Feedback Response Paper

Transcription

1 Differential Premium System Feedback Response Paper July 2008 Ce document est également disponible en français

2 DPS Feedback Response Paper July, 2008 Page 2 Executive Summary This paper provides an overview of the recommended modifications to proposals for a revised Differential Premium System (DPS) based on the feedback received from stakeholders and recommendations of an industry advisory panel. It also contains detailed responses to the main comments and suggestions made by respondents to the original proposals outlined in the DPS Consultation Paper. An updated deposit insurance calculation tool is available on DICO s website which reflects the recommended modifications. Any further feedback on the modifications proposed may be sent by mail by September 30, 2008 to the following address: Deposit Insurance Corporation of Ontario 4711 Yonge Street, Suite 700 Toronto, ON M2N 6K8 Attention: Richard White, Manager, Policy & Research Alternatively, comments may be provided by to rwhite@dico.com. The modifications outlined in this paper also reflect input from an advisory panel of senior industry representatives. We wish to thank Jack Vanderkooy (DUCA Credit Union), Rob Wellstood (Kawartha Credit Union), Luc Racette (L Alliance des caisses populaires), Jean-Guy Laflèche (La Fédération des caisses populaires) and Adrienne McKenzie (Victory Community Credit Union - Advisory Council of Smaller Credit Unions) for their valuable assistance. Copies of this document and other related materials may be obtained by calling DICO at or by downloading them from DICO s Internet website at DICO proposes to seek Government approval for a revised process incorporating these changes as soon as possible after the revisions to the Act have been proclaimed. It will recommend adjustments to premiums as necessary to ensure that the reserve fund remains within an appropriate target range, and will also consider transitional measures for institutions that might be adversely affected in terms of their total assessment under the new system. At this time, given the uncertainty of the proclamation of the revisions to the Act, it is anticipated that the proposed changes to the DPS would not likely come into force before mid Recommended Modifications Risk Premiums: No change is proposed in the calculation of risk premiums. Administrative Premiums (Recovery of DICO Costs) Minimum Administrative premium of $250 for the first three years, increasing by $ per year thereafter for each of the next five years to a maximum of $1,000. Base Premium Rate of $0.30 per $1,000 for assets up to $50 million Secondary Premium: Balance of DICO s costs pro-rated for assets $50 million and over (declining effective premium rates) Premiums based on audited financial year end results (for both institutions and DICO)

3 DPS Feedback Response Paper July, 2008 Page 3 Background A Consultation Paper outlining proposals for a revised Differential Premium System (DPS) was circulated for comment in November These proposals included a Risk Premium to cover potential loss exposures to the Deposit Insurance Reserve Fund and an Administrative Premium to recover DICO s operating costs and, as summarized below: Risk Premiums The proposed Risk Premiums would have been based on a single capital measure: Leverage capital for Class 1 institutions Risk weighted capital for Class 2 institutions Premiums would be based on a continuous structure with premium rates ranging from $0.20 (for additional capital of 100% or more) to $0.70 (for additional capital of 0%) A penalty premium rate of $1.40 rate would apply for high risk institutions that do not meet minimum regulatory capital requirements. Administrative Premiums The proposed Administrative Premium would have included: a base premium (using asset tiers) a pro-rated premium for the balance of DICO s costs a minimum premium of $1,000. Impact on Institutions Aggregate premiums are expected to decline by approximately 25%. Feedback on the DPS Consultation Paper For the most part, responses to the proposals were positive and all responses indicated general support for the direction proposed as summarized in Table 1 below. Respondent Overall General Support Table 1: Comparison of Feedback Responses Dual Premium Structure Risk Premium Structure Risk Measure DICO Cost Recovery Base Assessment (Tiers ) Pro-Rated Assessment 1 Yes Yes No No No No 2 Yes Yes Yes Yes No Yes 3 Yes Yes Yes Yes No Yes 4 Yes Yes Yes No No No 5 Yes Yes Yes No Yes Yes 6 Yes Yes Yes Yes No Yes 7 Yes Yes Yes No Yes Yes 8 Yes Yes Yes No No Yes 9 Yes Yes Yes Yes No Yes 10 Yes Yes Yes Yes Yes Yes

