Willis Towers Watson 2020 FIA Pricing Survey
Pricing and benefits in the U.S. fixed index annuity (FIA) market have been subject to the pressure of low interest rates for some time, and changes to reserving regulations are also on their way. A Willis Towers Watson FIA pricing survey has investigated how companies are handling key pricing assumptions and methodologies.
The combination of low interest rates and high equity market volatility has continued to make FIAs an attractive proposition for many investors. Demand for these products has grown dramatically over the past decade as variable annuities have felt more pressure from the economic and regulatory environment. Even if rates rise, we expect further solid demand for FIAs.
Good news as this may be for writers of FIAs, the very factors that make the products interesting to customers who want relatively low risk in the current environment are something of a double-edged sword, affecting pricing options and assumptions as well as the caps and benefits companies can offer.
Add to this, the market is facing impending new reserve regulations — both the extension of principle-based reserving (PBR) to FIAs and changes to U.S. generally accepted account principles, or GAAP, treatment of long-term duration insurance contracts. While the changes should result in lower statutory reserves for living benefits under PBR, they are also expected to increase volatility.
These are, concurrently, vibrant but also potentially challenging times for FIA writers — requiring them to tread carefully.
So, these are, concurrently, vibrant but also potentially challenging times for FIA writers — requiring them to tread carefully. Many are reevaluating their business models and approach across the board, from pricing methods, products and profitability targets to investment strategy, reinsurance and distribution.
Given the regulation changes, as well as the current economic environment, we focused our 2020 Fixed Index Annuity Pricing Survey on the most critical assumptions and methodologies relating to those items, including key policyholder behavior and investment assumptions and methodologies used in pricing. The survey set out to drill down into current practices and trends to see how companies are responding to these pressures. Companies that responded represent approximately 80% of top 20 market sales, with 17 of the top 20 writers of FIAs in the U.S., representing approximately $19 billion in FIA sales for the first half of 2020.
To put the overall results in context, we first asked companies about the nature and principal features of their product offerings.
Most companies surveyed offer products with a range of surrender charge schedules, with a 10-year option constituting almost half of sales. The remainder of products sold are mostly shorter durations, with a seven-year surrender period also being broadly popular. Shorter surrender charge sales have grown in recent years, and sales for products with surrender charges longer than 10 years have dropped down to minimal levels. Independent agents represented about 40% of sales, with broker/dealers account and banks providing about 20% each.
Of these sales, all companies in the survey said they offer a lifetime guaranteed minimum withdrawal benefit (GMWB), a feature included in almost 30% of sales where policyholders are looking for lifetime income without loss of liquidity. Election rates do vary significantly by company, depending primarily on rider competitiveness. Most policies sold include only a basic death benefit, with enhanced death benefit sales at very minimal levels. The low-rate environment has squeezed out most enhanced death benefits.
Point-to-point crediting strategies based on S&P 500 continue to comprise the majority of sales, predominantly based on an annual strategy (Figure 1). While a fixed methodology is used by all respondents, it only represents about 9% of sales.
| Crediting methods | Number of companies | 15th percentile | Average | 85th percentile |
|---|---|---|---|---|
| Annual point-to-point | 17 | 36.8% | 64.0% | 87.6% |
| Multiyear point-to-point | 12 | 0.0 | 19.2 | 50.5 |
| Monthly sum | 6 | 0.0 | 2.3 | 5.2 |
| Daily/Monthly average | 6 | 0.0 | 2.2 | 2.2 |
| Fixed | 17 | 3.8 | 8.9 | 14.3 |
Crediting methods based on the S&P 500 options are capturing about 44% sales, with volatility-controlled indices just barely behind that level.
Numerous factors underpin the why, what and how of product design and pricing.
Numerous factors underpin the why, what and how of product design and pricing. A breakdown of the key ones from the survey findings follows.
Of the companies surveyed, only 30% reported using a full asset liability (ALM) model, with the rest using a liability-only model. We have seen more companies expand their pricing model sophistication in recent years and expect to see more do that as PBR demands more rigor be placed on ALM strategies. ALM models can be particularly helpful in calibrating investments strategies, as well as renewal rate/cap setting. Also, as more companies look toward more exotic index strategies, as well as Registered Indexed Linked Annuities (RILA), ALM interactions become more important.
Underscoring the need for reliable results, many are validating pricing models annually and typically rely on the company’s risk management function to do so. Others validate from within the pricing team.
The majority of companies are split between either using a level deterministic scenario or real-world stochastic scenarios for base pricing, with many using the other for sensitivity testing. Most use current market prices to price initial index options and then project consistent with economic scenarios. Companies with deterministic pricing tend to also assume static option prices.
Most are sensitivity testing investment yields, base surrenders, statutory interest rates, and a combination of policyholder behavior and economic conditions. Given recent market conditions, unsurprisingly several indicated they are testing additional scenarios with lower rates (including longer periods of lower rates).
