Back to basis Evolving technical matters

Transcription

1 Back to basis Evolving technical matters Savings and retirement products with guarantees: how to get a better return with lower risks? Prepared by Clement Bonnet Consulting Actuary Clement Bonnet Consulting Actuary Hong Kong T: E: clement.bonnet@milliman.com

2 Contents The problem so far 2 Development of a simple ALM framework 10 Development of an optimisation framework 17 Key takeaways 31 Disclaimer 34 1

3 2 The problem so far

4 The drivers for changes Low interest rate environment Competition Need for more sophisticated ALM techniques? Regulatory environment Product evolution Need for higher returns 3

5 10-year government bond yield Jan-11 Jun-11 Nov-11 Apr-12 Sep-12 Feb-13 Jul-13 Dec-13 May-14 Oct-14 Mar-15 Aug-15 Jan-16 Jun year government bond yield Jan-11 Jun-11 Nov-11 Apr-12 Sep-12 Feb-13 Jul-13 Dec-13 May-14 Oct-14 Mar-15 Aug-15 Jan-16 Jun year government bond yield Jan-11 Jun-11 Nov-11 Apr-12 Sep-12 Feb-13 Jul-13 Dec-13 May-14 Oct-14 Mar-15 Aug-15 Jan-16 Jun year government bond yield Jan-11 Jun-11 Nov-11 Apr-12 Sep-12 Feb-13 Jul-13 Dec-13 May-14 Oct-14 Mar-15 Aug-15 Jan-16 Jun year government bond yield Jan-11 Jun-11 Nov-11 Apr-12 Sep-12 Feb-13 Jul-13 Dec-13 May-14 Oct-14 Mar-15 Aug-15 Jan-16 Jun year government bond yield Jan-11 Jun-11 Nov-11 Apr-12 Sep-12 Feb-13 Jul-13 Dec-13 May-14 Oct-14 Mar-15 Aug-15 Jan-16 Jun-16 Low interest rate environment Source: investing.com 4

6 A generalised risk-based capital regime framework in the region Hong Kong (1) RBC 2017/2018 ICA / SST Solvency / Solvency II All EU Countries 2016 China (2) 2016 Switzerland 2006 Thailand (3) 2011 United Kingdom 2005 South Korea Malaysia The Philippines (4) Singapore Taiwan Indonesia RBC RBC (1) A phased implementation of new RBC framework is expected to happen in 2016/2017 (2) C-ROSS reporting has been introduced in 2016 (3) RBC 2 is on its way. QIS 1 results submission is scheduled in Q Vietnam India US Canada Solvency I NAIC RBC / MCCSR Before 2000 (4) New RBC rules are currently under discussions between the industry and the Regulator. 5

7 Product - Example: low interest rate and participating products (1/2) Reserving Current asset prices make it difficult to achieve historical target returns assumed in pricing and reserving Customers Producing an attractive guaranteed or illustrated return may be challenging Pricing Guarantees in products may be unsustainable or the products may be too risky if bonus earning capacity is reduced (higher time value of financial options and guarantees) Distribution Low nominal rates mean less room available to cover profits and expenses including distribution costs Distributors like old-style products 6

8 Product - Example: low interest rate and participating products (2/2) Is high(er) equity exposure an answer? How do we earn 5% when interest rates are at 2%? Higher return can only mean lower charges (or lower returns to the shareholder) or higher nominal investment returns Higher nominal returns might require higher equity exposure with an impact on risk management (capital requirements, ALM) and on risk adjusted profitability measures Further, greater riskiness might require a switch towards terminal bonuses. This may delay the emergence of shareholder profits too. Deferral of bonus equals deferral of shareholder dividend? Attractive needs equity exposure? High equity exposure demands lower guarantees and/or deferral of bonus? 7

9 How to get a better return with lower risks? One potential solution 8

10 How to get a better return with lower risks? or put in place a robust ALM framework based on a variety of simple and more complex actuarial technics that will allow you (i) to better understand your return and risk trade-off, (ii) to better manage your regulatory and internal capital requirement. Basic ALM framework Intermediate ALM Framework Advanced deterministic ALM framework Advanced stochastic ALM framework 9

11 Case study 1: Development of a basic ALM framework Illustration of an anticipated non participating whole life plan 10

12 ALM fundamentals Asset and liability management (ALM) is the practice of managing the performance and the risks that arise due to mismatches between the assets and liabilities. Liabilities are dictated by competition and customer constraints and cannot be as easily adjusted as the assets. Liabilities A-L interaction Assets ALM focuses on optimising the asset portfolio given the liabilities that have been sold 11

