Grasp Your Actuarial Report In 15 Minutes Mujtaba Datoo, AS, MAAA, F Actuarial Practice Leader Aon Global Risk Consulting SEPTEMBER 13-16, 2016 SOUTH LAKE TAHOE,
Contextual Let s set up the context of the report so it can be grasped fully This session applies to P&C actuarial report for public entities There are other contexts for Pensions, Post Retirement, etc. 2
2 Time Perspectives Outstanding losses 2014 2015 December 2015 2016 2017 Projected funding 3
Time Perspective Liabilities estimate unpaid amounts at end of financial year, Or some other valuation date, say mid year They incorporate past events Loss forecasts project future costs Events have not occurred yet Pre funding losses 4
Conceptually 2 Actuarial Conclusions Liabilities are estimates of what is still left to be paid to claimants at close of fiscal year Claims have already occurred and can be reasonably estimated Needed for financial reporting Management's / Board s fiduciary responsibility Forecasts of next year Claims have NOT occurred yet For budgeting, basically pre funding costs 5
GASB Statement 10 Accrue liability for unpaid claim costs, including Incurred But Not Reported (IBNR) claims Liability should be based on: The estimated ultimate cost (including effects of inflation and other societal and economic factors) Using past experience adjusted for current trends Other factors that would modify past experience (GASB Statement No. 10, Paragraph 22) Accrue liability for claims adjustment expenses: ALAE expenses associated with specific claims (e.g. legal) ULAE expenses not associated with specific claims (e.g. claims administration) Discounting is neither mandated nor prohibited (GASB Statement No. 10, Paragraph 24) 6
Liabilities = Case Reserves + IBNR Claim adjusters (or staff) establish case reserve on individual open claims Actuaries estimate IBNR in aggregate 7
Who Owns the Liabilities? Adjusters set individual case reserves Actuaries estimate IBNR Actuaries provide central (expected) Auditors opine on fair representation of financial statement transactions Management records its best estimate Board owns the results! 8
Central Estimates Range of reasonable estimates Actuaries select central estimate estimate within a range of reasonable estimates based on several actuarial methods Also known as expected level at about 55% confidence level 55% based on skewed distribution of claims vs. symmetrical bell shaped distribution of 50% Reasonable estimates do not connote all possible outcomes 9
Actuarial Methods Actuaries use various standard methods to estimate ultimatecosts by year Select reasonable estimate based on underlying data, trends and judgment Sometimes non standard methods are applied 10
Standard Actuarial Methods Loss development Bornhuetter Ferguson analysis Frequency/severity Loss rate analysis Select ultimate losses based on appropriate methods and actuarial judgment 11
Key Actuarial Adjustments Loss development estimate ultimate costs today as if all claims were closed Trend brings costs to a common period, like CPI adjustment Adjust for varying SIRs and amount of losses retained Benefit changes WC, tort limits, etc. 12
Development Methods 100% 80% 60% 50% 67% 75% 60% 80% 65% 83% 70% 86% 75% 88% 78% 80% 82% 85% 90% 40% 40% 20% 0% 25% 20% 15% 5% 5% 5% 3% 3% 2% 2% 2% 0 12 24 36 48 60 72 84 96 108 120 132 Months of Development Incremental Paid Paid Reported 13
Estimated Ultimate Losses by Policy Year $40 $35 $30 $25 $20 $15 $10 $5 $0 to 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 5 3 1 3 2 1 1 2 1 1 0 22 20 22 15 17 17 14 15 3 5 3 6 3 3 2 24 26 20 20 7 4 18 10 6 14 17 13 6 Millions Paid Case Reserves IBNR 14
Claim Liabilities = Case Reserves + IBNR $40 $35 $30 $25 $20 $15 $10 $5 $0 to 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 $117.0 M (case reserves of $55.1 M + IBNR of $61.8 M) 5 1 1 0 1 2 1 1 3 3 2 3 5 3 3 2 6 3 7 10 4 6 17 13 Millions Case Reserves IBNR 15
Trends What is the frequency (number of claims per exposure measure) trend? What is the severity (average cost per claim) trend? What is the trend of large claims? What is the trend of large emerging claims, e.g. EPL? TIP: View graphically to quickly detect trends! 16
Distribution of Incurred Claims Below $100,000 2006 to 2015: about 7,000 non zero claims $100,000 $90,000 $80,000 $70,000 $60,000 $50,000 $40,000 $30,000 $20,000 $10,000 $0 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Jan-14 Jan-15 Jan-16 Incurred (Closed) Incurred (Open) 17
