Dr.W Pfau: "Efficient Frontier for Retirement Income"

[speedbump101]

Well worth the read IMO.

"This paper outlines a different way to think about building a retirement income strategy that dramatically moves away from the concepts of safe withdrawal rates and failure rates. The focus is how to best meet two competing financial objectives for retirement: satisfying spending goals and preserving financial assets."

http://www.fpanet.org/journal/ABroaderF ... tFrontier/

SB...

[Levett]

Thanks, SB.

Wade Pfau thinks "out of the box."

I find all his contributions to be measured and thoughtful, and I much admire the way he invites--rather than dismisses--other viewpoints.

Lev

[555]

Earlier thread
viewtopic.php?f=10&t=103191

[LadyGeek]

Earlier, but not the same thing. The earlier thread ("An Efficient Frontier for Retirement Income") was a review of wade's paper.

The OP's link is to the finished product from the February 2013 issue of Journal of Financial Planning. The link in human readable form: A Broader Framework for Determining an Efficient Frontier for Retirement Income

To the new investors: The author is Wade Pfau, posting here as wade.

[speedbump101]

Considering that Wade is moving from Tokyo to Philadelphia this year I hope he can find time to attend Boglehead's 12 this fall.

Cheers,
SB...

[rj49]

Interesting article, in shifting thoughts away from the traditional stocks/bonds allocation that people use to adjust risk. Instead, he advocates annuities balancing out the risk of stocks and providing a relatively safe floor, which would allow a higher stock allocation (a big step, considering how retirees have flooded into bonds for perceived safety at the risk of diminished long-term wealth and portfolio survivability).

The other interesting aspect of the paper is that the risk of portfolio depletion in the former SWR studies aren't completely realistic, because most people have Social Security and house wealth to fall back on, along with the natural inclination to cut back spending if wealth is being depleted too quickly.

In the end, though, it leaves retirees shell-shocked after the last decade in a bit of a bind as to whether to follow the recommendation of a high-stock allocation with annuity/TIPS/pension buffers insuring that basic needs are met, or whether to still play it relatively safely with a balanced stock/bond portfolio, even if pensions/annuities/house wealth already serve the same risk-reduction purpose as traditional bonds. Then there's the recency of stock market gains and constant predictions and DH fearful posts of imminent collapse in bonds, which of course would encourage one to join the mythical Great Rotation back into stocks.

[Cut-Throat]

Interesting article, in shifting thoughts away from the traditional stocks/bonds allocation that people use to adjust risk. Instead, he advocates annuities balancing out the risk of stocks and providing a relatively safe floor, which would allow a higher stock allocation (a big step, considering how retirees have flooded into bonds for perceived safety at the risk of diminished long-term wealth and portfolio survivability).

The other interesting aspect of the paper is that the risk of portfolio depletion in the former SWR studies aren't completely realistic, because most people have Social Security and house wealth to fall back on, along with the natural inclination to cut back spending if wealth is being depleted too quickly.

In the end, though, it leaves retirees shell-shocked after the last decade in a bit of a bind as to whether to follow the recommendation of a high-stock allocation with annuity/TIPS/pension buffers insuring that basic needs are met, or whether to still play it relatively safely with a balanced stock/bond portfolio, even if pensions/annuities/house wealth already serve the same risk-reduction purpose as traditional bonds. Then there's the recency of stock market gains and constant predictions and DH fearful posts of imminent collapse in bonds, which of course would encourage one to join the mythical Great Rotation back into stocks.

Shifting away from a Stock/Bond Portfolio reeks of recency. I think I'll stay the course, and do nothing.

[bobcat2]

From the paper.

An interesting result is that fixed SPIAs dominate inflation-adjusted SPIAs in the retirement portfolio for this example. ...
With payout rates of 5.84 percent, the fixed SPIA payout is 51 percent larger than the 3.875 percent payout of the inflation-adjusted version. With inflation fluctuating around 2.1 percent, it takes almost 20 years, on average, for the income from the inflation-adjusted SPIA to grow larger. ...

Nevertheless, it is important to emphasize that clients with particular concerns about inflation may expect that the breakeven inflation rate is too low and may seek the additional protection provided by an inflation-adjusted SPIA.

