As deferred annuities go, this one has legs.
Since its inception in 2010, the registered index-linked annuity, or RILA, has grown significantly in popularity. RILAs are deferred variable annuities that tie their performance to an index, such as the S&P 500. Unlike a simple S&P 500 fund, however, these products limit the downside risk of loss (e.g., via buffers and floors) and consequently the upside gains (e.g., via caps and participation rates) in various ways and over various terms (e.g., over one-, two- and six-year periods). Since 2017, sales of RILAs (also known as index-linked variable annuities, buffered annuities or structured annuities) have more than tripled to a quarterly sales volume of $10 billion as of the fourth quarter of 2021 (Figure 1). Secure Retirement Institute data show that RILA sales grew by 61% over 2020, amounting to $38.7 billion of the total annuity sales of $254.6 billion in 2021.
As a result, other deferred annuity (DA) product lines are losing ground. Since 2017, more traditional DA products have experienced market share dips ranging from 1% to 5%. In the fourth quarter of 2021 alone, RILAs comprised 17% of all DA sales, according to Wink, Inc. data.
sales volume of $10 billion as of the fourth quarter of 2021.
Source: Wink
Yet why is the growth in RILA products outpacing other DA products? The answer lies in the one-two punch that RILA products offer. They deliver downside risk protection — a safety net with various levels of protection — but with stronger upside potential than their fixed indexed annuity, or FIA, cousins . This new risk/reward trade-off has increased the popularity of RILAs among buyers looking for a more balanced growth strategy.
How RILAs differ from the rest of the DA pack helps drive sales. Structurally, RILAs meet the IRS definition of an annuity for tax-deferral purposes by including annuitization options. Yet unlike FIAs, RILAs are registered at the federal level with the Securities and Exchange Commission (SEC), much like other variable annuities. This is required because RILAs allow for losses that can penetrate the minimum Standard Nonforfeiture Law’s floor values, which all fixed annuities regulated at the state level must meet.
However, that’s where the similarities with variable annuities begin to blur. Unlike variable annuities, where growth is tied to unitized separate account funds, RILA crediting is formula-based, like FIAs. That is, the crediting structure does not directly tie to the value of that fund. Instead, RILAs credit interest based on a formula that matures at the end of a one or multiyear term.
Another key differentiator when compared with variable annuities is that loss protection is built into a RILA. The most popular loss protection structure is the buffer, which prevents negative credits unless the index losses exceed the buffer. For example, if the index loses 12% and the RILA has a buffer of 10%, the account value would experience only a 2% decrease.
It is reasonable to expect that in order to provide downside protection via a buffer or other floor limitation, there must also be an analogous limit on the upside — especially when these limitations operate on well-known indices, such as the S&P 500.
Because RILAs offer more significant downside risk than do FIAs (which usually limit downside risk to an interest credit of zero), they likewise offer higher caps or participation rates — all else being equal.
For a one-year point-to-point crediting strategy, the hedge strategy for a zero floor FIA would be to buy a call spread, which is a long at-the-money (ATM) call option coupled with a short out-of-the-money (OTM) call option, with the OTM strike equating to the defined cap rate for the index crediting period. Solving for the OTM strike is a simple iterative process with the option budget being the limiting parameter (option budgets are typically defined as the invested asset yield less a profit spread). The payoff structure of this hedge strategy is illustrated in Figure 2 with the red payoff line showing the lack of downside risk as well as, depending on the option budget, the relatively low upside credits given the cap of 3%.
as well as, depending on the option budget, the relatively low upside credits given the cap of 3%.
Source: WTW
For a one-year point-to-point crediting strategy, the hedge strategy for a RILA with a buffer would be to buy a call spread, similar to the FIA zero floor example, but additionally to short an OTM put option, with the strike equating to the buffer level. The payoff structure, shown in Figure 3, illustrates eliminated downside loss until the buffer is reached, at which point excess negative returns reduce the account value. The upside mirrors the zero-floor strategy with a call spread payoff, but here the short put price implicitly increases the option budget and pushes the cap rates higher than the zero-floor example.
negative returns reduce the account value.
Source: WTW
For a one-year point-to-point crediting strategy, the theoretical hedge strategy for a RILA with a negative floor would be to buy a call spread, similar to figures 1, 2 and 3, but additionally to sell a put spread (long OTM put option and short ATM put option) with the long put option having a strike equating to the negative floor. The put spread enables a higher option budget that is directed toward a higher cap rate, which in this example is 15%. The hedge allows for a 1:1 index credit payoff until the respective floors and caps are met, as illustrated by the red net payoff line in Figure 4.
illustrated by the red net payoff line.
Source: WTW
Because of these limits on loss and on gains, RILAs tied to the S&P 500 will generally perform better than the S&P 500 index in down markets and worse than the pure index in strong up markets. Much of a RILA’s value lies in how the crediting formula is structured, how volatile or steady the market is, and to which index the RILA is linked.
In many cases, RILA products still link to the S&P index; however, last year the market saw significant growth in the use of customized indices, which are emerging to address lower interest rates while providing some stability in returns. Among the variety of bespoke RILA indices, policyholders can select volatility-controlled funds that provide protection against a bear market while also still achieving attractive caps and participation rates — key measures of value to policyholders and agents alike.
The S&P 500, when subjected to various volatility control limits, can produce a range of results as shown in Figure 5.