13 Net cash-flow Duration matching (1/2) Key product features: Anticipated whole life plan Pay premiums for first 5 years Cash payment every five years and sum-assured paid on death Policy is in-force and beyond payment term (now in policy year 9) Objective: hold an asset portfolio which has the same duration as liability duration. In this example, liability duration is 9.6 years, so match with a 9.6 year ZCB? Sensitivity of present value of cash flows to interest rates Liabilitiy cash flows Projection year Sensitivity to interest rates -1.5% -0.75% -0.25% -0.1% Central +0.1% +0.25% +0.75% +1.5% Assets (A) 54,583 50,961 48,964 48,036 47,602 47,173 46,538 44,487 41,595 Liabilities (B) 55,009 51,057 48,704 48,037 47,602 47,175 46,547 44,565 41,880 Net (=(A) (B)) (426) (96) (10) (2) 0 (2) (9) (79) (285) 12

14 Duration matching (2/2) Liability cash flows : Duration = 10 years! Asset cash flows: Asset strategy 1 : asset duration = 10 years Asset cash flow 1 (year 10) Duration = 10 years => No ALM risk Duration = 10 years x Asset strategy 2 : asset duration = 10 years Reinvestment risk Asset cash flow 1 (year 5) Liquidity risk Asset cash flow 2 (year 15) 13

15 Net cash-flow Duration and convexity matching Objective: hold an asset portfolio that has the same duration AND convexity as liabilities. Asset portfolio: Bond 1 Bond 2 Bond 3 Bond 4 Term Coupon 6.4% 6.0% 4.7% 4.6% Weight 33% 11% 9% 46% Present value of future cash flows: Sensitivity to interest rates -1% Central +1% Assets (A) 51,545 46,788 42,784 Liabilities (B) 51,545 46,788 42,784 Net (=(A) (B)) Reinvestment risk Liquidity risk Liabilities Assets Year

16 Net cash-flow Cash-flow matching (1/2) Objective: hold a portfolio of assets that will generate a stream of cash-flows that match the cashflows the insurer expect to need to pay out as liabilities (typically a portfolio of bonds that are constructed to generate the matching cash-flows). Year on year cash flow matching Liabilities Assets i -1% Central +1% Assets 51,545 46,788 42,784 Liabilities 51,545 46,788 42, Net Year The cash flow matching constraints can be on a year on year basis of on a cumulative basis Need to consider whether new business is included in the cash flow matching 15

17 Cash-flow matching (2/2) Pros Cons Remove liquidity risk Remove reinvestment risk Remove volatility in the economic balance sheet? Lack of liquidity of the financial markets in Asia Liability cash flows are not certain, and subject to lapse / insurance / expense risks Some type of liabilities may not be easy to match 16

18 Case study 2: Development of an optimisation framework (risk / return tradeoff) Illustration based on a multinational company selling primarily endowment participating policies 17

19 Optimisation process (1/3) The optimisation of the asset portfolio is generally based on some variation of the following process: Selection of performance indicators Selection of risk indicators Definition of constraints Calculation of risk / return measures for selected asset allocations Creation of an efficient frontier Selection of an asset allocation maximise return for a given level of risk minimise risk for a given level of return Iterative process Testing of selected asset allocation against constraints 18

20 Optimisation process (2/3) Visually this will appear as the following: Performance indicator (Present value of future profits calculated using best estimate assumptions) Minimisation of the risk indicator under a performance constraint Maximisation of the performance indicator under a risk constraint Legend: Initial asset allocation Asset allocation tested Efficient frontier Asset allocation in line with constraints Optimisation process Constraint 1 Constraint 2 Risk indicator (Present value of future profits calculated using stressed assumptions) 19

21 Optimisation process (3/3) The implementation of the ALM process requires the definition of the 3 key components Selection of performance indicators Selection of risk indicators Selection of key constraints It is important to note that there is no right answer to these questions. The most important is ensuring the buy in of management in the selection of these indicators. If the company has an internal risk appetite framework, this will be a significant contribution to how these elements will be defined. 20

22 Selection of performance / risk indicators There are a number of key issues to consider when selecting performance and risk indicators: Deterministic or stochastic valuation? How to define the risk indicators? How many years of future cash flows to take into account? In-force business only / in-force + new business? Which RDR to use? 21

23 Selection of key constraints Duration / Liquidity (duration, convexity, cash flow matching) Shareholder (accounting performance, economic performance, volatility) Regulation (e.g. capital requirement, limits on assets) Policyholder (e.g. savings return to customer, competitiveness) The different sets of constraints above will act as a maximum / minimum limits, as approved by the Board or management. They will restrict the optimisation process. These constraints are often directly linked to the Risk Appetite of the company. More specifically, the elements of the Risk Appetite that do not figure in the performance or risk measure should figure in the constraints. 22