Given the market volatility, several companies indicated they are looking more frequently and closely at investments. Allocations by asset class vary considerably by company, with allocations to private placement, high yield, mortgages, structured assets and real estate often driving higher yields. To illustrate the general trends in allocations, Figure 2 groups companies into three categories based on their relative propensity to focus more on public investment grade bonds, commercial mortgages or alternative assets. Responding companies were fairly evenly split between these three categories.
| Asset class | Traditional | Commercial mortgage | Alternative | Average allocation |
|---|---|---|---|---|
| Public IG bonds | 62.8% | 43.3% | 27.4% | 42.9% |
| Private placement | 22.6 | 18.0 | 3.0 | 13.8 |
| High yield | 0.0 | 0.3 | 0.0 | 0.1 |
| Commercial mortgages | 14.6 | 28.7 | 7.3 | 17.1 |
| MBS/CMBS | 0.0 | 7.8 | 4.5 | 4.5 |
| Asset-backed securities | 0.0 | 0.0 | 9.7 | 3.5 |
| Structured assets | 0.0 | 0.5 | 8.4 | 3.3 |
| Emerging market bonds | 0.0 | 0.0 | 1.3 | 0.5 |
| Real estate | 0.0 | 0.4 | 8.0 | 3.1 |
| Other | 0.0 | 0.9 | 30.4 | 11.4 |
Average net spreads are relatively similar when comparing non-GMWB investment strategies with GMWB. Spreads vary from under 200 basis points (bps) to near or over 300 bps for a few companies (reported as of June 30, 2020). These spread assumptions are usually the most significant driver of profitability and competitiveness; however, equity return assumptions are also varied, generally ranging from as low as 5.0% to over 9.0%. Some also just assume a significantly lower index credit (e.g., 3.0%, as a percentage of the cap), often then taking a deterministic scenario approach and thus not reflecting stochastic scenarios. Equity/index return assumptions have been a key driver of FIA product profitability as well, both the levels themselves as well as how leveraged a given company’s products and investment strategy are.
As shown earlier, all companies surveyed offer point-to-point crediting strategies based on the S&P 500. The vast majority use volatility control indices as well (Figure 3). However, for pricing purposes, many are not directly modeling more than annual point to point. In projecting caps, most assume a fixed spread, although many still take a fixed-rate-and-cap approach, despite this likely leading to increasing spreads over the projection.
| Crediting methods | Number of companies | |
|---|---|---|
| Offered | Modeled | |
| Annual point-to-point | 17 | 13 |
| Multiyear point-to-point | 12 | 5 |
| Monthly sum | 6 | 1 |
| Daily/Monthly average | 6 | 1 |
| Fixed | 16 | 14 |
| Other | 5 | 3 |
| Indices | Offered | Modeled |
| S&P 500 | 17 | 13 |
| Volatility control | 15 | 8 |
| Hybrid indices | 4 | 1 |
| Fixed | 15 | 13 |
| Other standard market indices | 11 | 8 |
| Other customized indices | 4 | 3 |
Base lapse assumptions can vary dramatically by company for shock and ultimate levels, especially for contracts with GMWBs. Most companies surveyed have updated surrender assumptions within the past two years. While the use of predictive analytics is growing, it is still used by only a minority of companies.
The majority of companies reflect interest rate-based dynamic surrenders on the base product in pricing to account for the interest rate sensitivity, with interest rate-based dynamic surrenders more prevalent on the non-GMWB product.
Initial utilization assumptions are often lower when there is a GMWB rider. Many companies vary partial withdrawal assumptions by policy characteristics, such as duration and age, and several companies have assumptions that vary by tax status (qualified versus non-qualified). They also typically include layering in some degree of required minimum distributions (RMDs) on top of base utilization rates.
All but one company in the survey reported a duration and issue age-based assumption, with highest utilizations at typical retirement ages. The majority of companies also allow for a small percentage of people who never start the GMWB benefits. Further, about half of companies assume non-rider withdrawals prior to GMWB activation.
The majority (70%) of respondents reflect GMWB dynamic surrenders, with most basing their formulas on the present value of benefits. On average, those who use present value of benefits (as opposed to benefit base) show greater sensitivity in surrender multiples to levels of account value. Company views on the impact of GMWB rider value on full surrenders can vary considerably.
The general squeeze of low interest rates on typically more conservative investment policies is a key driver for FIA writers’ use of reinsurance. Despite the current economic environment, reinsurers have maintained a strong appetite for annuity products, enabling FIA writers to potentially access a broader range of asset strategies and diversify risk. Moreover, the reinsurance market supporting new business flow transactions is active and growing with newer entrants that specialize in asset-intensive product lines.
Many FIA writers are looking for reinsurance solutions to:
If appropriately structured, new business reinsurance arrangements can be established in a manner to minimize risks while leveraging investment and product expertise of the reinsurer.
All these prevailing practices and trends aside, the COVID-19 pandemic has inevitably had some knock-on effects for FIAs.
All these prevailing practices and trends aside, the COVID-19 pandemic has inevitably had some knock-on effects for FIAs.
Several companies say they have moved quickly to protect their FIA business, making adjustments in areas such as product investment strategies, scenarios considered, crediting methodologies, and pricing targets and methodologies. Specific actions taken have included lowering mean reversion targets, adjusting the credit quality/assets mix of investments, lowering crediting rates/caps, accepting lower levels of profitability (below targets) and more frequent discussions with the investment team. Some say they have taken more significant steps, such as opportunistic asset purchases and reducing guaranteed GMWB income levels, or discontinuing GMWB altogether. Many have also sought out reinsurance solutions for the reasons noted above. We are also seeing companies look toward index products for future growth, with traditional FIAs, RILAs and other new innovative index products.
But even when the shadow cast by COVID-19 over markets is lifted, FIA writers will still have plenty to think about in pricing strategy. Insurers should be ready to walk a tightrope of “lower for even longer” interest rates and the need to reassure investors about how they are managing the potential for volatility introduced by the incoming reserving requirements.
Rick is a Director in WTW’s Insurance Consulting and Technology Life practice and leads WTW’s Annuity Initiative for the Americas. He has extensive experience with financial modeling and reporting and is mostly focused on annuity-based pricing, model building and validation, and M&A activity.
