Session 70, PBR, VM 20, AG 48, and Investment Strategy: Are Changes Ahead? Moderator: Alan J. Routhenstein, FSA, MAAA

Transcription

1 Session 70, PBR, VM 20, AG 48, and Investment Strategy: Are Changes Ahead? Moderator: Alan J. Routhenstein, FSA, MAAA Presenter: Jason E. Kehrberg, FSA, MAAA Alexandre Lemieux, FSA, MAAA Alan J. Routhenstein, FSA, MAAA

2 2016 Valuation Actuary Symposium Session 70: PBR, VM-20, AG 48 and Investment Strategy Are Changes Ahead? Modeling Assets Under VM-20 Tuesday, August 30, 2016 Jason Kehrberg, FSA, MAAA Vice President, PolySystems Inc.

3 Modeling Assets Under VM-20 Agenda Overview and Starting Assets Modeling Existing Assets and Reinvestments Prescribed Default Cost Methodology 2

4 Overview and Starting Assets 3

5 What type of asset models may be required for VM-20? Net Premium Reserve (NPR) No asset model is needed since asset returns are prescribed, not modeled Similar for the deterministic reserve exclusion test Deterministic reserve (DR) Gross Premium Valuation Method Uses net asset earned rates (net investment earnings / invested assets) to discount Direct Iteration Method Discount rates not required, but need fully integrated asset/liability model to project surplus Stochastic Reserve (SR) Need fully integrated asset/liability model to project surplus Discount rates equal 105% of one-year Treasury rates 4

6 How many asset models may be required for VM-20? You will likely need a separate asset cash flow model for each model segment. A model segment is defined as a group of policies and associated assets that are modeled together to determine the path of net asset earned rates. Model segments are to be established consistent with company s asset segmentation plan, investment strategies, or approach to establish statutory investment income allocation. 5

7 The prescribed Economic Scenario Generator (ESG) When calculating the DR, SR, and SRET ratio, economic scenarios are used to model investment income. VM-20 requires use of the prescribed ESG Generates 16 scenarios for the SRET ratio (number 12 for the DR) and the full set of 10,000 scenarios Included representative scenario picking tool (based on 20-year Treasury) can be used to generate representative scenario subsets of 1,000, 500, 200, or 50 scenarios 10 Treasury rates: 90 days to 30 years 9 total return indices: money market, gov bonds, corp bonds, fixed income, balanced, large cap equity, global equity, small cap equity, aggressive equity Mapping proxy fund returns to the prescribed fund returns is left to actuarial judgment, but must be consistent with the prescribed returns. 6

8 Make up of starting assets for a model segment The following are included in starting assets: Due and accrued or unearned investment income All separate account assets supporting the policies All policy loans supporting the policies All derivative instruments supporting the policies General account assets to target level of starting assets Must be consistent from one period to the next For asset segments that back both VM-20 and non-vm- 20 business, need equitable method to select the subset of assets for VM-20, e.g. pro-rata, based on asset assignment for credited rates 7

9 Level of starting assets for a model segment Initially set equal to estimate of final modeled reserve (max DR, SR) and PIMR balance on valuation date. Can use reserve history, previously observed relationships to NPR, AV, or CSV adjusted for the economic environment, and inforce changes, etc. For all model segments combined, if starting assets < 98% of the modeled reserve, or > max(npr, 102% of modeled reserve), then document modeled reserve is not materially understated, or iterate to within collar. Iterate if not within 2% of the aggregate modeled reserve or demonstrate that the modeled reserve produced is not materially understated. 8

10 Modeling Existing Assets and Reinvestments 9

11 Modeling fixed income assets Project for each scenario: Gross investment income and principal repayments for each economic scenario, per contractual provisions for existing assets, or at prescribed gross spread to Treasuries for reinvestments. Prescribed default costs and prudent estimate investment expenses. Capital gains and losses on asset sales. The exercise of embedded options such as calls, puts, prepayments and extensions. Statement values, including accruals. Grouping is allowed if you can show it won t materially understate the reserve 10

12 Modeling equity Allocate equity assets to chosen proxy funds (e.g., large cap stocks, real estate), often expressed as a linear combination of the recognized market indices available in the prescribed ESG Project future gross investment returns for each proxy fund, subtracting prudent estimate investment expenses, and including realized and unrealized capital gains/losses Model associated asset sales and derivatives Process is same for GA and SA equity assets Similar to AG43 and C3P2 but with fully prescribed VM- 20 ESG 11

