Annuity Shopping: GLWBs and Asteroid Insurance

New Reader? Get free regular updates from Can I Retire Yet? on saving, investing, retiring, and retirement income. New articles weekly. Join more than 18,000 subscribers. Unsubscribe at any time:

Vanguard made waves in late 2011 by introducing a budget-priced Guaranteed Lifetime Withdrawal Benefit (GLWB) rider on its variable annuity. I knew I would be in the market for guaranteed lifetime income at some point. And, though I didn’t know much about the option then, I felt better knowing Vanguard offered one. Plus, there seemed to be at least murmurs of approval in the personal finance world….

So it came as a bit of a shock to learn that, effective this coming May 1st, Vanguard would be raising its rider fee (from 0.95% to 1.2%) and reducing its benefit (by 0.5% for certain ages) on GLWBs. Whether or not they are a good deal now, they are clearly going to be a worse deal going forward! Was I about to see the opportunity of a lifetime (or a retirement), slip away? Like any shopper faced with a limited-time offer, I swung into action….

I spent a good part of my free time this past week studying up on GLWBs. I’m not an academic, just an engineer going shopping for lifetime income, so I relied heavily on the analyses performed by others. I read about a dozen articles, and poured over many simulations of GLWBs against past history and future projections of current market conditions. I also called Vanguard and spent some time quizzing one of their annuity sales team members.

Before we get into my conclusions, let’s review annuities in general and GLWBs in particular….


The Holy Grail of retirement security is guaranteed, lifetime, inflation-adjusted income. Most of us will get some of that — though not enough — in the form of Social Security. A few will get the balance of what they need in the form of pensions. But the rest of us will have a shortfall between our living expenses and our guaranteed income. We could try to fill that gap by managing our investment assets to support systematic withdrawals indefinitely in retirement. But the insurance industry has what sounds, at least on the surface, like a simpler, safer idea: an annuity….

In its most essential form an annuity exchanges something of value (such as a career, or a lump sum of money) for a stream of payments — an income. Annuities can potentially help to ensure comfort and peace of mind in retirement. Unfortunately, more often than not, annuities, especially the complex variable annuity, have become a quagmire of pushy salesmen, hidden expenses, and dizzying complexity.

With annuities, as for investment products, we can greatly simplify our lives by sticking with the offerings from the most highly trusted, consumer-oriented companies. Let’s start with Vanguard. If you go to their landing page for annuities you’ll see they offer essentially three products for retirement income: (1) an immediate fixed annuity (SPIA), (2) an immediate variable annuity, and (3) a deferred variable annuity with Guaranteed Lifetime Withdrawal Benefit (GLWB).

Immediate annuities, where you buy a constant income stream over a certain period for a lump sum, are a relatively simple and well-known offering. They pool your lifetime risk with that of others to make your money go farther. I’ve discussed them elsewhere, and will surely be returning to the subject again.

Variable annuities remain at the eye of the retirement income hurricane. They are both the most popular, and the most despised, products. I have yet to see any independent, trustworthy source whole-heartedly recommend one. When I click on Vanguard’s link it takes me to an underwhelming page from American General Life that primarily touts the product’s many options, with no hard data on why I’d choose one. I remain unsold on variable annuities, in general.

That leaves one unexplored option at Vanguard: the GLWB (which is actually a rider or option for a variable annuity). Are GLWBs a viable alternative to immediate annuities? Let’s talk about what they are, and how they stack up against the alternatives.


GLWBs sound almost too good to be true. (Danger ahead.) They promise downside protection, lifetime income, upside potential, and the flexibility to cancel and withdraw your assets at any time. They seem like the perfect retirement income solution, until you start taking a hard look at the details….