Yes. I certainly would not count on the breakeven expected inflation rate of about 2.1% to hold over the next 20 years. This looks like a classic case of recency bias. Any inflation forecast over a 20 year period is likely to be inaccurate and the 2.1% breakeven expected inflation is far below the 3.3% or so inflation the US has experienced on average since 1900. (A long run average rate of inflation that would be much higher if it were not for the severe deflation from the mid 1920s thru the 1930s.) And there is no guarantee that inflation won't be higher than the historical average of 3.3% over the coming decades. Certainly there is no shortage of commentators warning that our large federal debt will lead to increasing inflation in the medium and long term. I believe it is reasonable to ask which type of life annuity (nominal or real) will dominate if inflation averages 4% or more over an extended period of time. Extended periods of inflation this high or higher have occurred over long periods since WWII. For example, my 2005 copy of SBBI shows inflation averaged 4.0% for the fifty year period from 1955-2004 and nearly 5% for the quarter century from 1975 to 2000. In short the aged retiree is assuming all the inflation risk when she purchases a nominal life annuity. Much better IMO that the inflation risk in grandma's retirement be placed with the insurance company rather than on the shoulders of granny.

BobK

[Browser]

I think Moshe Milevsky has already covered this ground with his extensive work and writings on the subject. See for example his book: "Pensionize Your Nest Egg: How to Use Product Allocation to Create a Guaranteed Income For Life."

[Rick Ferri]

Bottom line is this:

"The evidence suggests that optimal product allocations consist of stocks and fixed Single Premium Income Annuities (SPIAs), and clients need not bother with bonds, inflation-adjusted SPIAs, or VA/GLWBs. Though SPIAs do not offer liquidity, they provide mortality credits and generate bond-like income without any maturity date, and they support a higher stock allocation for remaining financial assets."

Wade is saying that retirees would be better of replacing their bond funds with immediate fixed annuities (not-inflation adjusted) and investing the rest in 100% equities.

Rick Ferri

[Cut-Throat]

Bottom line is this:

Wade is saying that retirees would be better of replacing their bond funds with immediate fixed annuities (not-inflation adjusted) and investing the rest in 100% equities.

Rick Ferri

What's your opinion of that advice?

[Garco]

My intuition is that laddering of SPIA's would be a cheaper way to provide for inflation projection than paying a special premium for an inflation-protected SPIA. I think Pfau referred to such an idea but I haven't seen it simulated. Say take three smaller 2-life SPIAs at ages 69, 74, and 79 rather than just one larger SPIA at age 69.

[VictoriaF]

Considering that Wade is moving from Tokyo to Philadelphia this year I hope he can find time to attend Boglehead's 12 this fall.

Cheers,
SB...

I did not know that Wade is moving to Philadelphia. It would indeed be great pleasure meeting him at BH12.

Victoria

[Rick Ferri]

Bottom line is this:

Wade is saying that retirees would be better of replacing their bond funds with immediate fixed annuities (not-inflation adjusted) and investing the rest in 100% equities.

Rick Ferri

What's your opinion of that advice?

While this idea is thought provoking, there are issues with this strategy.

1) Bonds can be liquidated in an emergency while SPIAs cannot. Stocks can be sold in an emergency, but what do you do in the middle of bear market when stocks are down 40% ?

2) There is no consideration to the behavioral aspects of a high equity allocation in a retirement portfolio. The present value of SPIAs do not show up on statements. So, they have no dollar value in a portfolio after purchase. This could create a problem. Despite having SPIAs, people will focus on the volatility of their investment portfolio in isolation. Investors will view a portfolio that has a high allocation to stocks as risky. They will not see it as a balanced portfolio with stocks and SPIAs (or as a balanced portfolio of stocks and fixed income). This may cause excessive plan failures during bear markets as investors sell because they're over their tolerance for risk in their investment portfolio.

3) Taxes are not being considered. SPIAs become less tax-efficient as the principal depletes later in life and all income becomes fully taxable. At the same time, required distributions from IRA accounts increase (which are all equity bow and thus worth more), thus increasing taxes even more.

Rick Ferri

[bobcat2]

There are several things I find odd about this paper, here are a couple.

Results are provided for a 65-year-old couple. Their Social Security benefit is equal to 2 percent of their retirement assets.

I would venture to guess that for at least 90% of 65 year old couples the Social Security benefit represents more (in most cases much more) than 2% of their retirement assets. I believe it is the case that about 60% of retirees get a majority of their retirement income from SS. That doesn't happen, or even come anywhere close to happening, if SS income is 2% of their retirement assets. For example, lots of couples will have about $30,000 between them in SS benefits, but few couples with about $30,000 in SS benefits will have $1,500,000 or more in portfolio retirement assets.

