NAIC VA Reserve and Capital Reform Poojan Shah SOA Antitrust - PDF document

Equity-Based Insurance Guarantees Conference Nov. 5-6, 2018 Chicago, IL NAIC VA Reserve and Capital Reform Poojan Shah SOA Antitrust Compliance Guidelines SOA Presentation Disclaimer Sponsored by

NAIC VA Reserve and Capital Reform 2018 EBIG Conference November 6 th , 2018 (1030 – 1200 hours) Poojan Shah, FSA, CERA, CFA, MAAA

Background • The current AG43 (reserving) and C3P2 (capital) standards are complex and have several shortcomings.  Application of the standards across companies is inconsistent  Introduces significant volatility in funding requirement  Motivates the use of captives • The NAIC has commissioned an initiative to identify and propose changes to the current frameworks. The goal is to satisfy the following overarching objectives:  Ensure robustness of funding requirements  Promote sound risk management  Promote comparability across insurers and products  Preserve current statutory construct where feasible  Minimize implementation complexity 2

Current VA Framework Overview CTE 90 CTE 90 CTE 70 CTE 70 Adjusted Hedging Best Efforts Hedging Best Efforts Hedging Adjusted Hedging Weighted Weighted Avg. Avg. Min E-factor = 30% C3P2 C3P2 AG43 AG43 Stochastic Standard Scenario Standard Scenario Stochastic MAX • Assumptions MAX • RW Scenarios • Assumptions Reserves Total Asset Requirement • Methodology (TAR) MAX Volatile Total Funding Requirement 3

Recommendations: Standard Scenario Current Recommended + + PV Accumulated Net Basic Adj Rsv GPVAD Starting Assets Revenue (ANR) (AG33) • Allow company specified market paths • Prescribed market paths  Final Reported Reserve =  Initial Shock: -13.5% AG43; -20% C3P2 Stochastic Reserve + Additional Reserve  ANR is discounted at:  Additional Reserve (cannot be less than 0) =  Locked-in valuation rates (AG43)  10-year CMT + 50 bps; 3% floor (C3P2) Std. Scenario Amt. – CTE (70) Adjusted – Buffer • Non-guaranteed revenue sharing not projected • Same revenue sharing guidance as stochastic • Hedges run off over the first year • Model hedges identical as CTE adjusted • Current assumptions unchanged since • Behavioral assumptions aligned with industry framework creation experience • Aggregation not permitted in AG43 • Aggregated modeling 4

Recommendations: Stochastic Current Recommended • Use VM20 generator for interest rates and • No adequate guidance on generating interest separate account returns rates and separate account returns • Remove working reserves from GPVAD • Working reserves (CSV) are part of the GPVAD calculation calculation • Maximum allowable credit for CDHS: 95% • Maximum allowable credit for CDHS: 70% • Prescribed spread and default cost assumptions • Allowed to reflect companies’ own spread and for calculating net investment income on General default cost assumptions for calculating net Account assets investment income on General Account assets • Non-guaranteed revenue sharing linearly grades • Non-guaranteed revenue sharing for AG43 to 80% over 5 years linearly grades to 50% over 6 years  No cap  Cap = 0.25% of separate account value per annum after the sixth projection year 5

Proposed VA Framework Overview CTE 70 (Pre-Tax) CTE 70 (Pre-Tax) CTE 98 (Post-Tax) CTE 98 (Post-Tax) Best Efforts Adjusted: No Hedge Best Efforts Adjusted: No Hedge or No Rebalancing or No Rebalancing AG43 AG43 Stochastic CTE 98 Reported Additional Standard (Min of 5% weight to (Min of 5% weight to Projection Amount Adjusted) Adjusted) SUM Reserves C3 = 25% of (CTE 98 + Additional Standard Projection Amount x (1 – Tax) – Reserves) 6

Disclosure Requirements • Stochastic Scenarios – only for funds generated using a proprietary ESG  Disclose Sharpe ratio and correlations • Best Efforts Hedging (CDHS)  Compare CDHS modeling vs. actual experience over past 12-36 months  Explicit hedging: Project hedges along realized market path, and compare to actual experience  Implicit hedging: Compare actual experience with fair value movements in hedged liability  Low error factor permitted only if the modeled hedge performance is close to actual  Position of Best Efforts CTE relative to fair value (FAS 133) and unhedged CTE  Necessitate further discussions with regulators if BE CTE < min (FV, unhedged CTE) • Standard Scenario  Compare projected cumulative decrement pattern using company vs. prescribed assumptions  How do mortality, annuitization, lapses, etc. contribute to AV run off over the projection path?  Under CTE approach for determining standard scenario, analyze average of worst 30% of scenarios 7

Modeling/Operational Challenges • Standard Scenario behavioral assumptions  Modeling Challenges:  Moneyness based on GAPV  Withdrawal delay cohorting  Attribution analysis to understand impact of each assumption change • Stochastic scenario generation  Map all funds to a combination of funds from the VM-20 generator  Allow usage of proprietary generator only if it does not materially reduce TAR  Disclose Sharpe ratio and correlation • Companies reflecting CDHS must make further disclosures  Actual vs. Projected Best Efforts hedge gain/loss  Position of Best Efforts CTE relative to fair value and unhedged CTE 8