Source: Dow Jones Indices
Since the S&P 500 price index (shown in purple in Figure 5) has uncontrolled volatility, its volatility can range significantly over various time periods. The 15% volatility controlled (15% VC) version of the S&P 500 (in yellow) outperforms the uncontrolled index in some periods, notably during and after the 2007 – 2008 bear market. This occurs partly because during the very-high-volatility bear market of 2007 and 2008, the 15% VC version is constrained to a volatility of 15%. Also, during the lower volatility bull market following these years, the volatility of the 15% VC version is levered up. By comparison, the highly constrained 5% VC version of the S&P 500 (in magenta) experiences neither the significant drops nor the significant growth of the unconstrained version.
Since volatility is a key driver of the participation rates of RILA indexing, a high participation rate (e.g., 100%) can be offered on a 5% VC version of the S&P 500. For the same hedge cost, the participation rate on the uncontrolled S&P 500 price index might only be 40%, for example. We show the comparison of these two hypothetical offerings in Figure 6, during the historical period 2003 to 2014.
Source: Dow Jones Indices
Although the participation rate is higher (100%) on the 5% VC version of the S&P 500, the performance is similar — even slightly worse by 2014 — when compared with the 40% participation rate on the unconstrained index. Note that in this example, we use consistent participation rates of 40% and 100% across the 11-year period. In practice, the affordable participation rates might vary year by year, particularly the 40% on the unconstrained index, as the volatility of the underlying index changed over time.
In addition, the RILA market has seen some additions of guaranteed living withdrawal benefit (GLWB) riders onto RILA products. Of course, for those RILAs that include a GLWB, there are additional risks to consider. As compared with a FIA, RILA returns can be both negative and more positive. The added return volatility in RILAs increases GLWB pricing risks. Figure 7 shows the GLWB moneyness (defined as the ratio of benefit base to account value) for a typical RILA and FIA product design for the period 2000 through 2020, assuming both products were linked to the S&P 500.
product design for the period 2000 through 2020, assuming both products were linked to the S&P 500.
Source: WTW
Given the FIA’s 0% floor and relatively low cap, as well as the rollup design of the GLWB, the ratio of the benefit base to the account value fairly predictably steadily grows over time. This is in direct contrast to the RILA. Given the 10% buffer and higher cap, the account value would experience greater volatility, and therefore the moneyness of the rider is harder to predict. In periods of large negative returns (as in 2009 in Figure 7), the account value drops significantly below the benefit base; however, after periods of positive returns, causing the account value to recover, the moneyness of the RILA rider actually dips below that of the FIA (2016 to 2020 in Figure 7).
Another consideration is how the income payments are defined. While the rollup designs generally lock in benefit amounts once GLWB utilization occurs, other designs in which the benefit base is based on account value performance allow the benefit to either increase or increase and decrease based on the index returns. How these benefits fair compared with their locked-in counterparts will greatly depend on the index performance over the life of the policy — something insurers and policyholders will no doubt be watching closely.
As the RILA market matures, additional GLWBs and new designs are expected to be brought to the market. As new features are introduced, insurers will need to think through a variety of product design approaches to mitigate these risks.
The hybrid nature of RILAs has made them sought-after products. They live in a product space that provides higher risk and return options than does the typical FIA but somewhat lower risk and return than the typical variable annuity ; however, getting RILAs registered and filed does take some time for companies wanting to issue them. The filing requirements for registering securities have made the process onerous.
Congress has taken notice of some of these difficulties, and as a result, a bipartisan effort is under way to simplify registration. The Index-Linked Annuities Act, currently two bills in the House and Senate, would call on the SEC to develop new registration rules specific to RILAs.
The calculation of the value of the RILA between anniversaries (commonly referred to as the interim value) is a key area of interest and, in the absence of regulation clarity, currently varies from company to company. To this end, the the National Association of Insurance Commissioners (NAIC) has proposed Actuarial Guideline ILVA, which aims to outline the methodology for interim value calculation. One perspective is that the RILA has emerged as a result of the freedom deriving from this lack of regulatory limitation.
Unfortunately, the current proposed approach is being viewed by some in the industry as somewhat extreme. Although clearly well intentioned, the proposed regulation would outlaw the same kind of interim value calculations that are embedded in every FIA currently for sale. Again, in the absence of the strict regulation that the NAIC is now proposing, the FIA value proposition involves providing an interest credit only at the end of the crediting term (usually one year). There has been some industry pushback, but the technical nature and regulator-driven process has led to a stringent proposed regulation in this area.
The interim value regulation proposal arises out of the concepts grounded in the nonforfeiture requirements created for traditional variable annuities’ unitized structure. For non-unitized structures, there is currently no existing regulatory limits on the interim values. Although the NAIC acknowledges that interim values have not been a significant concern, the proposed framework could trigger product restructuring, repricing and refiling for currently approved products.
The RILA will likely continue to evolve as an important part of the annuity market lineup. A hybrid of fixed and variable annuities, RILA provides a new paradigm in the risk/return trade-off, and one that is of particular interest to pre- and postretirement customers.
Hopefully, regulations will remain sensible and allow for the range of innovative concepts that have continued to emerge. If so, growth is expected to continue as more customers and insurers are attracted to these well-balanced risk-and-return products.
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.