I also question whether individual bonds would be left out if they were laddered for consumption. It's not clear how bonds are being handled in the analysis, but if the bond holdings are treated as bond funds that is very different from laddered nominal Treasuries or (even better) TIPS, whose maturing principal each year is used for consumption.

BobK

[Browser]

Good points Rick. I don't think it makes any more sense to include SPIAs as a bond surrogate in one's portfolio than it does to include social security and pension income. These are income sources, not portfolio assets. Having them reduces the need to take on risk in the portfolio via high equity allocations in order to achieve a given withdrawal rate, so it affects one's allocation strategy. If these income sources are sufficient to secure one's basic spending needs, an individual might choose to shoot the moon with residual savings invested in a portfolio highly tilted toward equities. But clearly the individual does with with conscious awareness of the risk and that they might end up losing.

[bobcat2]

Another point I found odd about the paper is the point Rick has made very well as his point 1).

1) Bonds can be liquidated in an emergency while SPIAs cannot. Stocks can be sold in an emergency, but what do you do in the middle of bear market when stocks are down 40% ?

As Rick notes above there is so little financial flexibility in this approach.

In addition I question the motivation for only stocks and life annuities. Supposing that SS and life annuities provide you with a safe income floor, why wouldn't you target your portfolio of investment assets to achieve your aspirational retirement income goal, rather than putting the portfolio of investment assets totally into stocks and apparently wishing for the best. It is unlikely that all of your investment portfolio needs to be in risky stocks to meet your aspirational income level. And certainly, given the volatility of stock returns, it can't be the case that year in and year out 100% of stocks is required to meet the aspirational income goal, unless you are falling short of your aspirational income goal and simply hoping that equity returns will be high or very high.

BobK

[bobcat2]

My intuition is that laddering of SPIA's would be a cheaper way to provide for inflation projection than paying a special premium for an inflation-protected SPIA. I think Pfau referred to such an idea but I haven't seen it simulated. Say take three smaller 2-life SPIAs at ages 69, 74, and 79 rather than just one larger SPIA at age 69.

Purchasing life annuities in chunks over time is a fine idea. It is even a finer idea if at least half of the life annuities bought in chunks over time are inflation-protected. The nominal SPIA bought at age 79 cannot rescue the nominal SPIA bought at age 69 that has been ravaged by high inflation in the intervening 10 years.

BobK

[speedbump101]

I did not know that Wade is moving to Philadelphia. It would indeed be great pleasure meeting him at BH12.

Victoria

"Wade D. Pfau, Ph.D., CFA, will be joining the American College in the spring as a Professor of Retirement Income in their new Ph.D. program on Financial Services and Retirement Planning. He is a past selectee for the InvestmentNews Power 20 for people expected to shape the financial advisory industry, and is a recipient of Financial Planning magazine’s Influencer Awards."


http://tinyurl.com/a6uc7uc

SB...

[bobcat2]

One more point that I find troubling about this study is the tremendous amount of sequence of return risk there is in this retirement income strategy. If the investment portfolio is 100% stocks and the first and second years of retirement see a 50% or so drop in the stock market, then given that the retirees are making significant withdrawals from the portfolio, there is very little chance they will ever recover from that sharp market decline. Instead they will be in all likelihood go thru retirement with a much lower living standard than they were anticipating at the onset of their retirement.

BobK

[ourbrooks]

I note that Wade Pfau's Monte Carlo simulations will, of necessity, include cycle in which there are poor stock market returns early in retirement. Even under those conditions the stocks/annuities approach was less likely to fail than a stocks/bonds mix. The reason why is simple to understand: Bonds can also decline in both value and interest rate and this can occur at the same time as stock market decline; the panic of 2008/2009 is a good example. In contrast, the annuity income doesn't decline in nominal terms so there's never a "double disaster" of simultaneous poor stock and bond returns.

Life is better still if you follow the Merton/Bodie/Bernstein approach and have enough in annuities to cover basic expenses. Then, the effect of 2008/2009 is to cause retirees to defer expenses. Those who did this in early 2009 had to continue driving the old clunker until this year if they wanted to avoid a permanent loss of any kind, but, even if they bought the new car last year, they didn't do nearly as badly as someone who withdrew the same amount continously, regardless of their portfolio mix.

