Executive Summary
The fundamental promise of guaranteed living benefit riders on annuities is the potential to invest more aggressively for growth, while maintaining an income guarantee no matter what, in case things don't work out as expected. In other words, at the most basic level, using income guarantees is a "floor with upside" approach to investing for retirement income. The floor may not be the most appealing income level, but it's tolerable on the downside when there's a hope and real chance that things could turn out much better.
Yet in today's environment, the combination of lower prospective returns and higher annuity costs is presenting a troubling new alternative: that guaranteed income riders no longer representative a "floor with upside" but instead are a floor that will be almost impossible not to fall down to and hit. In other words, what was supposed to be a floor that would only be relied upon in a worst case scenario may instead turn out to be the best case scenario and an unavoidable outcome instead!
Unfortunately, though, if the reality is that the income guarantee of the annuity is no longer a worst case scenario but actually a best case one, then it no longer represents the fundamental trade-off that made it so attractive in the first place. Instead, if there really is little upside to the income guarantee, then it may be more appealing to skip the guaranteed living benefit rider altogether and instead just buy a single premium immediate annuity for a comparable guarantee and invest the rest, or just invest and spend conservatively using a safe withdrawal rate approach and simply ride out any intervening market volatility!
Understanding The Guaranteed Living Benefit Rider
The guaranteed living benefit rider, first associated with variable annuities and more recently with equity-indexed annuities, was first created in the late 1990s and rose dramatically in popularity in the early 2000s after the technology crash. The basic concept of the riders was relatively straightforward: for a modest fee, the client could remain invested as he/she wished, but be assured of a minimum "benefit base" that would grow over time and could be converted into income via either withdrawals or annuitization. At best, the cash value would grow favorably with returns, but at worst at least the income guarantee would be available to fall back on.
Thus, for instance, if the client put $300,000 into a variable annuity with a living benefit rider, then an amount equal to $300,000 growing by 5% per year could be withdrawn against or annuitized in the future (depending on the terms of the contract), and in the meantime the assets would remain invested and grow (or not) based on whatever the market delivered. If market returns were good, the client could walk away with the cash value plus growth (minus the cost of the guarantee itself, which often was a fairly modest 0.25% - 0.50% {in addition to other applicable annuity and investment charges}). If the markets crashed, the benefit base amount would be available to generate retirement income. Heads and the client could win, tails and the client would at least not lose as much, which was a reasonable win-win outcome.
Of course, the terms of the guaranteed retirement income from such riders were not great. Often they included age setback provisions, unfavorable mortality tables, and were built upon very low internal rate of return assumptions. Many provided withdrawal guarantees that ultimately amounted to little more than the client spending their own money for 15-20+ years before the annuity company was ever actually on the hook for anything. But as a worst-case-scenario floor, at least it was something, especially since ultimately the goal and intention was to do far better than the income guarantees anyway; the guarantees were only there as a last resort, just in case.
Living Benefit Riders In Today's Environment
So what's changed in today's environment? There are three factors that have changed together in recent years, which all make the risk/reward trade-offs of living benefit riders far less appealing than they once were.
First of all, the raw costs of living benefit riders themselves have increased; as volatility rose in the years after the financial crisis, and many companies realized they needed to charge more simply to safely back the guarantees they were providing, what was once a 0.25% - 0.50% rider guarantee now often costs 0.75% to 1.25%. Often riders that have a lower price have the option to increase it to a higher level at the discretion of the company (sometimes as far as 1.50% to 2.50%), which unfortunately may be an all-too-probable outcome if the markets crash and the rider really needs to be relied upon (although in fact, some companies have already been raising the fees on existing riders, even though markets have been trending higher). And of course, that's in addition to the raw M&E cost of the annuity itself, and the expense ratios of the underlying annuity subaccounts, which can bring the total cost up anywhere from 2% to 4% depending on the details of the contract.
Second, today's investment environment has a lower expected return over at least the intermediate term environment, which limits the potential for upside in the first place. With bond yields at levels that are several hundred basis points below typical long-term returns, intermediate real TIPS yields at near-zero or even negative levels, and long-term (Shiller) P/E ratios at elevated levels that are typically associated with depressed 10-year real returns, the expected return for today's portfolios are well below average in the coming years, potentially no more than the mid-single-digits for a balanced portfolio. This is especially problematic if ongoing withdrawals are occurring. If the retiree is taking out 5% annual withdrawals, and dragging a 3% total expense for the cost of guarantees plus the cost of the annuity plus the cost of the investments themselves, it may be almost impossible for growth to exceed the drag of withdrawals and costs before reaching the point of inevitable depletion.