A GLWB is an extra-cost rider that you purchase on top of a variable annuity. It guarantees that some fixed percentage of your assets can be withdrawn each year for lifetime. That amount will never decrease in nominal terms (not adjusted for inflation). The actual percentage you can withdraw is based on your age at the time of the first withdrawal, and whether you choose a single or joint life option, to cover your and your spouse’s lifetime. (The current withdrawal percentage at Vanguard for a couple age 59 is 4%.)

In theory, annual withdrawals from a GLWB might increase in rising markets, as your paper Total Withdrawal Base (a high water mark) ratchets up, for potential upside. But, if regular withdrawals, fees, or market declines take a toll on the real Contract Value of your underlying investments, then the Total Withdrawal Base will never ratchet up. However, if the Contract Value goes to zero, then the insurance will kick in and continue to make your guaranteed payments for life.

Costs vary by company and by feature, but figure the insurance company will collect about 1% minimum per year on your Total Withdrawal Base. (Note how that supercharges the fees, drawing your Contract Value down even quicker, since the Total Withdrawal Base will likely be much higher than your Contract Value over time.) Add to that any additional management expenses for the underlying funds. Even frugal Vanguard, with its suite of dirt cheap mutual funds and ETFs, seems to incur an extra 0.5% or so in expenses when managing those as part of a variable annuity. Other companies are much worse — with total expenses for a variable annuity with GLWB running around 3% annually in many cases.

Pros: Short List

Let’s start with the few clear positives for GLWBs:

  • A GLWB does promise lifetime income. Assuming the insurance company stays in business, that could be a great comfort in your later years. According to Wade Pfau, current economic and market conditions could be creating a climate where the guarantees will be “in the money.” But, if the times are so far out of the historical norms, and all our annuities run dry at once and must be backstopped by the insurance provisions, will the insurance companies really be able to pay? And just how much will they be paying? That guarantee in nominal dollars at your retirement date could be woefully inadequate 30 or 40 years into retirement, depending on the effects of inflation. (More below.)
  • GLWBs do let you change your mind. You can pull your assets out of the variable annuity at any time without punishing penalties. This is a good thing. Though you will have been paying additional fees and your assets will have been exposed to market risk, so there is no guarantee of getting all your original money back. The flexibility to change your mind costs you while you’re waiting. And there are other, simpler ways to keep your options open — for example splitting your assets between stocks and bonds in an investment portfolio, where you could withdraw principal if necessary.

Con: Inflation

Unfortunately, GLWBs come with a long list of real-world negatives, nearly all of which are surprising, given the sales spiel. Let’s start with perhaps the most surprising:

GLWBs do not protect against inflation. How can that be? Since the underlying annuity investments nearly always contain an equity component, why wouldn’t those stocks offer the traditional defense against inflation, growing along with other real world assets in at least nominal dollar terms? In a nutshell, it’s because the relentless rider and investment fees eat away at the growth that would have warded off inflation.

All of my trusted experts concur on this point. Jim Otar, in his book Unveiling the Retirement Myth, writes that guarantees like the GLWB convert longevity and market risk to inflation risk, which “is not handled well.” Joe Tomlinson agrees.

And Wade Pfau backs them up with his own extensive research, noting that GLWB guarantees are not inflation-adjusted and would not have been worth much in rolling periods of history. In another paper Pfau observes that inflation takes a “stark toll” on GLWBs and similar products, noting the steady decline in probability that the guaranteed income will keep up with inflation.

Con: Where’s the Upside?

Perhaps it’s not so surprising, now that we know the GLWB inflation protection is largely illusory, to learn that its “upside” is shaky as well. It sounds good on paper that you can keep some of your money in the market, through the underlying variable annuity, and participate in market upside. That’s the theory. But, in practice, it rarely works out:

Otar reports, for one of his historical analyses, that in every year between 1900 and 1938 portfolios ran out of money and the guarantees were activated. That’s a long time to go without any upside.