[bobcat2]

I note that Wade Pfau's Monte Carlo simulations will, of necessity, include cycle in which there are poor stock market returns early in retirement. Even under those conditions the stocks/annuities approach was less likely to fail than a stocks/bonds mix. The reason why is simple to understand: Bonds can also decline in both value and interest rate and this can occur at the same time as stock market decline; the panic of 2008/2009 is a good example. In contrast, the annuity income doesn't decline in nominal terms so there's never a "double disaster" of simultaneous poor stock and bond returns.

If the bonds were laddered TIPS and I-bonds then there was no decline in the value of the bond holdings. The retiree sold the I-bonds on schedule in 2008/2009 and used the maturing TIPS for retirement income in 2008/2009 as scheduled. She would have suffered no loss during those two years from either the TIPS or the I-bonds. She would have received from the TIPS and I-bonds exactly the amount of real income she was expecting. Nor was the rest of the laddered portfolio of TIPS or I-bond holdings affected by the downturn.

I see no reason why the retiree cannot hold a TIPS ladder and I-bonds in addition to both equities and life annuities. The life annuities offer more longevity risk protection than TIPS. (The risk of outliving your money.) The TIPS offer more spending flexibility than the annuities. You can always dip into next year's maturing TIPS for unexpected expenses this year with little duration risk. Dipping into next year's life annuity income is impossible, short of selling the annuity at steep discount. Do I hear the siren call of JG Wentworth?

BobK

[Munir]

Bottom line is this:

Wade is saying that retirees would be better of replacing their bond funds with immediate fixed annuities (not-inflation adjusted) and investing the rest in 100% equities.

Rick Ferri

What's your opinion of that advice?

While this idea is thought provoking, there are issues with this strategy.

1) Bonds can be liquidated in an emergency while SPIAs cannot. Stocks can be sold in an emergency, but what do you do in the middle of bear market when stocks are down 40% ?

2) There is no consideration to the behavioral aspects of a high equity allocation in a retirement portfolio. The present value of SPIAs do not show up on statements. So, they have no dollar value in a portfolio after purchase. This could create a problem. Despite having SPIAs, people will focus on the volatility of their investment portfolio in isolation. Investors will view a portfolio that has a high allocation to stocks as risky. They will not see it as a balanced portfolio with stocks and SPIAs (or as a balanced portfolio of stocks and fixed income). This may cause excessive plan failures during bear markets as investors sell because they're over their tolerance for risk in their investment portfolio.

3) Taxes are not being considered. SPIAs become less tax-efficient as the principal depletes later in life and all income becomes fully taxable. At the same time, required distributions from IRA accounts increase (which are all equity bow and thus worth more), thus increasing taxes even more.

Rick Ferri

Rick, You are far more knowledgeable in these matters than most of us are, but may I raise some questions about your three points:

1. One can have a separate fund for emergencies such a home equity line of credit (HELOC) instead of planning to use either stocks or bonds (bonds are becoming more volatile, anyway) for an emergency.

2. One needs to look at one's portfolio differently if Pfau's advice is followed. Behaviors can be unpredictable no matter how a portfolio is structured or named, but fortunately are more controllable than market forces. This should not be a significant contraindication for a rational person.

3. If the investor is a retiree in the RMD stage and his SPIAs are funded with tax-qualified assets, the SPIA payouts are considered RMDs. I do not see a tax consequence is such a sitaution. (I happen to be in this caegory).

Maybe psychologically it is too radical to abandon bonds totally, but how about having SPIAs be 80% of the fixed income portion of a portfolio? That's the message I would draw from Mr. Pfau's advice. I also still think that currently SPIA payouts are historically quite low and it may be best to change into them gradaully- but that's a controversial and separate topic which Mr. Pfau has addressed in a separate article.

[umfundi]

My intuition is that laddering of SPIA's would be a cheaper way to provide for inflation projection than paying a special premium for an inflation-protected SPIA. I think Pfau referred to such an idea but I haven't seen it simulated. Say take three smaller 2-life SPIAs at ages 69, 74, and 79 rather than just one larger SPIA at age 69.

Purchasing life annuities in chunks over time is a fine idea. It is even a finer idea if at least half of the life annuities bought in chunks over time are inflation-protected. The nominal SPIA bought at age 79 cannot rescue the nominal SPIA bought at age 69 that has been ravaged by high inflation in the intervening 10 years.

