pwc.com.au NT WorkSafe Actuarial review of Northern Territory workers compensation scheme as at June 2014

Transcription

1 pwc.com.au NT WorkSafe Actuarial review of Northern Territory workers compensation scheme as at June June 2013

2

3 Key findings Funding ratio The following table shows the funding ratio for the scheme for insurers and self -insurers both separately and combined. The definition of the funding ratio is contained in the Glossary. The PwC central estimate excludes a risk margin while the insurer s provision includes a risk margin and the self-insurer s provision includes the 50% loading for the bank guarantee. Funding ratio Actual PwC central Difference ($) Funding ratio provisions (a) estimate (b) (b) - (a) (a) / (b) Insurers 267, ,777 26,121 91% Self-insurers 8,349 6,772-1, % Total 276, ,549 24,544 92% Notes: see section 2 of this report In this context the funding ratio is measuring the liabilities held by the insurers (the notional assets) compared to the aggregate outstanding claims liability calculated by the scheme actuary. This represents the ability of the scheme in aggregate to meet its liabilities. A funding ratio of less than 100% does not necessarily mean that individual insurers will be unable to meet their obligations to the NT Scheme. This is because: The insurers actuaries have access to more detailed claims data than we have to estimate the liabilities Insurers hold assets in excess of their liabilities due to APRA capital requirements and their own risk appetite While in aggregate the funding ratio is 91% in actuality some insurers could be holding greater than 100% of our notional allocation to them while another may be holding less than the 91% The total aggregate funding ratio of 92% has increased from 79% last year. This is mainly driven by the reduction in the net provision estimated by ourselves compared to the previous scheme actuaries due to changes in modelling and assumptions. Both the insurers and self-insurers funding ratio have increased; they were 78% and 112% respectively last year. The above projections involve assumptions about future uncertain claim events and economic, social and legislative conditions and hence the actual outcome may well be different from the results shown above. This should be borne in mind whenever using the results. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June 2013 i

4 Key findings Insurer premium adequacy The following graph shows the calculated break-even premium rates, using hindsight data and compares this to the actual premium rates charged by insurers, and the projected 2013/14 break-even premium rate. Adequacy of past premium rates and projected rate for 2013/14 3.5% 3.0% 2.5% 3.2% 2.6% 2.8% 2.6% 2.6% 2.5% 2.5% 2.4% 2.5% 2.2% 2.2% Calculated break-even premium rate Actual premium rate charged 2.3% 2.0% 2.2% 2.2% 2.1% 1.9% 2.2% 2.1% 2.2% 1.5% 1.0% 0.5% 0.0% Underwriting year The break-even premium rate is calculated on an inflated and discounted basis, is gross of reinsurance and does not allow for a profit margin. As the claims costs and expenses have been discounted to the point that the premium is received, the calculated break-even premium rates have reduced since the previous scheme actuary s review. The actual premium rates charged by insurers over the most recent three underwriting years is slightly higher than the hindsight break-even premium rate based on claims experience to date. With hindsight the actual premium rate charged was more than sufficient to cover the break-even cost for all accident years except 2009 and Our projected break-even premium rate for 2014 is 2.2% which is slightly higher than the 2.1% for 2013 but similar to We are concerned that the wage and premium information provided each financial year on the consolidated ANZSIC data includes wage or premium adjustments that relate to historical underwriting periods. This will distort comparisons, particularly for a growing economy, as wage adjustments for previous years will artificially increase the current year s wages and premium received. Before the next valuation we will investigate this so the wages and premium for each underwriting year may change in next year s report. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx ii 11 June 2014

5 Key findings Key scheme trends Claim statistic Insurer Self-insurers Number of claims incurred Slight decreasing trend, though 2013 is 2.9% higher than 2012 with 2,567 claims Claim frequency is declining at 3.6% for 2013 and 3.8% for 2012 Decrease in 2012 and 2013 to 139 claims due to change in reporting of small claims by one self-insurer Average claim size Relatively stable at just below $37,000 per claim in 30 June 2013 values Lower than insurers at $17,000 for is lower than 2012 Incurred cost Relatively stable, 2013 higher than 2012 due to increase in number of claims 2013 is below 2012 but higher than most previous years Gross loss ratio Improved in 2013 to 75% from 85% in 2012 due to growth in earned premium n/a Distribution by payment type Redemption and non-economic lump sums and weekly benefits combined account for two-thirds of the total incurred cost and payments each financial year. The distribution of payments for the last seven accident years has remained fairly stable n/a Data enhancements To enhance the quality of future reviews we have suggested a number of improvements on existing data collected and additional data items which are detailed in section 5.3. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June 2013 iii

6 Contents Key findings i 1 About this report Context for our review Disclaimer Compliance with standards 3 2 Insurer outstanding claims liabilities Outstanding claims liability Claims statistics Actual vs expected claims experience over 2012/ Reconciliation of estimates 12 3 Self-insurer outstanding claims liabilities Outstanding claims liability Claims statistics Actual vs expected claims experience over 2012/ Reconciliation of central estimates 18 4 Break even premium rates Adequacy of past premiums Forecast break even premium rate 24 5 Data and methods Data provided Data quality and reconciliation Data enhancements and additional data Projection methods for outstanding claims Approach to estimate break-even premium rates 30 6 Assumptions Financial Assumptions 33 L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx iv 11 June 2014

7 PwC 6.2 Superimposed inflation Expenses Reinsurance 35 7 Uncertainty Uncertainty in the estimates Determination of provisions 38 Appendix A Detailed data description 43 Appendix B Assumptions 49 Appendix C Insurer outstanding claim valuation 59 Appendix D Insurer claims statistics 75 Appendix E Insurer financial year claims experience 82 Appendix F Self-insurer outstanding claims valuation 91 Appendix G Self-insurer claims statistics 98 Appendix H Insurer break-even premium rate 103 Appendix I Glossary 110 NT WorkSafe v

8

9 1 About this report Key points of this section NT WorkSafe have engaged us to value the outstanding claims liability for the scheme as at 30 June 2013 and review the adequacy of premium rates charged by insurers This is the first time that we have conducted this review. The previous review as at 30 June 2012 was conducted by Marsh. We have complied with the relevant actuarial and accounting standards when performing this review. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

10 About this report PwC 1.1 Context for our review This report has been prepared for NT WorkSafe and the Scheme Monitoring Committee in accordance with contract number D , dated 31 October Under this contract we have conducted the following analyses which are detailed in this report: Calculation of the funding ratio based on 30 June 2013 outstanding claims liability valuations for insurers and self-insurers Calculations of the break even premium rate for each prior accident year using data to 30 June 2013, including a review of the trends in the required premium and a comparison to the actual premium rates charged by insurers. An estimate of the break-even premium rate for the 2013/14 underwriting year, based on historic data and future inflation assumptions. This is the first time that we have prepared this report for NT WorkSafe. The previous valuation as at 30 June 2012 was conducted by Marsh and we have used their report, dated 30 April 2013, for comparative purposes in this review. The scope of this report is the same as the previous report. We used 30 June 2013 data for this, our first, scheme review based on the same data formats used previously. In section 5.3 we have suggested future data enhancements to expand the insight offered in this report. Our review is for the following five active insurers: Allianz Australia Insurance Limited CGU Insurance Australia (Part of Insurance Australia Limited) GIO Insurance Australia QBE Insurance Australia Territory Insurance Office and the following four active self-insurers: Catholic Church Insurance Coles Supermarkets Australia Pty Ltd Westpac Banking Corporation Woolworths Supermarkets. The analysis excludes Government Self Insurance and uninsured claims. We have performed separate analyses for insurers and self-insurers. The report is structured as follows: Sections 2 and 3 of this report present the outstanding claims liability valuations for insurers and selfinsurers respectively. Section 4 analyses the break even premium rates for past underwriting years and the adequacy of the rates actually charged by insurers. Section 5 details the data and methodology we have used for this. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx 2 11 June 2014

11 PwC About this report Section 6 and section 7 outline the assumptions adopted in this review and considers the uncertainty in the work we have carried out, including some key risks faced. 1.2 Disclaimer Report and Advice This report has been prepared for the sole use and benefit of NT WorkSafe. It should not be used or relied upon by any other person for any purpose. You agree to use this report only in connection with the purpose in respect of which this report is provided being to present the outstanding claims liability as at 30 June 2013 and review the adequacy of premium rates charged by insurers operating under the scheme. We therefore accept no liability or responsibility for any loss or damage arising from use of the report for any other use or purpose. Judgements based on the contents of this report should be made only after studying the report and the appendices in their entirety, as conclusions reached by a review of an aspect or section in isolation may be misleading. The advice contained in this report has been prepared on the instructions of NT WorkSafe in accordance with the terms of reference in the tender document referred to in section 1.1 above and is based on the information and data provided to us. The conclusions reached in this report are reliant on the completeness and accuracy of information compiled and provided by NT WorkSafe, and by insurers and self-insurers to NT WorkSafe. Other than preliminary data checks, we have not conducted an independent review of this information. We do not accept any liability or responsibility for errors or omissions arising from the provision of inaccurate or incomplete information to us. Third Parties This report and the advice contained in it are confidential. You agree not to disclose the report and/or our advice to third parties by any means (including orally or in writing) without our prior written consent. We may, at our discretion, withhold or give our consent subject to conditions, including: the report is to be released in its entirety in response to a request, including all appendices we accept no liability or responsibility to any other person or entity other than NT WorkSafe in relation to this report and no-one other than NT WorkSafe should rely on this report for any purpose. 1.3 Compliance with standards Outstanding claims liabilities The approach for calculating the outstanding claims liabilities is consistent with that required by the Accounting Standards for private and State Government general insurers (AASB1023), and APRA s prudential standard NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

12 About this report PwC GPS320 Actuarial and Related Matters where applicable. It also complies with the Institute of Actuaries of Australia's Professional Standard PS300 to the extent possible given the data available. We have not performed a full review of asbestos liabilities due to lack of available data Premium rates Our advice to you constitutes a Professional Service as defined in the Code of Professional Conduct (the Code) issued by the Institute of Actuaries of Australia and our advice complies with the Code in this respect. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx 4 11 June 2014

13 2 Insurer outstanding claims liabilities Key points of this section Our estimate of the net outstanding claims provision as at 30 June 2013 is $336.5 million, which is $40.6 million (10.8%) lower than the provision prepared by the previous scheme actuary as at 30 June 2012 This provision is $68.8 million (25.7%) higher than insurers own provisions of $267.7 million. This difference is less than the 30 June 2012 difference of $119.5 million (31.7%) The funding ratio is 91% which is higher than 78% last year Over the last few accident years there has been a declining to stable trend in the number of claims incurred and claim frequency. Average claim size has been relatively stable over the same period The experience over the 2013 financial year showed that claim reports in relation to the 2012 and prior accident years were higher than expected in the previous scheme actuary s 30 April 2013 report, however claim payments were lower than expected The reconciliation of our gross estimates, excluding claims handling expenses to the estimates prepared by the previous scheme actuary as at 30 June 2012 shows a release of reserves of $38.7 million, which is 11.9% of the opening estimates. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

14 Insurer outstanding claims liabilities PwC 2.1 Outstanding claims liability Our estimates The table below shows our central estimate results by payment type group in current values and excluding claims handling expenses: Accident year ending 30 June Estimates of outstanding claims at 30 June 2013 ($000s) (a) (b) Allied Health, Vocactional Rehabilitation, Non- Compensation Weekly Medical And Payments Other Goods Benefits Hospital (Other), Death And Services Legals Redemptions And Non- Economic Lump Sum Active large claims allowance ,666 6,265 5,992 4,618 3,769 32, , ,835 2,597 2,921 2,799 3,096 27, , ,569 1,624 1,711 1,684 2,228 20, , ,945 1,322 1,342 1,261 1,821 17, , , ,089 9, , , ,944 5,388 21, , ,633 7,100 18, , , , , , , & earlier 7,908 2,193 1,767 2,605 2,156 18, ,692 Total 68,928 17,577 16,505 15,618 16, ,684 12, ,234 Total Notes: (a), (b) from appendix C4 To generate the gross central estimates, the current value estimates are inflated and discounted, as follows: Gross estimates at 30 June 2013 excluding expenses ($000s) Accident year ending 30 June 13 Inflated Infl/disc 30 June values values values ,812 85,362 75, ,849 58,374 51, ,323 40,601 35, ,265 33,490 29, ,030 20,030 17, ,573 25,030 21, ,719 21,473 18, ,994 8,401 7, ,977 7,192 6, & earlier 34,692 41,895 34,981 Total 295, , ,392 L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx 6 11 June 2014

15 PwC Insurer outstanding claims liabilities An allowance for reinsurance recoveries, a claims handling expense and a risk margin are included on the gross inflated/discounted estimate to find the net outstanding claims provision: Estimates at 30 June 2013 ($000s) Accident year Gross o/s Reinsurance Net o/s Claims handling Net central Risk Net ending 30 Jun liability (a) recoveries (b) liability (c) expenses (d) estimate (e) margin (f) Provision (g) Total 298,392 21, ,148 16, ,777 42, ,494 Notes: (a) (b) (c) (d) (e) (f) (g) from table above allows for 100% reinsurance recoveries on large claims and 3% on all other claims = (a) (b) = (c) x 6%, see section 6.3 for details of the claims handling expenses = (c) + (d) = (e) x 14.54%, see section for details on the risk margin = (e) + (f) The net provision at 30 June 2013 is $336.5 million, which is $40.6 million (10.8%) lower than the $377.0 million estimated provision prepared by the previous scheme actuary as at 30 June The main cause of this decrease is the change in methodology from an aggregate payment per claim incurred (PPCI) method to six payment type groups valued using either the payment per active claim (PPAC), payment per finalised claim (PPCF) or PPCI method depending on their characteristics. The change in valuation approach has decreased the gross central estimate, excluding expenses. We have separated the payment type groups to provide a more detailed analysis of the outstanding claims liability and understand the trends driving the overall cost. The PPAC method is predominately used for payment types which are regular, ongoing payments while we use the PPCF method for lump sum style payments. The PPAC and PPCF methods are better for valuing the tail liabilities than the PPCI method as the number of active claims is considered in estimating the liability. However a further enhancement to reduce the uncertainty of the tail liabilities would be to use a projected case estimates (PCE) method which considers the size of any active claims. See section 5.3 for more information on this data improvement. The first table presented in this section shows that the largest component of the outstanding claims liability relates to the redemptions and non-economic lump sum payment group (50% of the total), followed by weekly benefits (23% of the total). 90% to 95% of redemptions and non-economic lump sum payments are Hopkins settlements. Net results have only been provided in total as reinsurance recoveries depend on the large claims experience in each accident year and individual insurer s reinsurance treaties. As we do not have information on either, the reinsurance recoveries are a high level estimate and therefore the results are not presented by accident year. We did refine the reinsurance recovery method by separately estimating recoveries on the largest known active claims. Further detail on the parameters adopted to calculate the outstanding claims can be found in Appendix C. For further analysis on the composition of the incurred cost of claims by payment group see Appendix E2. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