24 Case study - Introduction The company has a large block of endowment participating policies that do not pay any terminal bonuses. Minimum guaranteed rates are very low for the block. The company has total discretion as to the level of bonuses that are paid. The aim of the analysis is to find the optimal portfolio of assets between the 4 key asset classes : Equity Property Fixed Rate bonds Floating rate bonds The process was split into four different steps: Step 1: Define framework Step 2: Develop asset and liability model Step 3: Produce key inputs Step 4: Measure profitability / risk indicators and constraints 23

25 Case study Step 1: Define the framework (1/2) Key indicators selected as part of the process: Measure KPI / KRI Stress Limits / constraints Value Present value of future profits 95 th percentile of stochastic scenarios Must be positive Earnings Accounting profit over first 5 years Average of scenarios 95 th percentile of stochastic scenarios Greater than previous year s profit Maximum loss of 50M Capital Solvency I capital 95 th percentile of stochastic scenarios CAR >100% for all years Guaranteed cashflows Fixed interest bonds No negative values Liquidity / cashflow matching A L gaps in first 10 years All asset classes Combined stress : increase 50% lapses and 50% of expected new business volumes No negative values 24

26 Case study Step 1: Define the framework (2/2) chastic scprojected Performance results over indicator a 30 year Average of the present value of future profits/losses over the 10 year projection period using real-world scenarios Risk indicator 95% percentile of the present value of the future profits/losses over the 10 year projection period using real world scenario. Average over all scenarios Discounting at the risk free rate Projection taking into account the risk premium on the different asset classes Discounting at the risk free rate Determination of the 95th percentile by ranking the scenarios 25

27 Case study Step 3: Produce key inputs One of the key inputs that was required to produce the analysis was a real-world economic scenario table. USD 10-year zero coupon bond 26

28 Performance: mean of the PVFP Case study Step 4: Key results (1/4) Efficient frontier: 95% Value at Risk (VaR) vs Mean of the Present value of future profits (PVFP) Equity = 55% Equity = 40% Equity = 30% Equity = 16% Equity = 14% Equity = 22% Equity = 12% Equity = 10% Equity = 20% Equity = 8% Equity = 25% Equity = 25% Property = 0% Initial Allocation: - Equity = 12% - Fixed Rate Bonds = 77% - Floating Rate Bonds = 5% - Property = 6% Property = 0% Equity = 0% Risk: 95% VaR of the PVFP 27

29 Mean of the PVFP Case study Step 4: Key results (2/4) Constraint 1: Risk measure > Equity = 22% Equity = 20% Equity = 16% Equity = 14% Equity = 12% Equity = 10% Equity = 8% Optimal Allocations - Equity = 8-16% - Fixed Rate Bonds = 63-71% - Floating Rate Bonds = 10% - Property = 11% Potential allocations: Efficient frontier: Equity between 8% and 16% - Property increased to 11% - Floating rate bonds increased to 10% Property = 0% Equity = 0% % VaR of the PVFP Equity = 16% although on efficient very little additional upside Initial Allocation: - Equity = 12% - Fixed Rate Bonds = 77% - Floating Rate Bonds = 5% - Property = 6% 28

30 Profit Millions Case study Step 4: Key results (3/4) Constraint 2: Average profit does not drop by more than 10% of the previous year profit 800 Average of annual profit Initial Initial 10_69_10_11 12_67_10_11 14_65_10_11 16_63_10_11 22_67_4_6 20_69_4_6 Potential asset allocations 300 Constraint Potential allocations: Equity allocation between 10% and 16%. 29

31 Case study Step 4: Key results (4/4) Constraint 3: Cash flow matching Guaranteed cash flows only + fixed income assets (equity = 14% total allocation) Total Asset Cash Flows Total Liability Cash Flows Asset Liability Mismatch Result: OK for equity = 14% 30

32 31 Key takeaways

33 Tangible competitive advantages Stability Efficiency Performance 32

34 The ALM journey WHERE ARE YOU? Basic ALM framework Duration / convexity matching Considering nature of liabilities for investment strategy Intermediate ALM Framework Pricing based on actual assets Cash flow matching/ analysis Stress testing Defining risk metrics Advanced deterministic ALM framework Optimisation Dynamic assetliability interactions Advanced stochastic ALM framework Stochastic modelling of assets Value at risk calculations Static / dynamic hedges Stochastic balancesheet projections 33

35 34 Disclaimer

36 Disclaimer This presentation is intended solely for educational purposes and presents information of a general nature. It is not intended to guide or determine any specific individual situation and persons should consult qualified professionals before taking specific actions. Neither the presenters, nor the presenters employer, shall have any responsibility or liability to any person or entity with respect to damages alleged to have been caused directly or indirectly by the content of this presentation. 35

37 Contact information Clement Bonnet Consulting Actuary Hong Kong E: T: