Report describing the outcomes of those processes. In essence, the Summary Statement and the Report should complement each other.
For some institutions it was noted that significant amounts of information from the Summary Statement were restated in the Report. While it is expected that institutions may choose to have some overlap of information for ‘context setting’, there appears to be little value in restating information from the Summary Statement in the Report.
APRA acknowledges that the level of duplicated information will likely reduce over time as institutions become more familiar with the specific purpose and use of the Summary Statement and the Report. As was the case with Summary Statements, a number of institutions acknowledged in the Report that this was their initial attempt at compiling such a document and as such it will be refined and improved over time.
- Has the board identified its needs for the Report and ways to streamline content where appropriate? Is there a plan in place as to when and how this will be done?
Statutory/benefit and general funds and Level 2 insurance groups
As was the case with Summary Statements, a number of Reports did not explicitly refer to statutory/benefit or shareholder/management funds (for life insurers and friendly societies respectively), or licensed entities for general insurers.
Reports for life insurers and friendly societies should address all statutory/benefit funds and the shareholder/management fund. For general insurers, Reports should cover the Level 2 group (where applicable) and individual licensed entities. The level of detail should reflect the size, nature and complexity of each fund and institution.
- Has the board satisfied itself that the Report adequately discusses the capital position over the period and the capital forecasts at all relevant levels?
Review of ICAAP
Most Reports included detail on the level of independent review undertaken during the Report period, together with the outcomes from those reviews; or specifically stated that no independent review had taken place. Some institutions had undertaken independent reviews of certain components of their ICAAP. Almost all Reports acknowledged that the ICAAP must be independently reviewed in its entirety at least once over a three year period as required under LPS/GPS 110.
Better practice Reports incorporated a schedule of future review activity, detailing which components of the ICAAP would be subject to review in the future and the planned timing of each review. Such a plan makes it clear that the institution is cognisant of its obligations in this area and provides a degree of comfort that these obligations will be fulfilled.
- Has the board clearly determined and articulated the nature and timing of future independent reviews of the ICAAP?
Current capital compared to minimum regulatory capital
All Reports reviewed included a comparison of the current capital position against minimum regulatory capital requirements, however the level and quality of information provided in this area varied.
For life insurers, it is inadequate if the Report excludes information on the shareholder fund, even where there was no significant activity undertaken by the shareholder fund in the period. Some general insurers similarly failed to include information on individual licenced institutions i.e. some Reports focused on the Level 2 group capital position only.
Better practice examples included a comprehensive table showing details of key indicators of the capital position including:
- capital base;
- Prescribed Capital Requirement (PCR);
- surplus above PCR;
- target surplus; and
- excess assets above target surplus at each level at the end of the period.
- Has the Board satisfied itself that the Report adequately considers key indicators of the capital position and trends over the period at each relevant level for the insurer or group?
Breakdown of PCR and capital base
There was a large degree of variation in the quality of information provided in this area. In some cases there was no, or minimal, information on the individual components of the PCR.
Better practice Reports included a table showing the individual PCR risk charge components. This was coupled with explanations as to how each amount was determined and why particular business activities are capital intensive in nature.
- Has the board satisfied itself that the analysis of the capital base in the Report is sufficiently granular to facilitate effective capital management?
Actual vs. planned outcomes of ICAAP
There was variation in the quality of information provided in the Reports reviewed. In some cases no comparison was provided between actual and planned outcomes whereas in other cases this comparison was done well. An explanation of why actual capital outcomes may have differed from planned outcomes is one of the critical areas of a Report and APRA would expect that information in this area would be of high quality.
Better practice would see the provision of detailed explanation on all factors, positive and negative, which have contributed to the difference between planned and actual capital outcomes during the period.
- Is the board satisfied with the frequency of review of projected and actual capital outcomes?
- Is the board satisfied that the Report adequately addresses any divergence between projected and actual capital outcomes; and that (as necessary) adjustments have been made to the capital plan?
Description of capital instruments
Institutions whose capital base currently consists solely of Common Equity Tier 1 (CET1) capital generally provided little by way of commentary as to whether this will remain an appropriate capital structure going forward. For institutions which have a more complex capital base made up of two or more types of instruments, these were generally adequately described, including the contractual obligations arising from these instruments as well as an explanation of the key differences between capital instruments.