16 Insurer outstanding claims liabilities PwC Comparison with insurers We have compared our gross outstanding claim estimates in inflated and discounted values excluding claims handling expenses to the insurers estimates by accident year. As stated above we have not compared the net results by accident year due to the lack of appropriate reinsurance data to enable us to assess the expected reinsurance recoveries by accident year. Gross estimates at 30 June 2013 excluding expenses ($000s) Accident year ending Insurers' PwC Difference ($) Difference (%) 30 June estimate (a) estimate (b) (b) - (a) (b) / (a) ,860 75,688 6, % ,384 51,439 3, % ,432 35,682 7, % ,210 29,498 7, % ,166 17,157 3, % ,615 21, % ,159 18,997 1, % ,963 7,041 1, % ,517 6,020 1, % 2004 & earlier 33,130 34,981 1, % Total 264, ,392 33, % Notes:(a), (b) = gross inflated and discounted values excluding claims handling expenses The comparison of the total net outstanding claims provision is: Net provision at 30 June 2013 ($000s) Insurers' PwC Difference ($) Difference (%) provisions (a) provision (b) (b) - (a) (b) / (a) - 1 Total 267, ,494 68, % Notes:(a), (b) = net inflated and discounted values including reinsurance, claims handling expenses and risk margin This comparison shows that our gross estimate is higher than the insurers gross estimate for all accident years except The gross estimate is higher due to different methods and assumptions used by insurers and by us in the valuation. The total difference is higher for the net provision due to different calculations for reinsurance recoveries and different risk margin assumptions. It would be expected that the scheme risk margin is higher than for individual insurers as their risk margin is lower due to diversification benefits from writing other classes of business. While reviewing our estimate compared to insurers estimates we became aware of outstanding large claims (total estimate >$5 million) in each of the 2007 and 2008 accident years. We did not receive any data on case estimates but we did request extra information about these two large claims so we could properly incorporate an allowance for them in our provision. For future valuations we will request reinsurance recovery case estimates and conduct a more thorough large claims analysis. See section 5.3 for more details of this extra data request. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx 8 11 June 2014

17 PwC Insurer outstanding claims liabilities Funding ratio For insurers the funding ratio compares the insurers net provision (ie including risk margin) with our central estimate (ie excluding risk margin). This shows: Funding ratio Insurers' PwC central Difference ($) Funding ratio provisions (a) estimate (b) (b) - (a) (a) / (b) Total 267, ,777 26,121 91% Notes: (a) (b) as per table above, net provision including risk margin net central estimate, excluding risk margin In this context the funding ratio is measuring the liabilities held by the insurers (the notional assets) compared to the aggregate outstanding claims liability calculated by the scheme actuary. This is used to represent the ability of the scheme in aggregate to meet its liabilities. The aggregate funding ratio is 91% which is an increase from 78% last year driven primarily by the lower central estimates we have calculated relative to the previous scheme actuary s estimate last year. A funding ratio of less than 100% does not necessarily mean that individual insurers may be unable to meet their obligations to the NT Scheme. This is because: The insurers actuaries have access to more detailed claims data than we have to estimate the liabilities While in aggregate the funding ratio is 91% in actuality some insurers could be holding greater than 100% of our notional allocation to them while another may be holding less than the 91% Insurers hold assets in excess of their liabilities due to APRA capital requirements and their own risk appetite. 2.2 Claims statistics The following sub-sections show the claims experience by accident year. For more graphs of claims statistics, including by financial year, see appendices D and E. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

18 Insurer outstanding claims liabilities PwC Claim reports There is a slight decreasing trend in the number of claims incurred, although 2013 is 2.9% higher than 2012 Number of claims incurred 3,000 2,500 2,000 1,500 1, IBNR Claims reported Previous year Claims incurred Accident year The main points to highlight from this chart are: Since 2005 the number of claims incurred has been volatile from one year to the next, but has generally exhibited a declining trend Excluding the low in 2007 claims incurred between 2005 and 2008 were just over 2,700 For the 2009 to 2012 accident years the number of claims incurred has decreased to between 2,500 and 2,650 The 2013 accident year has 2,300 claims reported to date and we project approximately 300 claims are incurred but not reported, giving an incurred number of claims of 2,600 Incurred claims for 2013 are 72 (2.9%) more than the 2012 accident year. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx June 2014

19 PwC Insurer outstanding claims liabilities Declining claim frequency Claim frequency per $75,276 of real wages 7% 6% 5% 4% 3% 2% 1% 0% Accident year See Appendix D1 for the formula to calculate the claim frequency Gross average claim size Broadly stable trend in real average claim size since 2007 Gross average claim size in 30 June 2013 values 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 Current year Average O/S claim size Current year Average paid claim size Previous year Average claim size Accident year Since 2005 the average claim size (in 2013 values) has been: around $28,000 in 2005 and 2006 a significant increase to $38,000 in 2007 between $35,000 and $40,000 from 2008, except for a low $31,900 in 2011 $36,700 in 2013 which, whilst still uncertain, is on par with NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

