Log in / My SOA

RILA GLWB Designs and Market Risk Analysis

By Matt Heaphy, Nicholas Carbo, and David Elliott

Product Matters!, June 2023

pm-2023-06-heaphy-hero.jpg

Introduction

This article assumes an understanding of Registered Index-Linked Annuity (RILA) and Fixed Index Annuity (FIA) products. An introduction to these products can be found in “The Rise of Registered Index-Linked Annuity Products” from the August 2022 issue of Product Matters!

RILA sales have been growing at a 47% compound annual growth rate (CAGR) from 2013 to 2021. Sales growth slowed to 5% in 2022 in the presence of rising interest rates.

Figure 1
Annual RILA Sales

pm-2023-06-heaphy-fig1.png

Starting in 2018, carriers began adding Guaranteed Lifetime Withdrawal Benefits (GLWBs) to their RILA offerings. These GLWBs have a wide range of guarantee levels and inherent risk to the issuing carrier. With slowing market growth, more carriers may launch GLWBs to stimulate sales. This article examines different types of RILA GLWBs with a focus on their embedded market risk.

RILA GLWB designs

RILA GLWBs fall into three major categories: account value benefit base, step-up only, and step-up and roll-up. All contracts begin in an accumulation phase wherein the customer’s lifetime income benefit has the potential to increase. When the customer elects to utilize the rider, the income phase of the contract begins, and the insurer starts making payments.

Table 1
RILA GLWB Designs

pm-2023-06-heaphy-table1.png[1] Assuming no excess withdrawals are ever taken.

Figure 2 projects the benefit base for each design assuming a starting premium of $100,000. The roll-up is assumed to be 6% compound interest.

Figure 2
RILA benefit base illustration

pm-2023-06-heaphy-fig2.png

Product Development History

Given the growth in RILA sales, many carriers have entered the market. The first RILA GLWB was launched in 2018 by Allianz Life. CUNA Mutual, Prudential, Equitable, and Brighthouse have also launched RILA GLWBs. We expect additional carriers to follow suit to maintain their position in an increasingly competitive market.

Figure 3
RILA Product launch History

pm-2023-06-heaphy-fig3.png

Key Risks

Adding withdrawal benefits to RILAs adds risks for the issuing carrier:

  • Policyholder behavior: GLWB pricing involves assumptions for utilization of rider benefits and dynamic surrenders; profitability will vary if behavior is different from expected.
  • Interest rates and reinvestment: withdrawal benefits extend the duration of the liability; long-term benefits are harder to duration match, which adds reinvestment risk.
  • Index performance: RILA claims vary by index account performance. Carriers use hedging to mitigate this risk. The introduction of withdrawal benefits increases the complexity of hedging, especially for designs with a roll-up.

The remainder of the article focuses on index performance risk for RILA GLWBs, how it compares to FIA GLWBs, and the impact on index account value hedging.

Index performance and cost of funds analysis

A stochastic cash flow model was used to analyze contracts for a set of real-world scenarios. The model calculated cost of funds (CoF, defined below) over a range of long-term index returns. This metric is useful because (1) it serves as a proxy for the cost of the liability and (2) it is comparable across regulatory regimes since it makes no assumptions about reserves and capital.

CoF definition: The earned rate needed on assets to pay for all liability claims without any profit or loss. It is calculated as the internal rate of return of initial premium, policyholder benefits, hedge purchase costs and payoff amounts.

Pricing actuaries who analyze CoF look to answer two questions:

  1. Average CoF: Is the expected yield on assets sufficient to cover CoF plus a reasonable margin for profits and the cost of capital? The analysis assumes that an average of CoF of 3.0% is supportable.
  2. CoF stability: Is the CoF stable across a variety of risk factors and economic scenarios? If not, what changes to product and risk management strategies are required? This analysis focuses on the sensitivity to index performance.

Table 2
Sample product designs and key assumptions

GLWB

Base contract

Key assumptions

Two designs:

  1. Account value benefit base
  2. Step-up with a 6% simple interest roll-up for 10 years

Rider fee: 1.5% benefit base

GLWB withdrawal rates: Solved to produce a 3% CoF

Two annuity products fully allocated to a single index account:

  1. RILA: 1-year 10% buffer with a 23% cap
  2. FIA: 1-year 7% cap

Real world scenarios: 1,000 scenarios from the Academy Interest Rate Generator

Dividend yield: 2% per annum

Insured life: 55-year-old female beginning the income phase at age 65

Interest rates and crediting rates: assumed constant

Results are presented three ways with various levels of hedging:

  • Without hedging: Shows the variance in CoF before hedging is applied along with variability of claims paid to the policyholder.
  • With hedging: Displays the CoF with hedging 100% of account value (a common approach for RILA and FIA without GLWB).
  • Percent of account value hedged: A constant percentage of account value is hedged to reduce the correlation between CoF and index performance.