$1,000,000 $900,000 $800,000 $700,000 Incurred Claims Above $100,000 2006 to 2015: 35 claims $600,000 $500,000 SIR $400,000 $300,000 $200,000 $100,000 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Jan-14 Jan-15 Jan-16 Incurred (Closed) Incurred (Open) 18
Claims Histogram Use for modeling 21% Lognormal Distribution % of Observations 18% 15% 12% 9% 6% 3% 0% Claim Size 19
Volatility of Claims about the Average Distribution of Incurred Claims undeveloped, untrended, limited to $1M Loss Rate per FTE $800 $700 volatile # Claims % $$ % Below avg 6,237 88% $9.6M 18% Above avg 887 12% 43.2M 82% Total 7,124 100% $52.9M 100% Average: $8,600 $600 $500 $400 $300 $200 flat $100 $0 2006 2007 2008 2009 2010 Below avg 2011 2012 2013 2014 Above avg 2015 20
Trend, graphically actual 10% reform impact, moderating trend projection 21
Forecasting 2.00 1.80 1.60 1.40 1.69 1.49 1.541.57 1.84 1.46 1.39 1.39 1.41 1.26 1.25 1.31 Millions $40.0 $35.0 $30.0 $37,286,000 = $1.41 Loss Rate X $2.6 Billion Payroll 1.20 $25.0 1.00 0.80 0.60 0.40 $20.0 $15.0 $10.0 0.20 $5.0 0.00 $0.0 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 Historical (2015/16 level) Projected Latest 10-Yr Average Historical Future 22
Forecasting Future Years $40 $35 5 Millions $30 $25 $20 $15 $10 $5 $0 1 22 2 20 3 3 22 3 3 24 5 26 6 3 2 3 20 20 7 4 18 10 6 14 17 13 6 36 37 38 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 2017/18 Accident Year Paid Case Reserves IBNR Projected 23
Funding Components Add: Expected losses Investment income Risk margin Admin expenses Excess insurance premiums Rate: Divide funding amount by exposure units, say payroll per $100 excess premiums admin expenses risk margin investment income losses 24
Total Cost of Risk: TCoR Traditional definition Retained losses + overhead expenses + reinsurance As % of a base, say per $1,000 of revenue RIMS Survey TCoR 4% per $1,000 of revenue in 2014 Evolving definition of TCOR Includes costs for operation risks, cyber risks, etc. 25
Funding Allocation to Departments Balance between exposure and experience Balances responsiveness vs stability Experience modifiers (ex mod) provides deviation (credit or debit) from average rate Bets practices features: Length of experience period, 3 to 5 years Capping individual claims Credibility of loss experience Limiting swings, say changes to +/ 20% 26
Discounting Funds held to pay future claims earn investment income Investment income can be used to offset future payments of claims or expenses Discounted liabilities or funding based on: Payout pattern of losses (paid loss development factors) Investment yield (discount rate) of entity s assets 27
Financial Statement Perspective As of June 30, 2015 ASSETS ($ Millions) Cash and Investments $189.0 Receivables 0.5 Miscellaneous items 10.5 LIABILITIES AND NET ASSETS Liability for open and incurred but not reported (IBNR) claims $117 ULAE 5 Other 3 Total Liability 125 Net Position (Surplus) 75 Total Assets $200 Total Liabilities and net assets $200 28
Balance Sheet $Millions 20 18 16 14 12 10 8 6 4 2 0 Real Estate Receivables Cash Stocks Bonds Assets Other ULAE IBNR Case Reserves Liabilities Surplus 29
Risk, Liquidity and Solvency, Defined Risk: Deviation or variability around an expected outcome Liquidity: Ability to pay in cash short term liabilities, usually one year Solvency: Ability to pay all liabilities in the long run 30
Insurance Risks Reserve Risk Pricing Risk Asset Risk Fixed income, interest rate, equities Credit Risk Including reinsurance Operational Risk Catastrophe e.g. floods, earthquake 31
Spectrum of Financial Ratios Backward looking Forward looking IRIS RBC Scenario-Based Approach Probabilistic Models Simple Complex 32
Key Financial Ratios Focus on 6 key ratios: 1) Premiums (aka Contributions) to Surplus 2) Reserves to Surplus 3) Surplus to SIR 4) Reserve Development to Surplus 5) Operating Ratio 6) Liabilities to Liquid Assets 33
Confidence Levels: Measuring Think of percentiles, like SAT scores Variability is measured by: Standard deviation Simulating underlying claims process to mirror entity s experience, and then simulate thousands of such possible transactions to generate aggregate losses 34
Confidence Levels: Caution Measure of variability above average or expected level Beware: Amount of losses between expected and, say, 70% confidence level How many (large) claims will this cover? Actual losses may emerge higher! 35