4 DPS Feedback Response Paper July, 2008 Page 4 Feedback to the proposals was generally favourable and indicated strong support for: Separating premiums for DICO costs from risk premiums Adopting a continuous premium structure for risk premiums (elimination of tiers) Using capital as a risk measure Lower overall premiums However, several issues and recommendations were presented as follows: Utilization of capital as the only measure for risk premiums The method of calculation of the Administrative Premium for DICO costs The potential impact on smaller institutions The disproportionate level of DICO costs for larger institutions Full details of the feedback and DICO comments are provided in Appendix 1. Possible Options Based on the feedback received and input from an industry advisory panel, a number of possible options were developed for further research. These included: Risk Premiums Incorporating another risk measure (such as Net Risk 1 or Insurance Risk 2 ) using a 10%, 20% and 50% weighting in addition to the proposed capital measures Administrative Premiums Eliminating the base premium (tiers) and assessing all costs proportionately based on total assets Lowering the minimum Administrative premium payable to less than the proposed $1,000 Incorporating a secondary premium with a slightly declining effective premium rate structure as assets increase Using average assets as a basis for proportional allocation of costs 1 Net Risk is the risk profile rating established for each institution by DICO Insurance and Risk Management division. Risk Ratings are Low, Moderate and High. 2 Insurance Risk is Net Risk adjusted for capital and earnings. Risk Ratings are Low, Low-Moderate, Moderate, Moderate-High and High.

5 DPS Feedback Response Paper July, 2008 Page 5 Analysis Based on our extensive sensitivity analysis (using a number of combinations of possible options) we have found that: Adding another risk measure (either Net Risk or Insurance Risk) does not materially change overall premiums using a 10% or 20% weighting as both measures produce similar results. At a 50% weighting, minimum risk premium rates would need to increase significantly to meet target revenue levels Using a fully pro-rated structure for DICO costs and reducing the minimum premium to $250 significantly reduces the impact on smaller institutions and the number of institutions that could pay higher premiums. Incorporating a continuous declining effective rate structure for recovery of DICO costs will eliminate the inequities in the current proposals and still provide adequate premium relief for larger institutions. There are no material differences using average assets or total assets. However, as average assets for an institution could span two calendar years, this option adds unnecessary complication and therefore is not recommended. The proposed modifications satisfy the design principles set out in our initial DPS proposals. After further review and recommendations of the Advisory Panel, we are proposing to modify the DPS as outlined below. Recommended Modifications Risk Premiums No change is proposed in the calculation of risk premiums. Administrative Premiums (Recovery of DICO Costs) Minimum Administrative premium of $250 for the first three years, increasing by $ per year thereafter for each of the next five years to a maximum of $1,000. Base Premium Rate of $0.30 per $1,000 for assets up to $50 million Secondary premium: Balance of DICO s costs pro-rated for assets $50 million and over (declining effective premium rates) Premiums based on audited financial year end results (for both institutions and DICO) The changes proposed will help limit the impact on smaller institutions while recognizing the disproportionate share that larger institutions would pay for DICO s costs. The advisory panel strongly believes that smaller institutions should pay more towards the cost of DICO s operating expenses and that the minimum premium rate should increase over time. From our review, using additional capital as the single risk measure provides the most objective basis to measure risk to the Deposit Insurance Reserve Fund. Incorporating an additional risk measure will add unnecessary complexity and subjectivity. Extensive research has shown that there is little change in premium levels when incorporating an additional risk measure.

6 DPS Feedback Response Paper July, 2008 Page 6 The following table provides a comparison of the current DPS structure and proposed DPS structure to the design principles established for a differential premium system. Table 2: Design Principles Design Principles Fairness CURRENT DPS PROPOSED DPS Met Comments Met Comments Not Fully By incorporating tiers, institutions with higher risk can pay the same premiums as institutions with lower risk in the same premium tier. Yes Institutions with similar risk pay a similar premium. Continuous structure ensures that premiums are distinct for each different level of risk. Transparency Yes Structure and measurement criteria documented and published. However, institutions must understand five separate risk measures. Yes Administrative premiums more fairly applied based on costs. Uses a single statutory capital measure reported on the institution s audited financial statements Administrative premiums will be based on DICO s costs as reported in its Audited Financial Statements. Timeliness Not Fully Some risk measures are based on results over three years. This can result in a delay in premium adjustment for some institutions as a result of a change in risk. Also, Management component is based on latest OSV results which may be have been conducted up the 36 months earlier. Yes Uses capital as reported at fiscal year end when premiums are calculated. Administrative premiums will be based on DICO s latest audited results. Accuracy Not Fully Includes qualitative measure (Management) which is subjective in nature. Weights assigned to each measure may not always result in an accurate composite rating. Yes Yes. Based on a single quantitative risk measure using audited results. Administrative Premiums based on audited results. Efficiency Not Fully Premium calculations can be complex for amalgamations, purchase and sale transactions and changes to fiscal year ends etc. due to 3 year measures for earnings and asset quality. Yes Uses one single year end capital position for risk premium calculations. Administrative premiums based on DICO s year end results.