Figure 4 shows the relationship between average index performance in the first 10 years of the scenario set on the x-axis and the CoF for each design described above assuming no hedging on the y-axis.

Figure 4
Cost of funds versus index performance: No hedging

pm-2023-06-heaphy-fig4.png

Results

  • Correlation to index returns: the correlation of CoF to index performance is high for all products before hedging is considered.
  • Roll-up downside protection – RILA: the high guaranteed floor results in more stable CoF in scenarios with low index performance.
  • Comparison to FIA: FIA products have a more stable CoF given full protection from losses and more limited upside potential.

Figure 5 includes hedging 100% of account value. This is a common hedging strategy for indexed products without GLWB. The hedging options are the same basket of call and put options (RILA only) used to hedge index accounts with a cap.

Figure 5
Cost of funds versus index performance: 100% of account value hedge

pm-2023-06-heaphy-fig5.png

Results

  • CoF correlation to index performance: hedging reduces the correlation. The correlation to index performance turns negative on the roll-up and step-up riders—this suggests the percent of account value hedged should be less than 100%.
  • Tail risk: the cost of the roll-up and step-up rider significantly increases when average index returns are below the 6% roll-up rate. Without adjustments to the hedge program, this exposes the carrier to large losses when index performance is low.
  • Dispersion of results: there is a wide dispersion of CoF for both RILA riders; this suggests that a more nuanced view of hedging is needed to increase stability in CoF relative to FIA.

Figure 6 shows the impact of adjusting the hedging strategy to minimize correlation between CoF and index returns. A simplified approach was used with a constant percent of account value hedged. Each product design has its own hedge percentage that was derived to bring correlation close to zero. This simplified strategy is illustrative only. In practice, a hedging strategy would be much more rigorous with ongoing monitoring and adjustments required to minimize variance.

Figure 6
Constant percentage of index credits is hedged to reduce market sensitivity

pm-2023-06-heaphy-fig6.png

Results

  • RILA - account value benefit base: percent of account value hedged was increased to minimize the correlation of CoF to index performance.
    • Key drivers: index credits directly impact prospective income payments. The increased hedge percentage suggests that the present value of income payments for the assumed insured life (at the CoF) is higher than the account value. This will not always be the case and will vary by policy, economic scenario, and other actuarial assumptions.
    • Implications: market risk can be reduced by adjustments to the portion of account value hedged. There is still some variation in CoF, which may be reduced by refining the hedging strategy.
  • RILA – roll-up and step-up: hedge notional was decreased to minimize the correlation of CoF to index performance. Sensitivity remains for scenarios with very high or low index performance. The CoF increased to 4.0%, which is above the 3.0% target.
    • Key drivers: the impact of index credits on GLWB payments depends on whether performance outpaces the 6% roll-up rate. This implies that the amount of account value hedged should vary over time in response to past performance. This complicates hedging and can be seen in the skewed shape of the distribution after applying a constant hedge percentage.
    • Implications: simple adjustments to the percent of account value hedged will not be able to reduce the tail risk of very low or high index performance. Hedge strategies should consider other methods, such as dynamic hedging (e.g., as seen on traditional variable annuity GLWB). Given the substantial increase in CoF, withdrawal rates may have to decrease to get back on target.
  • FIA – roll-up and step-up: hedge notional was decreased to minimize the correlation of CoF to index performance. Little sensitivity remains, but the CoF increased to 0.4% above the target, which requires a reduction in withdrawal rates.

Conclusion

RILA products have enjoyed tremendous sales growth up until the rising interest rate environment of the past year. This growth included the launch of GLWB riders. When carriers first introduced GLWBs, income benefits had a strong link to account value performance. This had the advantage of closely aligning the index account hedging strategy with the future guaranteed income stream. In 2021 carriers added guaranteed roll-ups which increased both downside protection for the customer and risk for the insurer, requiring a more robust approach to risk management.

As innovation continues, pricing actuaries need to consider how GLWB riders interact with index credits and market risk ramifications. Risk mitigation strategies should be customized accordingly and reflected in pricing and product designs.

Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries, the editors, or the respective authors’ employers.

Matt Heaphy, FSA, MAAA, is a manager with the Actuarial Practice of Oliver Wyman. He can be reached at matt.heaphy@oliverwyman.com.

Nicholas Carbo, FSA, MAAA, is a senior principal with the Actuarial Practice of Oliver Wyman. He can be reached at nicholas.carbo@oliverwyman.com.

David Elliott, FSA, CERA, MAAA, is a senior manager with the Actuarial Practice of Oliver Wyman. He can be reached at david.elliott@oliverwyman.com.