Monday, January 16, 2012

On the Pros and Cons of GLWBs

Guaranteed Lifetime Withdrawal Benefit (GLWB) riders, also known as Guaranteed Minimum Withdrawal Benefit (GMWB) riders, provide a seemingly attractive offer for retirees: they protect on the downside with a guaranteed income for life, they include upside potential with step-up features for growing markets, and they are contracts which can be ended with remaining assets returned.
I recently explored the performance of GLWBs compared to systematic withdrawals from one’s savings using US historical data in the column, “GLWBs: Retiree Protection or Money Illusion?” at Advisor Perspectives. Subsequently, Bob Powell included an interview with me in a column he wrote for MarketWatch, “Variable-annuity guarantees disappoint over time.” After reading some harsh criticism in the comments there and at SmartMoney, I do still stand by all the points I made in the interview.
Variable annuities and GLWBs are controversial. Some observers are set dead against them, while others think much more highly about their potential. In my initial column, my conclusions tended a little bit toward the anti-GLWB side.
I am not pushing any agenda and am only trying to determine what potential role GLWBs may have helping retirees to meet their goals. We do need to be careful, because in these trying and volatile financial markets, we might be too eager to grasp onto something offering assurances that might really be too good to be true. Skepticism is warranted.
Some of the feedback about my column and the interview suggest that my somewhat negative conclusions were unfair, and that GLWBs really are more useful than I gave credit. Today I’d like to further explore four of the pro-GLWB arguments I’ve been reading.
1. I only considered the rather plain GLWB rider offered by Vanguard. Other GLWB riders offer all sorts of interesting features such as guarantees to double one’s wealth and so on. These alternatives are much better.
This issue is the main reason for writing today. My initial response to this is that while other products may offer more attractive guarantees than Vanguard, they must also certainly be accompanied by much higher fees as well. But this gets to heart of a rather confusing issue for GLWBs.
If two different GLWBs offer the same initial guaranteed withdrawal amounts, it seemingly doesn’t matter what their fees are. Or at least, the role of fees may not be clear for someone who is just starting to learn about GLWBs. But the fees do matter.
Joseph Tomlinson* has written his inaugural monthly column at Advisor Perspectives on this issue. His column is, “Income Annuities versus GLWBs: A Product Comparison.” He argues that the way to assess the impact of fees from a GLWB is to look at the remaining account contract value as time passes. One of the features of GLWBs is that they can be terminated with assets returned. But higher fees will mean less is available to be returned. Fees on the variable annuity and GLWB rider reduce the contract value of the remaining assets, which also reduces the likelihood of the step up features for the upside potential kicking in. Market returns have to be so strong that the account value can still increase despite of the fees eating away at the account value.
To repeat, two GLWBs may offer the same guaranteed initial withdrawals, but if one has higher fees, then the contract value of the remaining assets will be depleted more quickly, which also has the implication that there will be fewer opportunities for upside gains should markets perform well.
Joseph Tomlinson’s new column explores this. He compares systematic withdrawals, Vanguard’s GLWB offering, a cash-refund annuity, and a more typical GLWB product with much higher fees than Vanguard. He estimates the average remaining bequest value in order to make a comparison about the impact of fees for different strategies offering approximately the same payouts. It’s a way to show the cost of the fees in more understandable dollar terms.
What he finds, naturally, is that one can’t really expect to enjoy downside protection along with upside potential. Risk is commensurate with reward. After fees for the guarantee, GLWBs end up behaving more like fixed income than like stocks. The potential for upside growth is constrained by the degree of downside protection. You can’t have your cake and eat it to. There’s no free lunch. You get the point.
2. By guaranteeing income, GLWBs will help their owners to maintain a higher stock allocation and to avoid the mistake of panicking and selling stocks after a market decline.
This is a popular argument from pro-GLWB sources. Essentially, the argument is that people who would otherwise be terrible investors prone to buying high and selling low or who might be permanently scared away from stocks will be suddenly transformed into proper stay-the-course buy-and-hold investors with a healthy dose of stock holdings on account of the GLWB guarantee.  Examples of this point can be seen in letters to the editor about my column at Advisor Perspectives here and here
I’m not sure about this. Especially, GLWBs still have an account balance, and assets can be returned, so people might still be prone to worrying about the account value and making the same mistakes.
What’s more, I’m not sure how one would even go about determining whether this point is true. Even if GLWB users make fewer behavioral mistakes with their asset allocations, we cannot know if that is due to the guarantee, or due to the different psychological make-up of GLWB users. This is called the sample selection problem. We would need a randomized experiment in which some people are randomly required to buy GLWBs and some are not, in order to properly demonstrate whether this point is true. Relying on anecdotal evidence and memories about client behavior is not enough.
This is a question which really requires further data exploration. It may or may not be true, and I would like to know! If anyone reading this does have such data, I’d be quite interested to hear about it and to potentially work with you on some data exploration.
3. Perhaps because future returns may be more gloomy than what we’ve observed in the past, GLWB providers may have been too generous and are now starting to leave the business.
This is certainly an interesting and important point. I do want to explore GLWBs further with simulations that may be more realistic. As I’ve argued many times, US financial market returns in the 20th century were quite high by international standards, and perhaps higher than we can reasonably expect to see in the future. With more realistic (i.e. lower) return assumptions, GLWBs may come out looking much more attractive from a buyer’s perspective (which is bad news from the seller’s perspective). Of course, if a GLWB is so great that it ends up driving the seller out of business, this is bad news in terms of still getting one’s guaranteed income. But this is an interesting point that I want to explore further in the future.
4. GLWBs can provide their users with comfort and peace of mind, knowing that they have a guaranteed income source no matter how bad things get.
I don’t dispute this (subject to the insurance provider remaining viable). In my interview, this is what I meant which I mentioned that retirees must decide on a personal level whether they value the guarantees more than the fees. This really is a personal decision and I can’t give an answer for anyone. Only you can decide.
One commenter about the interviewer said that this point is stupidly obvious. Yes, it is chapter 1 of principals of economics that you only buy something if you value it more than the price. But the problem with GLWBs is that it is so hard to determine the real world meaning of the price. The benefit of the GLWB is also rather abstract as well.
In that regard, I do think the analysis provided in Joseph Tomlinson’s new column can be very helpful in quantifying the real world impact of the GLWB fees, and so help retirees to decide whether the guarantees are “worth it” for them.
Again, I am not condemning GLWBs. I suppose I am just saying that I would like to see greater transparency about the meaning of the fees so that people can make better decisions about them.
If you are thinking about purchasing a GLWB, make sure to do your homework and understand all of the features and fees.
*I have been getting to know Joseph Tomlinson as we share common interests both in comparing retirement income strategies, as well as in applying utility analysis to retirement income, which looks to find a balance between the safety of lower withdrawal rates and the added life enjoyment of being able to spend more with a higher withdrawal rate. You can find more of his writings at his webpage, and he will also have a research article I recommend checking in the upcoming February issue of the Journal of Financial Planning called, “"A Utility-Based Approach to Evaluating Investment Strategies."