7 DPS Feedback Response Paper July, 2008 Page 7 Design Principles CURRENT DPS PROPOSED DPS Met Comments Met Comments Flexibility Yes Can be modified although modifications can be more complex due to number of risk measures, weightings and tiers Yes Premium rates can be adjusted as necessary to reflect higher or lower premium requirements without change to the risk measure. The base rate for the Administrative premium can be changed as necessary to reflect higher costs if needed. Affordability Not Fully Use of tiers limits premium relief as all institutions within tier pay the same premium. Not Fully Revised premium structure provides improved differentiation of premiums based on risk. Institutions can minimize premiums by maintaining higher levels of additional capital. Lowest premium rate of $0.20 compares to $0.14 for CDIC. (2008 premium year) Administrative premium will only be used to cover DICO s costs. Effective aggregate premium rate projected to decline from $0.97 to $0.73. Further details and examples regarding these modifications are provided in Appendix 2.

8 DICO Feedback Response Paper Appendix 1 July, 2008 DETAILED STAKEHOLDER FEEDBACK AND RESPONSE Feedback from stakeholders is summarized below followed by DICO s response in italics. Risk Premium Structure While we support the continuous structure in theory this could cause credit unions and caisses populaires to maximize capital rather than optimize capital. DICO Response The selection of the maximum capital limit of 16% (BIS) in the continuous structure was selected to minimize the incentive to maintain uneconomical levels of regulatory capital in order to reduce premiums. A number of credit unions and caisses populaires are currently reporting regulatory BIS capital levels of 16% or higher. We expect institutions to determine their own optimal capital taking into account minimum statutory requirements and guidelines, their particular business needs (including premium costs) and strategic plans. Using a single capital measure for risk premiums may encourage institutions to maintain a higher proportion of investment share capital thereby increasing reputational risk in the event of failure. Increasing capital through issuing investment shares for the purposes of reducing premiums may not be an effective strategy. The additional costs associated in holding investment share capital will usually not offset any corresponding decrease in risk premiums and is not encouraged. Recommend that deposit insurance premiums are established similar to those of Canada Deposit Insurance Corporation (CDIC). As result of the size and strength of its insured institutions, as well as its reserve fund, CDIC is able to have premium rates substantially below those of DICO. In addition, the current DPS is based to a large extent on the CDIC model although with much fewer risk measures and a modified premium structure. For example, the CDIC model has 11 quantitative measures and 2 qualitative measures. It also has four premium tiers where premium rates double from one tier to the next. We believe that the proposed risk premium structure and measures appropriately address many of the concerns raised with the current DPS and do not recommend retaining the current DPS model or adopting the more complex CDIC model. Risk Measure A number of responses indicated concerns regarding the proposed risk measure. Recommendations include incorporating DICO s assessment of the institution s net risk and earnings performance and risk management practices. One response raised concerns that the risk measure does not reflect profitability while another raised concerns that BIS reflects only gross risk and does not reflect net or residual risk.

9 DPS Feedback Response Paper Appendix 1 July, 2008 Page 2 Net risk and risk management practices are subjective or qualitative measures employed by DICO s staff to assist institutions in better understanding their risks and where there may be gaps in managing them. Using each institution s Net Risk assessment as part of a premium system to provide incentives for modifying risk behaviour may create an unnecessary conflict with examiners and DICO staff. While we do not disagree with the points made, we believe that incorporating one or more additional measures into the calculations for risk premiums such as risk and earning, etc. would add further complexity and result in a structure with similar challenges faced under the current DPS. These include determining and obtaining support for the appropriate elements and relative weights of each measure. As part of its scoring system, CDIC uses an examiners rating based on OSFI s net risk rating for each institution. This additional element is intended to penalize those institutions that do not have a good risk management system in place and to act as an incentive for them to manage their risks appropriately. However, DICO does not believe that this type penalty/incentive necessarily translates directly into better risk management practices. In fact, our experience has been that poorly managed institutions are quite content to pay a higher premium as long as regulatory action is not taken. We are confident that capital is a good measure of current risk to the DIRF as it represents the deductible in case of an immediate failure. Net Risk provides a more prospective view of risk. While BIS is a relatively good indicator of risk it does not fully recognize differences in the quality of security (e.g., trading assets versus real estate). While some institutions may have higher quality security for some loans, the overall risk in the portfolio is generally expected to be similar for most institutions with the same balance sheet structure. The BIS regime is designed to address issues such as this. The proposal calls for a modification of BIS but does not provide any suggested methodology or supporting rational. In addition, DICO s Bylaw #6 sets out the requirements for setting aside the required allowances for loans. Incorporating another measure for underlying credit quality would penalize those institutions which have set aside appropriate general allowances. We expect the pending implementation of the new IFRS accounting standards will further enhance the reliability of the BIS capital regime.