Dilbert s Wisdom 36
Monte Carlo Simulation Use claim stratification information to simulate claims process Number of claims (frequency) Average claim size (severity) Select number of claims For each claim, select size 37
Monte Carlo Simulation Number of Claims 20% Uniform Distribution, i.e. equally likely to occur Poisson Distribution, Mean = 100 % of Observations 15% 10% 5% 0% 1 2 3 4 5 6 Number of Claims Number of Claims 38
Monte Carlo Simulation Claim Size 35% Lego Distribution Lognormal Distribution 30% 25% % of Claims 20% 15% 10% 5% 0% 1,000 2,000 3,000 4,000 5,000 6,000 Claim Size 100 500 2,500 10,000 50,000 100,000 250,000 Claim Size 39
Monte Carlo Simulation Example Simulated Claims Trial Number Claim Amount for Claim # Aggregate Number of Claims 1 2 3 4 5 6 Losses 1 2 $2,000 $4,000 $6,000 2 4 5,000 3,000 4,000 3,000 15,000 3 2 4,000 3,000 7,000 4 5 1,000 6,000 4,000 1,000 3,000 15,000 5 4 6,000 4,000 4,000 2,000 16,000 9,999 3 4,000 4,000 4,000 12,000 10,000 5 2,000 3,000 3,000 4,000 3,000 15,000 Avg 3.5 $12,000 40
Monte Carlo Simulation Example Ranked Simulated Claims Trial Number Aggregate Losses Rank Aggregate Losses Percentile Confidence Level Factor 1 $8,401,712 1 $1,020,320 2 1,497,651 2 1,024,065 3 3,516,291 3 1,029,627 4 1,797,246 5 2,870,778 6 4,187,925 7 1,029,627 9,999 1,954,018 10,000 2,509,543 5,000 4,025,944 7,000 5,353,140 9,000 8,849,910 10,000 155,734,676 Average $4,293,260 $5.35M / $4.3M = 70% 1.25 90% 2.06 $8.8M / $4.3M = 41
Monte Carlo Simulation Example Confidence Level % of Trials 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Expected level 70% confidence level 99.5% 1 in 200 year 1.6 1.9 2.2 2.5 2.8 3.2 3.7 4.3 5.4 6.0 7.3 8.8 12.2 19.4 Over Aggregate Losses ($Million) 42
43
Grasp Report By Question and Answer Approach Note: Answers here for illustrative use 44
Q: What is the scope of the report? Answer: Estimate liabilities and forecasts for workers compensation claims 45
Q: When are the liabilities valued as of? Answer: End of fiscal year June 30, 2015 46
Q: What are the liabilities? Answer: $117 million (case reserves of $60M + IBNR of $57M) 47
Q: Are the liabilities discounted? Answer: Yes, using 1% discount rate 48
Q: How do liabilities compare to last year? Answer: Say, increase of $5 million, or 4.5% Do atable of Reconciliation Are the changes material? Caused by emergence of large claims? 49
Q: What was the (loss) rate change? What was exposure change? Answer: Loss Rate change: 2015 = $1.38 per $100 of payroll 2016 = $1.41, an increase of 2% Exposure change: 2015 = $100M of payroll 2016 = $103M, an increase of 3% 50
Q: What are the cost drivers? Answer: Medical trend increasing 4% annually 51
Q: What are key changes from last year? Answer: Added new large member of JPA Multiple emergence of EPL claims 52
Q: Any program changes? Answer: SIR increased from $500K to $750k Added $250K corridor for law enforcement claims 53
Q: What is the financial position? What are the financial ratios? Answer: Financial position: Net position (surplus) of $75M. Financial ratios: Create exhibit of ratios over several years. Compare to benchmarks 54
Q: How are the claims emerging? Answer: Use scatter diagram: Large claims (say over 5 times the average) Look for recurring types of claims, claims from a specific member if in a pool, etc. Large claims follow about 80:20 rule 55
Q: Any data issue? Answer: Could not capture deductibles? 56
Q: Did claims exceed the SIRs? Answer: $$ excess recoverable from reinsurance 57
Estimates, Everywhere estimates Financial values are estimates They are the best estimates within the context and standards Actuaries estimate liabilities Investor advisers present fair value Accountants record value of assets, etc. Focus on material items 58
Reasonable Assumptions Connect the nine dots in a square grid with four straight lines without lifting the pen Used by management consultants to demonstrate the need to discard unwarranted assumptions. 59
Summary Understand the context of the report Focus on key findings Look for material changes Use graphs to highlight trends Ask if any major changes from previous year Emerging trends, large claims, etc. Don t sweat the details! 60
Questions? Thank you. Mujtaba Datoo, AS, MAAA, F Actuarial Practice Leader Aon Global Risk Consulting (949) 608-6332 mujtaba.datoo@aon.com 61
Thank You! 62