20 Insurer outstanding claims liabilities PwC The uncertainty about the future development means that the ultimate level and our estimates may differ from that projected for recent accident years. This is especially true for the 2013 accident year where a high proportion (82%) of the average claim size relates to uncertain future claims development. Since the previous valuation the gross average claim size has decreased significantly for the 2011 accident year and decreased slightly for the 2005, 2006 and 2012 accident years but has increased for 2007 to These changes reflect the change in method to take into account more information about the number of active claims. Appendix E contains the average claim size split by payment type. The mix of payment types across the accident years has remained fairly stable. Redemptions and non-economic lump sums (90% to 95% are Hopkins settlements) are the largest payment type, closely followed by weekly benefits. These two payment types account for two thirds of total incurred costs. 2.3 Actual vs expected claims experience over 2012/13 Actual experience compared to the expected experience over 2012/13 for claims incurred up to 30 June 2012 showed: Claim reports were higher than expected (272 actual vs 244 expected) Claim payments were similar to expected ($68.7 million actual vs $69.1 million expected). Expected experience is taken from the previous scheme report dated 30 April See appendix C2 for full details. The impact of this favourable experience is quantified in the reconciliation below. 2.4 Reconciliation of estimates The table below reconciles the gross outstanding claims central estimate, excluding expenses presented in 2.1 with the equivalent result as at 30 June Reconciliation of gross actuarial estimates, excluding expenses ($000s) Accident year ending 30 June Total & earlier A. Gross estimates at 30 June 2012 (a) 76,732 58,054 40,654 31,121 24,211 17,003 14,692 61, ,925 B. Gross payments 1 July 2012 to 30 June ,383 11,431 7,751 16,270 3,794 2, ,328 71,137 C. Assumed investment return (b) 1,949 1,561 1, ,744 8,595 D. = A - B + C 56,297 48,184 34,000 15,528 21,084 14,954 14,461 56, ,383 Updated gross estimates at 30 June 2013 E. Revised gross estimates at 30 June 2013 (c) 51,439 35,682 29,498 17,157 21,889 18,997 7,041 41, ,704 F. = E - D -4,858-12,502-4,502 1, ,042-7,420-15,873-38,679 Change 01 July 2012 to 30 June 2013 G. Proportion of change attributable to -5,092-3,618-2,945-2,126-2,304-1,812-1,010-5,992-24,900 changes in real rates of return assumed H. = F - G 234-8,883-1,557 3,755 3,109 5,855-6,411-9,881-13,779 Balance of change attributable to changes in underlying actuarial estimates I. Gross amount incurred and outstanding for 75, /13 accident year (e) J. = E + I 298,392 Total gross outstanding liability, excluding expenses at 30 June 2013 L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx June 2014

21 PwC Insurer outstanding claims liabilities Notes : (a) from Exhibit 7 of the previous scheme report dated 30 April 2013 (b) (c) calculated using 3.0% pa being the one year forward rate from section 6 of the previous scheme report dated 30 April 2013 from appendix C4 of this report. The table shows that: overall estimates show a release of reserves of $38.7 million, which is 11.9% of the opening 30 June 2012 estimates. This release is made up of a $24.9 million release (7.7% of opening estimates) due to changes in the real rates of return, and a $13.8 million release (4.3% of opening estimates) due to changes in the underlying estimates all accident years except 2007 to 2009 had releases in the outstanding claims estimate the releases due to changes in real rates of return are due to the real rates of return increasing from a flat -3% to real rates that range from -1.2% to +1.6%. these releases due to changes in underlying actuarial estimates are due to a change in method to the PPAC and PPCF methods which are more sensitive to the number of active claims, rather than the PPCI which only considers the total number of claims. the strain on reserves for the 2007 and 2008 accident years is due to the allowance for large claims in those years. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

22 3 Self-insurer outstanding claims liabilities Key points of this section Our estimate of the net outstanding claims provision as at 30 June 2013 is $7.8 million, which is $0.6 million (7.3%) lower than the provision prepared by the previous scheme actuary as at 30 June 2012 Our inflated and discounted central estimate, including claims handling expenses is $1.2 million (21.7%) higher than self-insurers central estimates of $5.6 million The funding ratio is 123% which is higher than 112% last year Claim incurred numbers reduced in 2012 and 2013 while the average claim size increased in 2012 but reduced in 2013 The experience over the 2013 financial year showed that claim reports in relation to the 2012 and prior accident years were higher than expected based on the previous scheme actuary s 30 April 2013 analysis, however claim payments were lower than expected The reconciliation of our central estimates, excluding expenses to the estimates prepared by the previous scheme actuary as at 30 June 2012 show a release of reserves of $1.3 million, which is 17.5% of the opening estimates. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx April 2014

23 PwC Self-insurer outstanding claims liabilities 3.1 Outstanding claims liability Outstanding claims provision The provision below is based on cumulated claims data across all payment types and self-insurers. Projected payments are inflated then discounted to get to the gross central estimate before application of an allowance for claims handling expenses and a risk margin to the provision. We do not allow for any reinsurance recoveries for self-insurers. A breakdown of our results are shown in the table below: Estimates at 30 June 2013 ($000s) Accident year Gross o/s Reinsurance Net o/s Claims handling Net central Risk Net ending 30 Jun liability (a) recoveries (b) liability (c) expenses (d) estimate (e) margin (f) Provision (g) Total 6, , , ,757 Notes : (a) in inflated and discounted values (b) (a) x 0% (c) (d) (e) (a) + (b) assumed to be 7% of the net outstanding liability = (c) + (d) (f) a risk margin to increase the provision to a 75% level of sufficiency, = (d) x 14.54% (g) = (e) + (f) The inflated and discounted net provision at 30 June 2013 is $7.8 million, which is $0.6 million (7.3%) lower than the $8.4 million provision prepared by the previous scheme actuary as at 30 June We have used the same methodology as the previous scheme actuary for the self-insurer valuation however our adopted payment rates are lower. Further detail on the parameters adopted to calculate the outstanding claims can be found in Appendix F Comparison with self-insurers estimates We have compared our assessment of the net central estimate to self-insurers estimates. The results are shown in the table below: Estimates at 30 June 2013 ($000s) Accident year Self-insurers' ending 30 Jun estimate (a) PwC estimate (b) Difference ($) (b) - (a) Difference (%) (b) / (a) & earlier % % % % , % ,504 1, % ,255 2, % Total 5,566 6,772 1, % Notes : (a), (b) in inflated and discounted values, including claims handling expenses NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