10 DPS Feedback Response Paper Appendix 1 July, 2008 Page 3 Administrative Premiums to cover DICO's Costs We are concerned about our ability to influence or control the nature and level of DICO s operating expenses. We also suggest any increase in DICO s costs be deferred pending equivalent reductions in FSCO s charges. Our proposal calls for billing the Administrative premium for DICO s operating costs in arrears so that the aggregate costs being recovered can be readily determined from DICO s audited financial statements. Operating costs will exclude any amounts for loss provisions and will only include direct operating expenses. While we anticipate an overall reduction in the overall costs of regulatory oversight, any potential reduction in FSCO charges should be addressed directly with FSCO. Will there be a mechanism to adjust premiums collected that exceed DICO s actual expenses? It is proposed that premiums to recover DICO s costs be based on actual year end audited results and billed in arrears. Therefore, there should be no excess premiums collected. There should be a discount for members of certain leagues where the league monitors its own network which materially reduces the network monitoring cost for DICO. As part of the proposed regulatory changes, institutions may enter into group capital arrangements for the purposes of determining capital adequacy. This may provide an opportunity for determining risk premiums under this approach, which may help reduce premium costs for group participants. There is also the ability to establish voluntary stabilization funds for the mutual benefit of participants which if included in a group capital arrangement would have the effect of increasing capital and thus reducing premiums. Use average assets rather than total assets in order to eliminate potential distortions. Using average assets for both the base and pro-rated premiums would have minimal impact on premiums as the effect would apply to all institutions. This structure was mirrored on the OSFI approach. However, given that the DICO cost recovery premiums will be assessed for costs already incurred by DICO, using total assets based on the latest audited information will be more timely and appropriate. Audited results for each institution will be available for the same year, whereas average assets based on audited information will cover two years for institutions with fiscal years ending other than in December.

11 DPS Feedback Response Paper Appendix 1 July, 2008 Page 4 Use all off-balance sheet assets (as opposed to only those used for calculating BIS capital). Do not use any off-balance items in the calculations. We do not believe that using all off-balance sheet items is appropriate. Any deviation from regulatory BIS capital rules will require extensive consultation and debate, and will likely not achieve any meaningful form of consensus. The base premium amount should only apply to institutions under $10 million. Limiting the base premium to only smaller institutions under $10 million does not fully address the issue of the higher relative costs for smaller institutions. Smaller institutions, especially Class 1 institutions, account for a disproportionate amount of DICO resources and expense. Under proposed modifications, the base amount will apply to assets up to $50 million. Several responses raised concerns and suggestions regarding the method used to determine the base premiums. Several respondents suggest that the tier structure leads to inequities. One respondent also questions the non-linear method for allocating these costs due to the use of tiers for the base portion of the cost recovery premium. Another respondent recommends eliminating the base premium and adopting a continuous premium structure (with a decreasing marginal increase in premiums as assets increase). The use of tiers for the base premium was proposed to modestly reduce the burden on larger institutions and is based on the OSFI approach for banks. Using tiers will always result in some inequities especially where institutions transition into a higher premium tier and see an immediate bump in the base rate. The alternative of eliminating the base premium and adopting a single fully continuous structure would resolve most of the issues raised but would lead to higher premiums for larger institutions. Implementing a base rate (for assets up to $50 million) and a pro-rated structure for the balance of DICO s costs will result in slightly declining rates for institutions with assets of $50 million or more and eliminate the inequities relating to the use of tiers. A number of concerns were raised over the impact of minimum assessments for smaller credit unions. While there will likely be modest increases in costs for some of the smaller institutions with the introduction of the minimum rate of $1,000, it is felt that these are fully justified in view of additional costs associated with smaller institutions. In order to reduce the immediate impact of this increase, the minimum rate will be initially set at $250 for the first three years increasing thereafter over five years to $1,000.