BobK

Not yet having read the entire thread, I think that SPIAs to provide your minimum needs is a fine idea. Laddering those SPIAs (buying them when you need them) is also a great idea, since the actual inflation outlook is low and inflation-protected SPIAs are relatively expensive. Actually, given the outlook for low interest rates, nominal SPIAs themselves are relatively expensive.

But, the cost of an SPIA must be compared to other current investments, not to the cost of an SPIA at some different time.

And, I have to add: Delaying SS for a year is equivalent to buying an inflation protected SPIA with survivor benefits at less than half the cost of the open market.

Keith

[Rick Ferri]

Rick, You are far more knowledgeable in these matters than most of us are, but may I raise some questions about your three points:

1. One can have a separate fund for emergencies such a home equity line of credit (HELOC) instead of planning to use either stocks or bonds (bonds are becoming more volatile, anyway) for an emergency.

People already look at these options regardless of their investment strategy. I have an instant access personal line of credit just for this reason. I am not retired, but it seems to me that taking out a loan in retirement isn't a comfortable choice. Loans have to be paid back and that means more cash flow is needed.

My point is that money invested in an SPIA is spent. That liquidity is no longer available. There is something to be said for the liquidity from bonds even though the returns are low.

2. One needs to look at one's portfolio differently if Pfau's advice is followed. Behaviors can be unpredictable no matter how a portfolio is structured or named, but fortunately are more controllable than market forces. This should not be a significant contraindication for a rational person.

That's easily said after the stock market has returned 135% since March 2009. It's was quite a different story in February 2009 when stocks were off more than 55% from their peak. Your retired, you have 100% in equity, and you just lost half your liquid net worth in 18 months. Are you going to stay that same rational person? Maybe, and maybe not.

3. If the investor is a retiree in the RMD stage and his SPIAs are funded with tax-qualified assets, the SPIA payouts are considered RMDs. I do not see a tax consequence is such a situation. (I happen to be in this category).

Yes, it does depend on where you put the SPIAs. I was bringing up taxes because it does need to be considered even though it may not affect everyone.

Maybe psychologically it is too radical to abandon bonds totally, but how about having SPIAs be 80% of the fixed income portion of a portfolio? That's the message I would draw from Mr. Pfau's advice. I also still think that currently SPIA payouts are historically quite low and it may be best to change into them gradually- but that's a controversial and separate topic which Mr. Pfau has addressed in a separate article.

How about 50% ? And that's only if you need that much.

One last item that I didn't mention is estate planning. SPIA money dies when you and your spouse die. It's not passed on to heirs. This may not be important to some people but it will be to others.

My bottom line is moderation. Too much of a good thing is not a good thing. Here is another example:

Zvi Bodie recommended 100% TIPS in his book Worry-Free Investing. It was all over the media. People thought it was a very good idea. But was it? Mike Piper, a.k.a. Oblivious Investor pointed out a fee issues with that plan, "I’m not arguing that a 2% real return is unreasonably optimistic. But relying on a given return (even from a portfolio comprised entirely of TIPS) is neither risk-free nor worry-free."

Rick Ferri

[YDNAL]

Bottom line is this:

"The evidence suggests that optimal product allocations consist of stocks and fixed Single Premium Income Annuities (SPIAs), and clients need not bother with bonds, inflation-adjusted SPIAs, or VA/GLWBs. Though SPIAs do not offer liquidity, they provide mortality credits and generate bond-like income without any maturity date, and they support a higher stock allocation for remaining financial assets."

Wade is saying that retirees would be better of replacing their bond funds with immediate fixed annuities (not-inflation adjusted) and investing the rest in 100% equities.

Rick Ferri

I see Wade addressing "the evidence" in Rick's quote, yet near the end of the paper he says:

Because using retirement-date assets to buy a fixed SPIA provides more income than needed until inflation sufficiently reduces the real value of the SPIA payments, and because excesses are invested in a portfolio of 100 percent stocks, a 100 percent allocation to a fixed SPIA can support more of the spending needs and result in the same median amount of financial reserves as a portfolio of 40 percent stocks and 60 percent bonds. Nevertheless, it is important to emphasize that clients with particular concerns about inflation may expect that the breakeven inflation rate is too low and may seek the additional protection provided by an inflation-adjusted SPIA.