Pfau reports, for one of his studies based on current poor market conditions, that “After 25 years, the contract value for the VA/GLWB falls to zero in more than half of the cases.” And he notes, under normal market conditions, a typical 50/50 retirement portfolio may not even need a GLWB-style guarantee, since the failure rate would only be 1.4%, and the wealth after 30 years would match the initial wealth in real, inflation-adjusted terms.

Again, Tomlinson reminds us of the reason for this: costs, at least in more expensive variable annuity/GLWB products, eat up 50-100% of the equity premium — the growth advantage that stocks should have given us.

Con: Downside

Ok, so maybe the GLWBs’ promised upside and hoped for inflation protection aren’t what they’re cracked up to be. But at least we can count on the downside protection, the guaranteed income, right? Well that downside protection may be partly an illusion as well:

For starters, there may be a realistic possibility of insurer insolvency. Pfau notes “In the event that someone replicating GLWB withdrawals on his or her own runs out of wealth, an outcome that no investor has faced historically, then GLWB owners will surely worry about the risk of default for the insurer.” And Otar, writing in 2009, thinks rider fees are unsustainable. What happens if insurers realize that too late?

Let’s dig deeper into that “guarantee” behind a GLWB. It’s really only as good as the company standing behind it, and any insurers standing behind that company. Pfau observes that “rider guarantees may not be protected by state guarantee associations.” So exercise due diligence on both insurance company ratings, and the state guarantees behind them, if any, before signing up for any GLWB!

Lastly, GLWB withdrawal maximums are actually quite minimal, once adjusted for inflation. Pfau found that GLWB withdrawal amounts met or exceeded their initial levels in real (inflation-adjusted) terms in only about 5% of the many 30-year periods he analyzed. In many cases, the guaranteed spending after those long time spans may be only 20-30% of the original level, in real terms.

What kind of “guarantee” requires you to give away virtually all of your upside while promising only 20-30% on the dollar? You’re giving away higher income in the majority of cases, to gain the dubious certainty of keeping one-third of your money over the long haul.

Con: Complexity

So, much of the certainty promised by GLWBs melts away when examined more closely. But one aspect of these products is a definite: They are among the most complex financial instruments ever devised for ordinary consumers.

I’ve just spent several days pouring over research papers from some of the brightest minds around, documenting reams of simulations with these products. Still their behavior isn’t clear. The only thing I’m confident about is that there is no way vague claims such as “guaranteed,” “upside potential,” or “downside protection” can really be taken at face value.

Pfau, one of the leading experts working today, says that there are continuing “disagreements about the actuarial value of the guarantees” and that the products’ values are “difficult to objectively quantify.” If the experts can’t accurately value them given unlimited time and resources, how can the rest of us do it given a few hours for retirement planning?

Well, maybe the friendly annuity salesman can simplify things for us. I spent about 30 minutes recently talking to a pleasant, helpful, and confident-sounding Vanguard rep. Unfortunately he disagreed with Vanguard’s online variable annuity kit on at least two key terms of the GLWB contract. Was he right, or was the online information right? Or did I just misunderstand something?

However that question is answered, it’s more proof, if you need it, that these things are highly complex. Better find a sales rep or advisor you can literally trust with your financial life, or shop for simpler products….

Thanks, But No Thanks

As we’ve seen, a GLWB is a complicated attempt at providing downside protection with upside potential for retirement income. But for financial products, as for physical products (think appliances), it may be best to unbundle the desired features so you have access to simpler, more reliable solutions in each department.

At its core, the GLWB attempts to offer two benefits:

  1. Lifetime income. But an immediate annuity offers the same feature, while keeping the principal of customers who die earlier, so it will be cheaper for the same benefit. All other things being equal, a SPIA will maximize your income vs. a variable annuity product. (In one scenario I evaluated for a 65-year old single male, the SPIA would pay out 1.6% more of the premium annually.) Conclusion: you’ll get more downside protection for the same amount of premium with an immediate annuity.
  2. Market upside. But insurance companies have access to the same markets as the rest of us, with no theoretical advantage over any other seasoned investor. And the kinds of ‘black swan’ market events that you might need protection from, are the same kinds of events that are likely to threaten an insurance company’s solvency anyway. The withdrawals for a GLWB are artificially low in most cases and could be duplicated by a systematic withdrawal plan from an investment portfolio, with much more upside potential. (Note: that systematic withdrawal plan may tilt payments toward earlier years, when a retiree is most likely to be alive to enjoy them, at more risk of depletion later. But the total value being delivered is probably the same in either case, before fees.) Conclusion: a systematic withdrawal plan from an investment portfolio is superior to a GLWB for capturing market upside.

The unpleasant probability with any kind of annuity product is that, over time, the insurance company will take all your money and the payouts won’t keep up with inflation. Further, if the markets do average or well, you will have left money on the table. And if the markets do poorly, you will be hoping your insurance company doesn’t go out of business. In the end, what you’re really paying for, is peace of mind, and money management. As we’ve seen, that peace of mind is probably not warranted, and you can do the money management yourself, if so inclined.

In the final analysis, a GLWB is like paying for a gatekeeper who forces you to take lower withdrawals from your nest egg — the same level of withdrawals that would allow a cheaper systematic withdrawal plan to be sustainable. Or it’s like buying fast food — settling for the lowest-common-denominator in order to get certainty, a meal that you can count on being mediocre. Or it’s like asteroid insurance. In spite of recent events, it’s unlikely that there will be a major asteroid strike in our lifetimes. And, if that does happen, it’s likely that nobody will be around to pay the claim….


  1. Darrow,
    This is the type of article that wins you a position as a trusted voice in personal finance. For me, you’re able to provide an initial, secure foothold from which to begin my own critical investigations and analysis. Thank you for sharing your research in a format that is so thoughtful and easy to understand! As a fellow engineer, I really appreciate the work required to present complex issues with clarity and simplicity.
    Be well and keep writing!

  2. Wow – well done! I keep coming back to annuities and VAs with GLWBs, myself, hoping for good news. I really like the way you cited analysis by other leading lights in the retirement income sphere – Pfau, Otar and Tomlinson. There has to be a better way than the 4% “rule”, but I don’t think we’ve found it yet. Amazing! And this is for people who have managed to save enough to worry about this stuff. What about most of the rest of the population?

    • Agreed. I think we’ll come up with better guidelines than the 4% Rule, but I don’t think we’ll ever achieve certainty in this area, since it requires predicting the future. And, you’re right, it’s sobering that this entire discussion is reserved for those few who have saved enough to have options in retirement. The vast majority will need to buy immediate annuities with all their assets and even then won’t have enough to preserve their lifestyle. Thanks for the comment Kathy!

  3. I also spent a lot of time researching these products including all the varying articles online that seem to be so mixed in terms of yes, no or maybe. In the end I finally decided to try Vanguard (I wouldn’t with the new fee structure and reduced payout coming) because as you said, you can change your mind and pull it all out and start over with looking at different ideas again.

    Here is why I decided to purchase two (different dates so the 1 year basis aniversary is different allowing a possibility of not having all my money in 1 annuity and having the bad luck that the market falls 900 points that day) from Vanguard. If you use the Balanced allocation it is managed by Wellington. I would like to actually get in their funds but they have been closed for so long I viewed this as an opportunity.

    The other reason is my personal mental makeup. If I am getting money every month from this and the market drops like it did a few years ago and I see a 40% loss in my account, I won’t think about selling because I am still getting the original amount (or higher) but when my money is in my regular account and the market drops, I am always tempted to pull out some and, in effect, try market timing and we all know how well that works. Like I said, it is my mental issue, not necessarily a reason for someone else but it does work for me.

    I should note that I put a small percentage of my retirement money in these and still have CD’s, moneymarket accounts as well as index funds with Vanguard and a rental home so I feel like I can take a shot on these and if they are as bad as some say, I can get out.

    Great article and thank you!


    • Thanks for the comment Paul — that’s invaluable perspective, to hear from somebody else who made this decision. Interesting that you hedged your dates, and also used it as a path into Wellington. As for market drops, you understand yourself, and that knowledge is power. You also sound very well diversified, so you aren’t betting the farm on this or any other retirement income stream. I probably wouldn’t characterize any offering from Vanguard as bad — just sub-optimal for me, in this case. Thanks again for your thoughts!

  4. —Darrow

    This is the best advice and analysis I have ever seen on annuities. It’s unbelievable that our regulators have failed to make such information clearly and easily available. You’ve done a great thing!

    The most important thing you did was to place sufficient emphasis on issuer risk. Insurance salesmen will claim that no American issuer has ever gone completely bankrupt. Regardless of the truth or falsity of that statement, annuities by definition involve a long period of issuer risk. The precedents set by the failure of AIG (a backstop for many annuities) and the government’s failure to achieve meaningful reform in the aftermath of one of the worst eras of financial crises, issuer risk hiding in the Black Swan portion of that long period/curve is not just possible but probable. [Note that the government did not rescue AIG to save the annuities it was insuring or reinsuring but rather did so to save a preferred class of cronies. There’s no guarantee that affected retirees will be so lucky Next Time. And, Next Time is not a probability but a certainty.]

    • Thanks David. Agreed, regardless of the past track record for annuities, we are in unprecedented times with high debt, ultra-low interest rates, and a wave of retirees on the horizon. Other than pooling mortality risk, the insurance companies have no more special sauce than the rest of us in the markets. If it sounds too good to be true, it probably is….

      • David M

        That was another reason (see my previous response) for using Vanguard, the money in the annuity is actually still in my Vanguard account, not with an insurance company. (No, I don’t work for them, I just couldn’t find anything better)

  5. DK-

    Excellent blog post! I’m also a fan of Pfau and Otar, both for their detailed and objective analysis and their somewhat conservative bent. My plan is to provide guaranteed (read: highly reliable and not subject to market volatility) income streams for “essential” expenses (SS, pension, etc.), and use the portfolio for discretionary expenses. As part of that strategy, I’m considering an SPIA. So, your article is informative and, at least, does not discourage me from that approach.

    Regarding investments in Wellington, if that’s VWELX, I don’t know of any restriction. In fact, I just invested in it this month.


    * fellow engineer

    • Thanks M. I’m also narrowing the field of choices down to a SPIA (plus Social Security) for essential expenses. Then a conservative, diversified portfolio for inflation protection and upside. I’ve posted some thoughts on how best to mix the two. Now I’m starting to give some thought to the timing when to buy a SPIA or the rungs in a ladder of them. It’s a complex problem, but I think the solution will be easier if we keep the tools simple.

      • DK-

        I’m interested in reading your thoughts on timing for SPIA(s). Is that a future blog or, does it already exist here?


        • That’s in the future. I’m just starting to get my head around the issue. On first look, with interest rates at all-time lows, it seems to make little sense to buy a SPIA now, since rates on annuities, and inflation, can only go UP in the future. As long as by delaying you don’t eat into your assets so much (and annuity rates don’t rise enough), that you can no longer create the necessary income for essential expenses in the future.

          If anybody is aware of existing research on this problem, I’d appreciate the reference…

  6. Darrow – A great overview of the US annuity Market. Compared to the UK it seems there are many more complexities to comprehend. I think it is worth pointing at however much we may despise annuity providers for high charges, low rates or lack of protection from inflation, an annuity does (in the UK at least) offer a guaranteed income for life, meaning you can never ‘outlive’ your pension income.

    • Thanks Simon. Seems anything retirement-related is more complicated in the U.S. 🙂 I’m not dismissive of annuities in general. I expect a simple, immediate annuity will play a role in my retirement at some point. But I’m not yet sold on the benefits of variable annuities.