13 Modeling existing derivatives Must model all benefits, costs, residual risks, and frictional costs associated with all relevant existing derivatives. For future derivatives if there is a CDHS (VM-20 encourages future consideration of graded approach for recognizing the impact of non- CDHS derivatives). Each derivative program must be classified as associated with assets or with liabilities for purposes of calculating reserves, esp. the deterministic reserve. If derivative cost and payoff generate relatively smooth increments to the net asset earned rate (more bond-like ), that s a good candidate for asset treatment. Otherwise, liability treatment may be more appropriate. For the DR, asset-associated derivatives are included in the net asset earned rate, whereas liability-associated derivatives are included in the liability cash flows. 12

14 Modeling future derivatives Model all benefits, costs, residual risks, and frictional costs associated with future CDHS derivatives CDHS requirements are similar to AG43 and C3 Phase II Examples of CDHS associated with VM-20 products Hedging for equity indexed products Hedging variable life long-term guarantees Rebalancing portfolio using derivatives to meet ALM targets Reflection of a qualified CDHS in the model may either increase or decrease the reserve Not reflecting a hedging program very similar to a qualified CDHS can also increase or decrease the reserve relative to what would be held if the program qualified for a CDHS The direction of such impact can vary with economic conditions Guidance Note encourages future consideration of a graded approach for recognizing the impact of non-qualifying hedging strategies 13

15 Other asset modeling considerations Pre-tax IMR (PIMR) does not need to be modeled for VM-20, simply reduce/increase the modeled reserve (DR or SR) by the positive/negative PIMR allocated to the model segment as of the valuation date. Policy Loans Model explicitly or substitute other assets as a proxy. Proxy assets Treat cash flows as asset cash flows and demonstrate the substitution does not lower the reserve. Explicitly model all policyholder loan behavior Treat cashflows as liability cashflows, e.g. starting loan balance and additional loan principal as outflow, interest and principal paid in cash as inflow, capitalized interest ignored. Regardless of which approach is used, the policy loan activity must reflect the company s expected utilization of policy loans by policyholders. Disinvestment strategy used to fund negative net cash flows must be consistent with investment policy. Spreads used to determine market values of assets sold must be consistent with prescribed spreads for asset purchases. Borrowing and negative assets are not prohibited techniques. 14

16 Modeling reinvestments Reinvestment asset gross returns are determined from the bottom up by adding gross spreads to Treasuries (unlike existing asset gross returns which are determined according to contractual provisions). For public non-callable bonds gross spreads are prescribed and grade from current to ultimate over three years. For other assets gross spreads are not prescribed. Prescribed reinvestment defaults are similar to existing assets but do not include the max net spread adj factor. Investment strategy must be consistent with actual investment policy and not produce a lower minimum reserve than would result using 50/50 mix of A/AA. Must provide documentation demonstrating compliance with this minimum floor requirement. 15

17 Default Cost Methodology 16

18 Defaults are prescribed under VM-20 Objectives provided by LATF: Default costs for the same or similar assets should be the same across companies. Companies should not be able to lower reserve by investing in riskier assets beyond some threshold. In the short term, default costs should reflect current economic conditions and grade into historic conditions over the longer term. The method should be relatively simple. 17

19 Inputs and tables used to determine defaults if NAIC designation exists Company determined inputs for each asset Investment expense Option adjusted spread (OAS) The average spread over Treasuries that equates a bond s market price as of the valuation date with its modeled cash flows Weighted average life (WAL) The weighted average number of years from the valuation date until 100% of the outstanding principal is expected to be repaid Ratings NAIC designations and ratings from Approved Ratings Organizations (ARO ratings) Prescribed VM-20 tables Table A Baseline defaults Tables F,G,H and I Current and ultimate spreads Table K Ratings conversion table Table J Current and ultimate swap spreads 18

20 PBR credit rating Use Table K to convert ratings to a VM-20 PBR numeric credit rating of 1 to 21. For an asset with an NAIC designation that is derived solely by reference to underlying ARO ratings without adjustment, the PBR credit rating is the average of the numeric ratings for each available ARO rating, rounded to nearest integer. For an asset with an NAIC designation that is not derived solely by reference to underlying ARO ratings without adjustment, the PBR credit rating is the second least favorable numeric rating associated with that NAIC designation. 19

21 Defaults on assets without an NAIC designation Prescribed defaults are determined differently if the asset has no NAIC designation Includes but is not limited to residential mortgages, i.e., whole loans Defaults are determined such that the net yield is capped at 104% of the Treasury rate most closely coinciding with the asset s purchase date and maturity structure, plus 25 basis points 20

22 Table K PBR Bond Ratings Moody S&P Fitch DBRS RealPoint AM Best NAIC NAIC Comm Mtg PBR Aaa AAA AAA AAA AAA aaa 1 Aa1 AA+ AA+ AA high AA+ aa+ 2 Aa2 AA AA AA AA aa 3 Aa3 AA- AA- AA low AA- aa- 4 A1 A+ A+ A high A+ a+ 5 A2 A A A A a 1 6 A3 A- A- A low A- a- 1 7 Baa1 BBB+ BBB+ BBB high BBB+ bbb+ 8 Baa2 BBB BBB BBB BBB bbb 2 9 Baa3 BBB- BBB- BBB low BBB- bbb Ba1 BB+ BB+ BB high BB+ bb Ba2 BB BB BB BB bb Ba3 BB- BB- BB low BB- bb B1 B+ B+ B high B+ b+ 14 B2 B B B B b 4 15 B3 B- B- B low B- b- 16 Caa1 CCC+ CCC+ CCC high CCC+ ccc+ 17 Caa2 CCC CCC CCC CCC ccc 5 18 Caa3 CCC- CCC- CCC low CCC- ccc- 19 Ca CC CC CC CC cc

23 If NAIC designation exists, defaults are the sum of three components Baseline default cost Table lookup using PBR rating (1-20) and WAL (1-10) Prescribed table is based on historical averages + margin Spread-related factor Based on the prescribed current and historical spreads that vary by PBR rating (1-20) and WAL (1-30) Factor has a max, min and grades to zero over 3 years Maximum net spread adjustment factor Only if portfolio net spread is greater than threshold Grades to zero over three years 22

24 Questions? Jason Kehrberg, FSA, MAAA Vice President, PolySystems Inc. (312)

25 Session 70 PBR, VM-20, AG 48 and Investment Strategy: Are Changes Ahead? 2016 Valuation Actuary Symposium Alex Lemieux August 30, 2016

26 Investment Strategy Discussion Topics Portfolio Guidance Portfolio Segments Long vs Short Tenures Analysis of Results PwC 2

27 VM-20 Guidance The valuation manual provides guidance on model grouping VM-20 ( the manual ) specifies that modeling should be done at the model segment level The company should define an appropriate model segment consistent with the following: 1. Asset segmentation plan 2. Investment strategy 3. Approach used to allocate investment income 4. Risk management strategy PwC 3

28 Investment Portfolio Implications Two potential approaches to portfolio management in a post-pbr world One Portfolio Simpler Less investment accounting effort May have to notionally allocate assets to individual products May complicate credited rate determination Portfolio by Product Requires a process to allocate assets More investment accounting effort to maintain Exacerbated by multiple products over multiple issue years Earned interest rate is observable by portfolio Credited interest rate process should be easier PwC 4

29 One Portfolio Approach Are calculated reserves better represented by one aggregate portfolio? The use of one investment portfolio by a company is supported by: Much easier to administer Better represents the pooled risks of the company Consider the respective and combined reserves for a portfolio of universal life products: One portfolio of single premium universal life policies One portfolio of fixed annual premium accumulation universal life policies Both portfolios use the same 10 year A-rated bond strategy A summary of the time 0 and time 5 stochastic reserves is presented below: UL SPUL Sum Grouped Year Million 12.3 Million 12.7 Million 12.3 Million Year Million 11.8 Million 15.0 Million 13.6 Million PwC 5

30 Finding ways to depict the range of scenarios that contribute to the calculated reserve provides useful information 0.25 Range of scenarios PwC 6

31 Illustration of 5 worst scenarios UL Scenarios Worst UL Scenarios PwC 7

32 Illustration of 5 worst scenarios SPUL Scenarios Worst SPUL Scenarios PwC 8

33 One Portfolio Approach Can different interest rate scenarios happen at the same time? Uncombined stochastic analysis causes different scenarios to be included in the reserves for different products, implying multiple different futures. Case in point, the 5 worst scenarios for the time 0 reserves by product are as follows: Scenario SPUL Scenario # UL ,756, , ,787, , ,178, , ,316, , ,650, ,936 Excluding scenario 993, no two scenarios have overlap. The one portfolio approach uses the same worst scenarios to determine the policy reserves for the company. PwC 9

34 Investment Strategy How will different portfolio postures affect reserves under PBR? Reserves under PBR will reflect actual portfolio holdings Assumptions are made as to how future cash flows will be invested How will stochastic reserves react to different investment postures under: A level interest rate scenario A decreasing interest rate scenario An increasing interest rate scenario PwC 10