24 Self-insurer outstanding claims liabilities PwC This comparison shows that our net central estimate is higher than the self-insurers estimate by $1.2 million (21.7%). Self-insurer funding ratio For self-insurers the funding ratio compares the self-insurers bank guarantee provision (the central estimate times 1.5) with our estimate excluding risk margin. This shows: Funding ratio ($000s) Self-insurers' provision (a) PwC central estimate (b) Difference ($) (b) - (a) Difference (%) (b) / (a) - 1 Total 8,349 6,772-1, % Notes: (a) bank guarantee provision, net central estimate (from table above) x 1.5 (b) as per table above net central estimate, excluding risk margin The aggregate funding ratio is 123%, which has increased from 112% last year. This increase is because the estimates we have calculated are lower than the previous scheme actuary s estimates last year. As the selfinsurers provision includes the 50% loading for the bank guarantee we would expect the funding ratio to be more than sufficient. 3.2 Claims statistics The following sub-sections show the claims experience by accident year. For more graphs of claims statistics, see Appendix G Number of claims incurred High increases in claims in 2011 appear to be reversing Number of claims incurred IBNR Claims reported Previous valuation Accident Year L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx June 2014

25 PwC Self-insurer outstanding claims liabilities The main points to highlight from this chart are: The number of claims incurred was stable from 2005 to 2009 at between 115 and 125 There were stong increases in the number of incurred claims in the 2010 and 2011 accident years, to 151 and 207 respectively. We are unaware of what drove this increase. Since the high in 2011 there have been reductions to 177 claims in 2012 and 139 claims in From a review the self-insurer reports we understand that one self-insurer has changed its management and recording of small claims which has contributed to the decrease Our estimate of the incurred number of claims is higher than that at the previous valuation for both the 2011 and 2012 accident years, very significantly for Gross average claim size Since 2006 the average claim size has fluctuated between $12,500 and $17,500, with 2013 in the middle of the range Average Claim Size ($000's) Average claim size in 30 June 2013 values Outstanding average claim size Paid average claim size Previous valuation Accident Year 2005 has a much higher claim size than other years due to more large claims at a higher average claim size than other years and 2011 have a lower average claim size than surrounding years which implies the increase in the number of claims related to smaller claims. Our projected average claim size for the 2009 to 2012 accident years is significantly lower than the previous scheme actuary s estimate at 30 June The uncertainty about the future development means that the ultimate level and our estimates may differ from that projected for recent accident years. This is especially true for the 2013 accident year where a high proportion (85%) of the average claim size consists of the uncertain future estimate. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

26 Self-insurer outstanding claims liabilities PwC 3.3 Actual vs expected claims experience over 2012/13 Actual experience compared to the expected experience over 2012/13 for claims incurred up to 30 June 2012 showed: claim reports were higher than expected (24 actual vs 18 expected) claim payments were lower than expected ($1.8 million actual vs $2.2 million expected). The expected experience is taken from the previous scheme report dated 30 April See appendix F for full details. The lower level of payments more than offsets the higher than expected claim reports. The impact of this favourable experience is quantified in the reconciliation below. 3.4 Reconciliation of central estimates The table below reconciles the gross outstanding claims central estimate, excluding expenses presented in 3.1 with the equivalent result as at 30 June Reconciliation of gross actuarial estimates, excluding expenses ($000s) Accident year ending 30 June ($000s) & earlier Total A. Gross estimates at 30 Jun 2012 (a) 2,287 2,194 1, ,189 B. Gross payments 1 July 2012 to 30 June , ,789 C. Assumed investment return (b) D. = A - B + C 1,195 2,073 1, ,589 Updated gross estimates at 30 June 2013 E. Revised gross estimates at 30 June 2013 (c) 1,515 1, ,328 F. = E - D ,261 Change 1 July 2012 to 30 June 2013 G. Proportion of change attributable to changes in real rates of return assumed H.= F - G Balance of change attributable to changes in underlying actuarial estimates I. Gross amount incurred and outstanding for 2, /13 accident year (C) J. = E + I Total gross outstanding liability, excluding expenses at 30 June ,329 Notes : (a) from Exhibit 25 of the previous scheme report dated 30 April 2013 (b) assumed to be 6.0% of claim payments in 2012/13 (c) (d) calculated using 3.0% pa being the one year forward rate from section 6 of the previous scheme report dated 30 April 2013 from appendix F4.4 of this report. The table shows that: there was a release of reserves of $1.3 million which is 17.5% of the opening 30 June 2012 estimates. This release is made up of a $0.3 million (4.1% of opening estimates) release due to changes in the real rates of return, and a $1.0 million (13.4% of opening estimates) release due to changes in the underlying actuarial estimates all accident years except 2012, 2008 and the 2005 and earlier accident years had releases in the overall outstanding claims estimate L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx June 2014