Of course, investors can try to leave room for greater upside by investing more aggressively, in a portfolio that at least has a possibility of out-earning 5% annual withdrawals and a 3%-ish cost drag. However, this too is no longer an option, as the third change in today's annuity contracts is that they are increasingly limiting investment choices to require diversified portfolios. In other words, most annuity contracts will require the policyowner to have as much as 40% of the contract invested in low-return fixed income investments where the annuity costs alone (never mind including withdrawals) exceed the expected return. With such a severe drag on the portfolio it's virtually impossible to have a sufficient amount invested aggressively enough to produce any real upside potential. After all, if 40% of the assets are generating no return (or a negative return after fees), then the remaining 60% of the portfolio must generate the growth for all of it; if 5% annual withdrawals are coming out, the remainder of the portfolio must generate an 8.3% net return just to avoid a path to assured depletion, which is actually a whopping 11.3% gross return before fees, which is at best very difficult to achieve in today's environment and unlikely to be achieved reliably.
Alternatives With Similar Floors And Greater Upside?
Given that such high equity returns are probably unrealistic in today's environment (after all, 11.3% gross returns are even higher than the overall long-term average, not to mention from an above-average-valuation starting point!), what are the alternatives?
The first is simply to spend more conservatively and stay invested with a lower cost portfolio; in other words, to utilize a safe withdrawal rate strategy (which is essentially a floor-with-upside approach) that deliberately sets spending low enough to ride out the market volatility. While following a 4% safe withdrawal rate approach may seem unappealing when an annuity income guarantee might offer 5%, it's crucial to remember that safe withdrawal rates assume inflation-adjusted spending, while living benefit riders generally only provide level cash flows that may be eroded by inflation over time. Which means while a safe withdrawal rate approach might mean spending $40,000 from a $1M portfolio initially, the spending is assumed to have risen to nearly $100,000 per year by the end (with a 3% inflation rate for 30 years), while the annuity guarantee will still be paying only $50,000 at the beginning and the end (while many annuities do have income guarantees that step up with market growth, remember that with 5% withdrawals and 3% costs and a low-return environment, the odds are not good that the income guarantee will ever materially rise with the market over time). Ultimately, then, the annuity approach generates $1.5M of cash flows for 30 years, while the safe withdrawal rate generates nearly $2M of rising cash flows (albeit from a slightly lower starting point); and the safe withdrawal rate approach also has far more upside potential to boot, as it's not dragging such hefty fees. In fact, one might wonder why it's worthwhile to pay for the annuity guarantee at all when it merely provides a series of cash flows that are almost 1/3rd less than was ever necessary even in the worst historical market return scenarios we've ever had! In other words, what's the point of paying for a guarantee that ensures less than what can be had by merely being conservative? The point of a guarantee should be to get a higher floor in exchange for conceding some upside, not to reinforce a floor that's even lower than what was already feasible without a guarantee!
Of course, one key distinction is that safe withdrawal rates are assumed for a 30-year time horizon, while at least some income guarantees from annuities are for life, however long that may be. In other words, the annuity also provides a longevity hedge at whatever income floor it offers, while a safe withdrawal rate approach is based on a time horizon the client at least might outlive. Yet if the goal is simply to secure an amount like $50,000/year (5% of a $1,000,000 portfolio) as a lifetime income and still have upside, the reality is that the retiree can simply buy a single premium immediate annuity to accomplish that goal! For instance, even at today's rates on ImmediateAnnuities.com, a 65-year-old couple can buy $50,000/year for life for $839,990 (or only $727,377 or $764,391 for a single male or female, respectively). This leaves the remaining ~$160,010 (or more) available to invest or to hold in reserves, without the cost drag of the annuity. In other words, if the reality is that a $1,000,000 variable annuity with a living benefit rider provides a guaranteed $50,000/year for life, but will have little or no upside in today's environment due to the combination of higher costs and lower returns, why not secure that same goal with only $839,990 (or less as rates rise) and leave the rest of the money invested elsewhere to produce a more feasible upside?
Understandably, though, many investors don't want to give up so much liquidity in a SPIA, and still aren't necessarily comfortable with the approach of just being conservative and riding out market volatility, and really want guarantees. Yet the challenge is that using an annuity with guarantees doesn't necessarily eliminate the risk, it simply shifts it - and potentially concentrates it - with the annuity company. After all, unlike immediate lifetime annuities that manage mortality risk by pooling together a large number of people so those who die young can help fund mortality credits for those who live long, with variable annuity living benefit riders a market decline forces the annuity company to set aside reserves to potentially pay guarantees for everyone at once. In turn, this can make the reserve demands and profitability of the annuity company more volatile, as was witnessed in 2008-2009; in fact, it's notable that while diversified investors have already recovered to new highs since the financial crisis, many annuity companies providing guarantees have permanently left the business after the financial crisis and are even trying to reduce their exposure to (and entirely get off the hook for!) their existing annuity guarantees. In other words, while the annuity guarantees were supposed to pool retirement assets and risk management together to protect retirees better than just being a diversified low-cost investor, the reality is that it was the standalone investors who came through with a quick recovery and some annuity companies that took a knock-out punch!