35 ULSG stochastic reserves under an increasing interest rate scenario From Investment Considerations for Principle-Based Approaches for ULSG Reserves, Risks and Rewards, February 2016 PwC 11

36 PBR example for a Single Premium Universal Life product As companies consider the impact of stochastic reserving and investment strategies, they should give careful consideration to: How will crediting rates be managed in a rising rate environment? How will floor rates be managed in a low rate environment? We prepared an example using 1000 scenarios for a SPUL product The product is exposed to dynamic lapses in a rising rate environment and to floor rates in a down scenario The product was modeled with 1, 5, 10, 15 and 20 year tenure investment strategies, assuming an average single A credit rating PwC 12

37 Result #1: CTE levels of time 0 reserves using representative investment strategies 17,000,000 16,000,000 CTE's for different investment strategies Short strategy offers lower reserve; at a cost to profitability 15,000,000 14,000,000 13,000,000 12,000,000 11,000,000 10,000,000 Average CTE30 CTE50 CTE70 Max 1 Year Strategy 5 Year strategy 10 Year strategy 15 Year Strategy 20 Year Strategy PwC 13

38 Result #2: CTE levels of time 5 reserves using representative investment strategies Note much of the business with the 1 year strategy has lapsed by year 5 18,000,000 Year 5 Reserves by Investment Strategy Impact of dynamic lapses more pronounced as block ages 16,000,000 14,000,000 12,000,000 10,000,000 8,000,000 6,000,000 4,000,000 2,000,000 - Average CTE30 CTE50 CTE70 Max 1 Year Strategy 5 Year strategy 10 Year strategy 15 Year Strategy 20 Year Strategy PwC 14

39 Graph of SPUL reserves in year 5 per remaining policy Note the change in axis from previous graphs 18,000 SPUL Reserves per policy 17,000 16,000 15,000 14,000 13,000 12,000 11,000 10,000 Average CTE30 CTE50 CTE70 Max 1 Year Strategy 5 Year Strategy 10 Year strategy 15 Year Strategy 20 Year Strategy PwC 15

40 Profitability must be considered in addition to reserve volatility Using a pricing curve equal to the forward curve; profitability becomes an issue when adopting the shorter strategies: 1 Year 5 Year 10 Year 15 Year 20 Year Strategy Strategy Strategy Strategy Strategy PVFP* (2) Million 1 Million 2 Million 3 Million 4 Million The desired reserve stability of the 1 year strategy is offset with an expected loss on the product due to negative investment margins. Note, the scenarios that triggered the greatest reserves were not the same by strategy PwC 16

41 Investment Strategy Are changes ahead? Will calculated reserves better reflect the underlying asset portfolio? Will the underlying asset portfolio better reflect the policy liabilities? Yes, the degree of which will be impacted by: Sophistication of model to portray investment strategies, credited interest rate strategies, and dynamic experience variables over a range of future economic scenarios Considering risk between the alternative investment strategies Incorporating profitability into the full evaluation PwC 17

42 Thank you Alexandre Lemieux Director (312) This publication has been prepared for general guidance on matters of interest only, and does not constitute professional advice. You should not act upon the information contained in this publication without obtaining specific professional advice. No representation or warranty (express or implied) is given as to the accuracy or completeness of the information contained in this publication, and, to the extent permitted by law, PricewaterhouseCoopers LLP, its members, employees and agents do not accept or assume any liability, responsibility or duty of care for any consequences of you or anyone else acting, or refraining to act, in reliance on the information contained in this publication or for any decision based on it PricewaterhouseCoopers LLP. All rights reserved. In this document, PwC refers to PricewaterhouseCoopers LLP, a Delaware limited liability partnership, which is a member of PricewaterhouseCoopers International Limited, each member firm of which is a separate legal entity.