12 DPS Feedback Response Paper Appendix 1 July, 2008 Page 5 Aggregate Premiums No institution should pay a higher premium under the new system. Other From our analysis, only a few institutions may initially pay a higher premium under the proposed modified structure based on current information. We plan to develop transition rules to ensure that no institution will pay higher premiums for the first three years after implementation of these changes. There should be an immediate premium reduction as the DIRF is within the target range The DIRF is currently only at 61 basis points which is well short of the long term upper range target of 68 basis points. In light of the current uncertainties in the Ontario and North American economy, we believe it would be imprudent to reduce premiums until our fund target has been achieved. Also, we believe that any premium reductions should coincide with implementation of a revised DPS. This option is, therefore, not considered to be viable at this time.

13 DPS Feedback Response Paper Appendix 2 July, 2008 TECHNICAL SUPPLEMENT: ADMINISTRATIVE PREMIUMS 1. Administrative Premiums to recover DICO operating costs are based on total assets (including specified off-balance sheet assets) as reported in the latest audited financial statements and will include: (a) Assets up to $50 million o Base premium rate of $0.30 per $1,000 of assets (b) Assets $50 million and over o Secondary Premium for balance of DICO s operating costs (c) Minimum Administrative Premium (Refer Table 1) Table 1: Minimum Premium Rate Schedule Year Minimum Premium Year Minimum Premium 1 $250 5 $550 2 $250 6 $700 3 $250 7 $850 4 $400 8 $1, An illustration of potential Administrative premiums and rates is provided in Table 2 below. Table 2: Examples of Premium Calculation 3 System A Total Assets $26,000,000,000 $10,000,000 $100,000,000 $500,000,000 4,000,000,000 B Total Assets up to $50 $6,500,000,000 $10,000,000 $50,000,000 $50,000,000 $50,000,000 C D E F G million Base Premium Fixed Premium rate $0.30 per 1,000 for assets up to $50 million Balance of Assets over $50 million Balance of DICO Costs to be Pro-Rated Secondary Premium Estimated rate per $1,000 Secondary Premium Amount $1,950,000 (B*$0.30/1000) $19,500,000,000 (A-B) $4,300,000 ($6,250,000 C) $ (E*1000/D) $4,300,000 (D*F/1000) $3,000 $15,000 $15,000 $15,000 $Nil $50,000,000 $450,000,000 $3,950,000,000 $ $ $ $ $NIL $11,025 $99,231 $871,026 H Total Premiums $6,250,000 (C+ G) $3,000 $26,025 $114,231 $886,026 I Effective Premium $ $ $ $ $ Rate per $1,000 (H*1000/A) 3 Assumes DICO costs of $6,250,000

14 DPS Feedback Response Paper Appendix 2 July, 2008 Page 2 3. A summary of the projected impact on premiums from the modifications proposed is provided in Table 3. Table 3: Summary of the Projected Impact on Premiums Current Estimated Risk Premiums $11,950,000 Current Estimated Aggregate Premiums $18,200,000 Projected Risk Premiums under revised DPS $7,250,000 Projected Reduction in Risk Premiums 39% Projected Aggregate Premiums under revised DPS $13,500,000 Projected Reduction in Aggregate Premiums under revised DPS 26% 4. Declining Effective Premium Rates The following graph illustrates the potential effective premium rates based on the latest available information. For assets up to $50 million, the effective premium rate is the same as the base rate of $0.30 per $1,000 of assets. The effective premium rate for institutions with assets above $50 million gradually declines as assets increase. Sample Administrative Premiums Rate per $1,000 of assets $0.35 $0.33 $0.31 $0.29 $0.27 $0.25 $0.23 $0.21 $0.19 $0.17 $0.15 Average Rate Effective Rate <$50 M <$100M $500M $4B Asse ts 5. Estimated Premium Rates With the proposed reduction in aggregate premium levels, the estimated aggregate premium rate in Ontario is expected to decline from $0.97 to $0.73. Sample comparative estimated rates are provided in Table 4 below. Table 4: Comparative Estimated Aggregate Premium Rates Based on Estimated Insured Deposits CDIC B.C. Alberta Sask. Manitoba Quebec Nova Ontario Ontario Scotia 2007 Projected $0.16 $0.50 $1.70 $0.30 $1.00 $0.40 $1.00 $0.97 $0.73