Furthermore, given the pain that the last bear market brought to companies providing retirement income guarantees, many of the guarantees shifted to contracts where the annuity company has more control: i.e., equity-indexed annuities. The upside for the annuity company of using equity-indexed annuities is that it's far easier to manage the costs of the guarantee when the company controls all the money that underlies the guarantee, and has the flexibility to set participation rates, cap rates, and spreads at levels that ensure the product will be viable. While that's not necessarily a bad thing - if the company offers genuinely competitive rates - the risk is that if another market event occurs, the insurance companies will simply reduce the participation rates, caps, and spreads at that time, and annuity-owners will have little recourse (as surrendering the annuity to reinvest for better returns at that point would forfeit the guarantees that were being purchased). In other words, it doesn't matter if the company guarantees its costs, if it also controls the money invested and has the right to dictate whether (or not) the policy will pay an appealing rate of return. This is ultimately the key concern for any equity-indexed annuity - similar to hybrid long-term care policies, it's easier for companies to offer more appealing guarantees when they have the control and capability to cut the rates of return paid to policyowners to make up for any shortfall! In fact, given that equity-indexed annuity policies already have less upside, simply by virtue of their inherent risk/return trade-offs, it appears increasingly likely that many equity-indexed annuities are destined for the same "no realistic upside" scenario as today's variable annuities with living benefit riders. On the other hand, as Moshe Milevsky recently noted, the raw annuitization payout rates on some equity-indexed annuity guaranteed living benefit riders are remarkably competitive to today's SPIA rates anyway, which suggests there may be little downside to at least taking a shot at risk-managed upside before annuitizing (if annuitization is already the plan), at least for assets that were going to be held in cash or invested for little or no return anyway.
Ultimately, though, the fundamental point to all of this is that the purpose of annuities with living benefit riders is to provide a guaranteed income floor, while also being able to invest for (and have a realistic chance of achieving!) upside growth. In the early days of guaranteed living benefit riders, this is exactly what the contracts delivered, allowing investors to put a floor under their portfolios (or better yet, under the most aggressive subset of investments in their portfolios) at a more modest cost, and without being required to drag a large portion of the portfolio in low-return investments that cannot outearn the cost of the guarantees wrapped around them. Accordingly, clients should still be very cautious about surrendering any existing annuities without a thorough due diligence process. But in today's environment, as insurance companies require investors to hold a portion of assets inside the annuity in less productive investments, and at a higher cost, and in a lower return environment, the trade-off is simply not what it once was. Suddenly, saving the cost and just investing and spending conservatively appears to be a remarkably appealing alternative.
So you are saying – stay away .
This is a great article that addresses a number of questions I have had recently. Thanks.
Very well stated – exactly as it is.
great summary of what really goes on inside these complex instruments
And of course the rider costs are on the income account value and not the actual account value, which makes actual rider costs even greater than the nominally stated expense. If the heavy cost drag weren’t enough, as soon as you get an account value decline, the costs increase and drown down the account value even more. To me this is the ‘variable annuity death spiral’ as the account value just gets eaten up by ever-increasing costs and more rapidly spirals to zero.
Kevin,
This varies by contract. Some calculate the rider cost as a percentage of the benefit base, others calculate it as a percentage of the cash value.
Frankly, I would make the case that applying the rider cost based on the income account value actually IS the more appropriate way to structure it. You SHOULD be paying the cost of your guarantee based on the size of your guarantee; to do otherwise risks mismatches that can misprice the guarantee and potentially create a risk for the insurance company and therefore those who are counting on that guarantee.
You are correct that it can lead to a cash value ‘death spiral’. Though again, ironically that actually means the client more quickly gets to the point where the insurance company is actually paying out from its own pocket, and then the policyowner gets it back. This is distinctly different than the death spiral that occurs with variable universal life policies. With a VUL, the death spiral causes the policy to lapse and the insurance company is off the hook; with a variable annuity, the death spiral causes the guarantee to activate and puts the insurance company ON the hook. Very different outcome!
– Michael
Great article Michael. Its scary to see an article earlier in the week discussing how annuities are again reaching record sales (especially with fees up and benefits down).
I’d be curious to see how many living rider benefits are turned on. And of those, how many have provided real value over the account value. My suspicision is nearly zero.
It’s also becomes a timing issue. For a 70 year old retiree to benefit from the rider, they might need to live to 90, 95 or longer.
In my opinion, these are a complete rip off. Its a shame that our industry and our compliance departments allow this nonsense.
I am in the broker-dealer world and have been for only the last 2 years. I am beginning on my own to see the negative factors of the GMWL/GMWB riders. Most of them I have seen have 1% avg sub account fees, 1.15%-1.25% M&E and 1% -1.25% Rider fee making these fees at least 3.25%. That being said a 5% roll up is really just 1.75% or less depending on the actual fees. I have yet to be sold on them myself and am glad I actually haven’t placed clients in them. There are a few out there that have a 7.2% increase but haven’t done the extensive research to know if these are in the same camp as the annuities mentioned above.