43 Term Life ( TL ) and ULSG ( UL ) Reserve Financing ALM Case Study Session 70 PD PBR, VM 20, AG 48 and Investment Strategy 2016 SOA Valuation Actuary Symposium Presented by Alan Routhenstein, FSA, MAAA August 30, 2016

44 Table of Contents Introduction to TL & UL Reserve Financing and AG48 TL & UL Base Case AG48 Liabilities and Assets TL & UL ALM Strategy Sensitivities & Stress Scenarios TL & UL ALM Observations for AG48 Covered Policies TL & UL ALM Considerations for Financing of PBR Policies 2

45 Many banks and reinsurers are willing to finance the excess of statutory over economic reserves The definition of economic reserves has historically been a negotiated deal term that for XXX/AXXX financing deals has often been GPV determined with best estimate assumptions. The graph below shows how TL XXX, economic and excess reserves can vary over time % 100.0% 80.0% 60.0% 40.0% Statutory Excess Economic 20.0% 0.0% New transactions financing policies not grandfathered under Actuarial Guideline 48 ( AG48 ) involve an Actuarial Method reserve ( AMR ). 3

46 TL AMR is often closer to best estimate economic reserves than to statutory reserves For TL, the AMR is a modified VM-20 reserve equal to the larger of a net premium reserve ( NPR ) and deterministic reserve (DR) For TL, the AMR generally peaks at 30% to 50% of statutory reserves 120.0% 100.0% 80.0% 60.0% 40.0% Statutory AMR Economic 20.0% 0.0% That compares with best estimate economic reserves that generally peak at 20% to 40% of statutory 4

47 UL AMR Relative to Statutory and Economic Reserves For universal life with secondary guarantees ( UL ), the AMR is a modified VM-20 reserve equal to the largest of a net premium reserve, a deterministic reserve and a stochastic reserve For UL, the relationship between statutory reserves, the AMR and economic reserves is highly dependent on policy design, the nature of the reinsurance treaty, the degree of asset/liability matching and the definition of economic reserves For UL blocks of policies that have been financed, best estimate economic reserves have generally peaked at 55% to 85% of statutory For UL AMR our sample size is still too small to generalize 5

48 For AG48 Covered Policies, the Cedant must back the AMR with Assets that are Primary Security Primary Security is defined to include: Cash meeting the requirements of Section 3.A. of Model 785 SVO-listed securities meeting the requirements of Section 3.B. of Model 785, but excluding any synthetic letter of credit, contingent note, credit-linked note or other similar security that operates in a manner similar to a letter of credit Primary Security And also, for funds-withheld and modified coinsurance reinsurance arrangements: Commercial loans of good quality and higher; policy loans; and derivatives used to support and hedge liabilities pertaining to the actual risks in the policies ceded.

49 In AG48, assets financing the excess of statutory reserves over the AMR are called Other Security Other Security assets need not qualify as NAIC admitted assets, but for each transaction they are subject to cedant and captive regulatory approval Other Security Primary Security

50 It is possible to also finance a portion of Primary Security, with a mezzanine layer of financing The mezzanine layer size is the excess of AMR over economic reserves Such mezzanine tranche must consist of Primary Security in a form approved by the cedant s regulator Other Security Mezzanine Tranche Economic Assets

51 Table of Contents Introduction to TL & UL Reserve Financing and AG48 TL & UL Base Case AG48 Liabilities and Assets TL & UL ALM Strategy Sensitivities & Stress Scenarios TL & UL ALM Observations for AG48 Covered Policies TL & UL ALM Considerations for Financing of PBR Policies 9

52 First, I want to thank two of my colleagues who prepared the actuarial models for our case study Elisha Hack, Little Falls, NJ Andrew Steenman, Indianapolis, IN 10

53 For our case study, the illustrative TL & UL Covered Policies have the following characteristics TL Level Term Period ( LTP ) mix includes 10, 20 and 30 year UL product is a shadow account product Gender, issue age, and underwriting class are consistent with industry averages TL and UL Covered Policies are closed blocks that do not include future new business 11

54 For our case study, the illustrative existing assets have a less risky than average ratings distribution Weighted average life ( WAL ): 21 years Weighted average PBR Credit Rating: 6 (A2/A) Distribution by PBR Credit Rating: Rating Distribution 3 (Aa2/AA) 19% 6 (A2/A) 65% 9 (Baa2/BBB) 16% 12 (Ba2/BB) 0% Weighted average book yield: 4.1% 12

55 For our case study, the illustrative reinvestment strategy has a more typical ratings distribution WAL: 20 years Weighted average PBR Credit Rating: 7.2 (A3/A-) Distribution by PBR Credit Rating: Rating Distribution 3 (Aa2/AA) 0% 6 (A2/A) 65% 9 (Baa2/BBB) 30% 12 (Ba2/BB) 5% 13

56 Our projections include an inner loop and an outer loop Inner loop is used to develop a robust DR calculation (and SR, for UL) at each node using forecasted in-force liabilities and assets, and estimates of future assumptions and economic parameters In the base case, using VM-20 assumptions (i.e. prescribed economic scenarios, default costs and public bullet bond credit spreads; and best estimate assumptions plus margins for mortality, lapse and expenses) Outer loop is used to forecast cash flows and inforce at each node In the base case, using best estimate assumptions 14