Better practice examples commented on the continued appropriateness of continuing to solely use CET1 and also whether any changes in the capital base are expected. Reports also usefully could detail whether an institution has any capital options available to it other than CET1.
- Has the board articulated its thinking as to the current and future mix of capital; why a particular mix is appropriate; and what factors may have arisen in the period that has confirmed or changed the board’s view?
Material changes in ICAAP
Most Reports reviewed included a reasonable level of commentary on material changes made to the ICAAP during the period covered by the Report, while others explicitly stated that no material changes had been made. The materiality of the changes discussed varied between Reports.
The inclusion of information on material changes made to the ICAAP, together with a description of the impact each change had on the institution’s capital management processes during the period is considered better practice.
- Is the board satisfied that the Report clearly documents any proposed or actual material changes to the ICAAP?
Changes in risk appetite and risk profile
In general, Reports adequately described any changes made to the institution’s risk appetite and risk profile during the period, or alternatively stated that no material changes had been made to the risk appetite or ICAAP.
Better practice Reports provided further explanation as to the rationale for any change in risk appetite, rather than simply providing a description of the change; and, following this, provided detail on the implications of the change in risk appetite on the capital position of the institution.
- Is the board satisfied there is a process which ensures that, where there is a material change in risk appetite, the impact is considered in the ICAAP and there is a clear description of why (or why not) the ICAAP has changed as a result?
Economic capital vs regulatory capital reconciliation
Where institutions utilise internal capital models (whether for internal or regulatory purposes), these generally result in different capital outcomes to regulatory requirements. Better practice Reports provided a more complete explanation of the key factors driving the difference between capital outcomes. In addition, a description of capital applied to risks under the economic capital model, which are not specifically covered by the regulatory framework, also adds value to the Report.
Does the board consider there is adequate coverage in the Report on the relationship between economic and regulatory capital; and does the Report adequately identify the reasons for any difference between these?
Three-year capital projections
LPS/GPS 110 requires an institution to compile capital projections covering a minimum of three years and include the results of these projections in its Report. While all Reports reviewed included three year capital projections, the quality of the projections varied significantly. Some institutions provided detail in regard to three year capital projections while others were less robust and relatively basic. A small number of institutions provided projections out to five years.
Most life insurers and friendly societies broke down their three year projections by statutory fund/shareholder fund and benefit fund/management fund respectively, however a number of general insurers failed to break down projections by licensed insurer and focused on projections at Level 2 only. In some instances projections were relatively high level and failed to include detail on how the projections were developed and the key underlying assumptions driving them, thus limiting their value.
A number of institutions also failed to align capital projections to the strategic direction described in strategic and business plans, while many did not make reference to how past performance was taken into consideration (and some institutions reported against historically exaggerated expectations of performance which diminished the value of their ICAAP comparison).
Better practice examples included:
- commentary on how the projections were developed;
- a description of key drivers and underlying assumptions;
- reliable and realistic projections clearly aligned with business plans; and
- a description of how key risks had been factored into projections.
Some institutions with a Pillar 2 supervisory adjustment applied in addition to their PCA considered this adjustment to be temporary in nature and thus forecast their capital position against their PCA as opposed to their PCR. APRA expects the PCR to be reflected in capital projections.
- Has the board satisfied itself that longer-term capital projections are specified in sufficient detail in the Report?
Impact of changes in risk appetite, risk profile and other factors driving capital needs
Most Reports included information on planned changes in risk appetite and risk profile however the quality of this commentary varied. Detail on how potential changes in risk appetite and risk profile will impact an institution’s capital projections is expected to be included in Reports.
In addition, APRA expects institutions to provide detail on any changes which occurred during the period to the drivers behind future capital requirements by i) business activity; ii) geographic spread of exposures and iii) risk types. Better practice would see:
- institutions which operate in offshore locations providing a breakdown of any changes in capital usage by country;
- institutions with solely domestic operations including detail on any changes to exposures across states or exposures to particular “higher risk” locations throughout Australia; and
- discussion on any changes to projected capital needs by risk type.