27 PwC Self-insurer outstanding claims liabilities these releases are due to lower than expected claim payments and adjustment in the adopted payments incurred. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

28 4 Break even premium rates Key points of this section The actual premium rates charged by insurers over the most recent three underwriting years is slightly higher than the hindsight break-even premium rate based on claims experience to date. With hindsight the actual premium rate charged was more than sufficient to cover the break-even cost for all accident years except 2009 and The break-even premium rate for the 2014 underwriting year is 2.2% of wages. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx April 2014

29 PwC Break even premium rates 4.1 Adequacy of past premiums The break-even premium rate is calculated on an inflated and discounted basis and is gross of reinsurance, as each insurer will have a unique treaty in place. No allowance for a profit margin has been made as insurers will set their own margin based on a multitude of factors. As we have allowed for the claims costs and expenses to be discounted to the point that the premium is received the calculated break-even premium rates have reduced since the previous scheme actuary s review which presented the rates on an inflated, undiscounted basis. The following table shows our calculated break-even premium rates and the actual rates charged by insurers: Calculated break even premium Actual premium Underwriting year Actual wages (a) ($000s) Discounted gross incurred cost (b) ($000s) Discounted Commission in other expenses financial year in the fin year (c) ($000s) (d) ($000s) Premium (e) ($000s) Calculated premium rate (f) Premium (g) ($000s) Actual premium rate charged (h) Difference (break even - actual) ,305,053 92,295 3,697 15, , % 122, % 11, ,704,972 85,421 2,864 14, , % 101, % -1, ,700,283 74,886 2,863 11,998 90, % 102, % 11, ,576,580 76,140 2,624 10,680 90, % 75, % -14, ,829,000 75,008 2,544 11,837 90, % 83, % -7, ,423,000 69,512 2,658 11,713 85, % 88, % 3, ,170,000 65,214 2,753 10,065 79, % 82, % 3, ,880,000 50,478 2,874 9,354 63, % 80, % 17, ,384,000 48,250 2,702 9,314 61, % 76, % 15,228 Notes : (a) from the consolidated ANZSIC data (b) (c) (d) (e) (f) (g) (h) calculated in Appendix H1 actual commission, from the consolidated Form A returns other expenses, from the consolidated Form A returns, discounted by half a year = (b) + (c) + (d) x (1+ one year historical interest rate) ^ (3/12) to allow for the fact that premiums are received 3 months after the commencement of the underwriting period = (e) / (a) from the consolidated ANZSIC data = (g) / (a) As the previous scheme actuary did for the previous valuation, we have used wages as a measure of exposure instead of the number of employees. The number of employees counts the full number of participating workers and will not reflect changes in the mix of full and part-time workers. Therefore we feel that wages gives a more accurate reflection of the actual exposure. We are concerned that the wage and premium information provided each financial year on the consolidated ANZSIC data include wage and/or premium adjustments that relate to prior underwriting periods. This will distort the comparison to prior years data, particularly when an economy is growing, as wage adjustments for previous years will artificially increasing the current year s wages. Also where the mix in burner policies is changing this could distort the level of premium received by year. Before the next valuation we will investigate this matter. This may result in changes to the wages and premiums values for each underwriting year. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

30 Break even premium rates PwC Adequacy of past premium rates 3.5% 3.0% 2.5% 3.2% 2.6% 2.8% 2.6% 2.6% 2.5% 2.5% 2.4% 2.5% Calculated break-even premium rate Actual premium rate charged 2.2% 2.2% 2.3% 2.0% 2.2% 2.2% 2.1% 1.9% 2.2% 2.1% 1.5% 1.0% 0.5% 0.0% Underwriting year The key points to highlight from the above chart and table may be summarised as follows: From 2005 to 2010 there was a decreasing trend in the actual premium rate charged by insurers, from 3.2% to 2.1% of wages Over the same period, 2005 to 2010, the hindsight break-even premium rate based on claims experience to date remained stable around 2.5%, except for a low of 2.2% in 2006 Since 2010 the actual premium rate charged by insurers has been stable to increasing, and was 2.3% in 2013 In contrast, the hindsight break-even rate decreased significantly between 2010 and 2011 to vary between 1.9% and 2.2% so that over this period it is equal to or less than the actual premium rate charged. With hindsight the actual premium rate charged was more than sufficient to cover the break-even cost for all accident years except 2009 and Historically the actual rate has fluctuated around the break-even premium rate, however we would expect that the premium charged by insurers to be consistently higher than our break-even premium as we have not included a profit margin whereas insurers would include a profit margin. Additionally we have calculated the break-even rate based on the gross risk cost, whereas the premium rate charged by insurers would factor in reinsurance (both recoveries and cost of a treaty) so this is another source of potential difference in the comparison as the reinsurance costs could be different to the gross risk cost. Using the difference in the actual premium and break even premium charged from the table above we have graphed the insurer performance on an accident year basis. The financial year basis comes from the Form A supplied by insurers. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx June 2014