57 Our case study includes a number of simplifying assumptions for the outer loop, including Liabilities YRT reinsurance premiums assumed to increase annually by explicit and implicit mortality margins Mortality is assumed to be updated annually to reflect anticipated mortality improvement TL lapse rates at the end of the LTP are assumed to be 100% UL crediting rates utilize an illustrative portfolio crediting rate strategy UL dynamic lapse behavior linked to NLG in-the-money-ness UL level premium pattern to attained age 110 Assets Modeling of structured securities as corporates with same WAL & OAS UL default rates are VM-20 rather than best estimate 15

58 Our case study includes a number of simplifying liability assumptions for the inner loop, including TL and UL Liabilities YRT reinsurance premiums assumed to increase by explicit and implicit mortality margins Assume no future changes to credibility or sufficiency data period TL Liabilities TL lapse rates at the end of the LTP are assumed to be 100% No future node starting assets true-up (at 98% to 102% of reserves) UL Liabilities UL crediting rates & dynamic lapse behavior have simplified margins 200 scenarios are used to estimate the SR at every fifth node 16

59 Our case study includes a number of simplifying asset assumptions for the inner loop, including Assets Modeling of structured securities as corporates with same WAL & OAS Long-Term VM-20 spread tables are assumed to not change Current VM-20 spread tables are assumed to equal to Long-Term Spread-related factor for default costs is assumed to be 0 Maximum net spread adjustment factor is assumed never applicable 17

60 In the base case scenario, the TL DR is estimated to exceed the NPR for most of the projection 18

61 Likewise, in the base case, the UL DR & SR are estimated to exceed NPR for most of the projection 19

62 In the base case, the TL AMR is closer to economic than statutory reserves for most of the projection 20

63 In the base case, the UL AMR is close to the average of economic and statutory reserves for most of the projection 21

64 In the base case, the TL Other Security peaks at about 82% of the Excess Reserves peak 22

65 In the base case, this case study s UL Other Security peaks at about 68% of the Excess Reserves peak 23

66 Table of Contents Introduction to TL & UL Reserve Financing and AG48 TL & UL Base Case AG48 Liabilities and Assets TL & UL ALM Strategy Sensitivities & Stress Scenarios TL & UL ALM Observations for AG48 Covered Policies TL & UL ALM Considerations for Financing of PBR Policies 24

67 We analyze two yield curve shift stress scenarios Best estimate scenario bps Best estimate scenario bps 25

68 We ve analyzed one alternative ALM strategy WAL + 5 Existing assets: The relative sizes of most existing assets were modified to increase the portfolio WAL by 5 years Reinvestment assets: The allocation between 10y and 30y reinvestments was modified to increase by 5 years the WAL of each reinvestment cohort 26

69 Table of Contents Introduction to TL & UL Reserve Financing and AG48 TL & UL Base Case AG48 Liabilities and Assets TL & UL ALM Strategy Sensitivities & Stress Scenarios TL & UL ALM Observations for AG48 Covered Policies TL & UL ALM Considerations for Financing of PBR Policies 27

70 TL AMR in this case study is sensitive to yield curve shifts 28

71 UL AMR in this case study is even more sensitive to yield curve shifts 29

72 TL Other Security is sensitive to yield curve shifts 30

73 UL Other Security is even more sensitive to yield curve shifts 31

74 If the TL financing has a mezzanine tranche, its size is sensitive to yield curve shifts 32

75 If the UL financing has a mezzanine tranche, its size is even more sensitive to yield curve shifts 33

76 Table of Contents Introduction to TL & UL Reserve Financing and AG48 TL & UL Base Case AG48 Liabilities and Assets TL & UL ALM Strategy Sensitivities & Stress Scenarios TL & UL ALM Observations for AG48 Covered Policies TL & UL ALM Considerations for Financing of PBR Policies 34

77 After a company adopts PBR for TL, financing of VM- 20 less economic reserves is like a mezz. tranche 35

78 After a company adopts PBR for UL, financing of VM- 20 less economic reserves is like a mezz. tranche 36

79 If the company is a small company in VM-20, TL Other Sec. would be sensitive to yield curve shifts 37

80 If the company is a small company in VM-20, UL OS would be even more sensitive to yield curve shifts 38

81 Milliman Contact Information Alan Routhenstein