Some institutions made no comment in this area and others referred to it only briefly. Only a few institutions included commentary which could be considered satisfactory. This is an area for improvement.
- Has the board satisfied itself that the Report is adequate to assist the board to understand clearly the capital implications associated with any changes in the institution’s risk appetite or risk profile during the period?
Self-assessed area for improvement
Stress testing was identified in many Summary Statements as an area where improvements were planned over the next 12-18 months. This was reiterated in a large number of Reports. In cases where institutions advised in their Summary Statement that they would provide a greater level of detail on stress and scenario testing in their Report, this was generally observed.
Better practice would see the process for stress and scenario testing being fully described in the Summary Statement, with the outcomes of this testing detailed in the Report. The Report should also consider what changes or enhancements to stress testing were planned for the period, including commentary on implementation or (if they were not implemented as planned), explanation as to why they were not implemented.
Use of outputs from stress testing/scenario analysis in decision making.
While the Reports revealed a marked increase in the level of attention given to stress and scenario testing by institutions, there remains a general lack of focus on how the outcomes from this testing are utilised. Many institutions do not appear to recognise the full value of stress and scenario testing as an input into key decision making processes including Risk Appetite Statements and the determination of capital buffers.
In a large number of cases, institutions used the outcomes from stress and scenario testing to justify their existing capital buffers and corresponding trigger points as being appropriate. We observed very few examples where the outcomes resulted in any adjustment to buffers and triggers.
It was also noted that relatively few institutions made changes to their ICAAP as a result of undertaking stress and scenario testing. This may suggest that such testing is still viewed primarily as a method of confirming that existing capital management parameters are appropriate rather than truly testing them.
Only a small number of Reports took the additional step of describing the management actions which would be considered when stress and scenario testing outcomes revealed a need. Better practice is to include the results of stress and scenario testing both before and after the impact of management actions in the Report.
- Has the board satisfied itself that the Report adequately discussed stress testing outcomes and links this to planned changes or enhancements to stress testing, capital buffers and triggers?
Sensitivity factors and scenarios
Sensitivity factors chosen by institutions for stress testing varied substantially, moving from relatively benign factors which had little impact on the institution’s future performance, to those which had a severe impact and sometimes resulted in a breach of PCR. Some institutions preferred to focus on less severe factors to show that they are capable of withstanding shocks which have a higher likelihood of occurring, while others chose factors which could be considered more severe but less likely to better test resilience to unforeseen stress.
Where scenario analysis was undertaken, some institutions appeared to underestimate the full impact of the scenarios, neglecting to consider potential “follow on” effects which might be expected to occur should a particular scenario ensue.
Stress tests should include a variety of severe but plausible sensitivity factors which are tailored to the business and risk profile of the institution as well as risks on the horizon. In addition to determining the individual stress factors, the likelihood of each factor occurring should also be taken into consideration.
Testing should place the institution under a degree of stress which might require remedial actions to be taken. This provides management and the board with a better understanding of the severity of factors that may cause concern, as well assisting in the development of a toolkit of mitigating actions that potentially could be implemented in appropriate circumstances.
Better practice Reports included scenarios that were highly relevant to the individual business operations of the institution.
Reverse stress testing
Only a small number of institutions included detail in their Report on the outcomes from reverse stress testing2. The level of detail provided on outcomes also varied and few institutions took the subsequent step of suggesting actions to be taken based upon the outcomes.
Better practice would be for institutions to use the Report as the primary method of advising the board of the outcomes from reverse stress testing, or detailing these outcomes in a separate document and providing a summary in the Report. Any capital management actions required as a result of reverse stress testing outcomes should also be highlighted in the Report.
- Has the board reviewed the outcomes from reverse stress testing and considered whether these outcomes require any change to future capital plans?
1APRA releases response paper on risk management consultation for private health insurers
2 Reverse stress testing is essentially a process in which an entity starts with the outcome (a given loss), and will then aim to identify a scenario that would cause this impact. It can be used to identify scenarios that would cause the entity to fail, relative to PCR or a different benchmark (or that would cause an unacceptable loss in a particular business or portfolio).