31 PwC Break even premium rates Insurer performance by accident year and financial year 40,000 30,000 20,000 Profit/Loss ($000s) 10, , ,000-30,000 Accident year result Financial year result The insurer performance on a financial year basis has generally been more variable than the performance on an accident year basis. The negative financial year result in 2009 and 2010 coincides with a reduction in the average premium rates charged from 2.6% in 2008 to 2.2% in 2009 and 2.1% in The lower gross written premium in 2010 lead to a lower earned premium while payments and outstanding claims liabilities increased, leading to the reduction. On an accident year basis the 2009 and 2010 actual premium charged is lower than the hindsight break-even premium based on current claims experience. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

32 Break even premium rates PwC 4.2 Forecast break even premium rate The following table shows the break even premium rate projected for the next financial year. For comparative purposes we have also shown the last five underwriting years as well, which are the same as in the above table: Underwriting year Actual wages (a) ($000s) Notes: (a) 2014 = wageroll for 2013 x (1+3.75%) Discounted gross incurred cost (b) ($000s) (b) 2014 = adopted claims incurred x adopted average claim size in 30 June 2013 values x (1 + wage inflation) x (1 + superimposed inflation) x inflation/discounting factor 2,573 x 36,636 x ( %) x ( %) x Expenses (c) ($000s) Premium (d) ($000s) Calculated premium rate (e) ,503, ,112 19, , % ,305,053 92,295 18, , % ,704,972 85,421 16, , % ,700,283 74,886 14,861 90, % ,576,580 76,140 13,304 90, % ,829,000 75,008 14,381 90, % (c) = (b) / (1 commission rate (3.2%) other expense rate (12.9%) ) (b) (d) = (b) / (1 commission rate (3.2%) other expense rate (12.9%)) x (1 + interest rate (2.5%)) ^ (3/12) to allow for the fact that premiums are received 3 months after the commencement of the underwriting period (e) = (d) / (a) Our projections of the break-even rate for the 2014 underwriting year is reliant on three key items: Actual wages are forecast to increase at 3.75%, the adopted level of future wage inflation The future gross incurred cost is the product of the number of incurred claims and an average claim size, allowing for future inflation and superimposed inflation and an inflation/discount factor to allow for the timing of payments Expenses are the sum of commission and other expenses which are both set as a percentage of the total premium. Superimposed inflation is a measure of the growth in claims cost in excess of wage inflation. More detailed analysis on the derivation of the four adopted assumptions for the projection (incurred claim numbers, average claim size in 30 June 2013 values, commission rate and other expense rate) are shown in Appendix H. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx June 2014

33 5 Data and methods Key points of this section NT WorkSafe provided us with the data required for this review The data provided was the same as was used for the 30 June 2012 review by the previous scheme actuaries. We have suggested a number of enhancements to this data and additional data items which would improve the quality and add value to this review This section also describes the methodologies used to estimate the outstanding claims for insurers and selfinsurers, and for the break-even premium rate. NT WorkSafe Actuarial review of NT workers' compensation scheme as at 30 June

34 Data and methods PwC 5.1 Data provided NT WorkSafe supplied data to us from two sources: The internal WIMS database which records details of all claims lodged under the NT scheme Insurers own systems giving details of claims lodged. Since this was the first time we have performed this review NT WorkSafe provided the same data used by the previous scheme actuary for the current and each of the previous four financial years. Section 5.3 below gives details of ways in which the supplied data could be enhanced and additional data provided. We were supplied with the following reports from the WIMS system and consolidated data forms sent from insurers. All data was supplied in electronic form. Report 1 Reconciliation to Form B. NT WorkSafe s comparison of the payments and reports in the WIMS system and provided from insurers systems Report 2 Data based on date of accident. Unit claims data for all claims lodged by insurers and selfinsurers, with accidents grouped by financial year and presented in separate files. This data contained payment information by payment type and development year Report 3 Number of new claims received Report 4 History of payments based on injury date. Claim triangles for reports and payments for each insurer and in total Report 5 List of claims and insurers. Lists all claims since scheme inception by unique identification number and the insurer the claim was lodged with Consolidated Form A. A simplified profit and loss account showing only the insurance aspects, by insurer for the most recent financial year and in total for all insurers since 1999 Consolidated Form B. The number of claims reported and paid during the most recent financial year, and the number of active claims and the outstanding provision at the end of the most recent financial year, by accident year and split by insurer ANZSIC data. Policies, employees, premiums, wages and the average premium rate for each ANZSIC category. We were also provided with various forms completed by each insurer and self-insurer for the scheme, and with copies of valuation reports for some insurers and self-insurers. We have separated the data to only include the information for the five insurers and four self-insurers which are active in the scheme. These nine companies are: Allianz Australia Insurance Limited CGU Insurance Australia (Part of Insurance Australia Limited) GIO Insurance Australia QBE Insurance Australia Territory Insurance Office. L:\FRM\Insurance\GI\NT WorkSafe\P000\FY2014\Reports\ NT WorkSafe June 2013.docx June 2014