Is he advocating "retirement-date Assets to buy 0/100 Stocks/SPIA," since superior to 40/60 Stocks/Bonds and provides more income than needed, then invest "excesses [income] in 100% Stocks?"

[Levett]

Rick Ferri wrote: "SPIA money dies when you and your spouse die."

That would depend on whether an individual or a couple did or did not choose a period certain feature and whether they did or did not die before the period certain feature had concluded.

If you die before the period certain feature is exhausted, your beneficiaries receive the commuted value of the remaining payments within the period certain.

I wholeheartedly agree about moderation, and with that comes a willingness to examine the possibilities of what Moshe Milevsky calls "product allocation."

Lev

[umfundi]

Rick Ferri wrote: "SPIA money dies when you and your spouse die."

That would depend on whether an individual or a couple did or did not choose a period certain feature and whether they did or did not die before the period certain feature had concluded.

If you die before the period certain feature is exhausted, your beneficiaries receive the commuted value of the remaining payments within the period certain.

I wholeheartedly agree about moderation, and with that comes a willingness to examine the possibilities of what Moshe Milevsky calls "product allocation."

Lev

Or, you can do it yourself by buying an SPIA plus Life Insurance.

I think Pfau's paper is excellent, except I have no clue what a VA/GLWB is. It sounds complicated and expensive.

With the demise of defined benefit pensions and more people needing options to generate lifetime income from a defined contribution lump sum, I think this conversation is both timely and necessary.

Keith

[Rick Ferri]

Rick Ferri wrote: "SPIA money dies when you and your spouse die."

That would depend on whether an individual or a couple did or did not choose a period certain feature and whether they did or did not die before the period certain feature had concluded. If you die before the period certain feature is exhausted, your beneficiaries receive the commuted value of the remaining payments within the period certain. Lev

Yes, but like an inflation-adjusted SPIA, the payouts are lower and that changes the probabilities. I'm not saying no...it's all up for consideration.

Rick Ferri

[Munir]

Rick Ferri wrote: "SPIA money dies when you and your spouse die."

That would depend on whether an individual or a couple did or did not choose a period certain feature and whether they did or did not die before the period certain feature had concluded. If you die before the period certain feature is exhausted, your beneficiaries receive the commuted value of the remaining payments within the period certain. Lev

Yes, but like an inflation-adjusted SPIA, the payouts are lower and that changes the probabilities. I'm not saying no...it's all up for consideration.

Rick Ferri

From the quotes I have obtained in the past, I found that a joint-survivor with a 10 or 15 period certain quote is still more generous than an inflation-indexed quote (I am 75 years old). It's probably best for an individual to obtain different quotes with different scenarios and from different reputable companies before making a decision on what type of an SPIA is best for what one's life expectancy is.

Thank you, Rick, for engaging in the conversation and for the thoughtful points you raise.

[555]

One problem with inflation-adjusted SPIAs (as mentioned in the article) is that they are priced unfavorably (in actuarial terms) compared to fixed SPIAs. Perhaps if the pricing improves that will shift the decision towards inflation-adjusted SPIAs.

The article seems to looking at constant (real dollar) spending. If you add in the need for sporadic large expenditures then that would shift towards needing more bonds/cash (essentially an emergency fund).

It would be interesting to see if there were ever such a thing as an "actuarially fair " variable annuity, whether that would make sense to hold.

But realizing that the conclusions come from a model with certain assumptions, I think the conclusions make sense.

[Levett]

Keith: I could not agree more with your observation that--

"With the demise of defined benefit pensions and more people needing options to generate lifetime income from a defined contribution lump sum, I think this conversation is both timely and necessary."

My sense about the future is that more people are going to have less of an opportunity to retire on their own terms--in large part, due to inherent weaknesses in DC plans (high expenses, poor options, behavioral errors). I've read little evidence that contemporary Americans, on the whole, are adept money managers.

My parents, on the other hand (born in the early 1900s), knew how to save and stretch a buck and still have some left over as an inheritance for the kids.

Lev

[Mel Lindauer]

....except I have no clue what a VA/GLWB is.

Hi Keith:

VA/GLWB = Variable Annuity with a Guaranteed Lifetime Withdrawal Benefit.

[umfundi]

....except I have no clue what a VA/GLWB is.

Hi Keith:

VA/GLWB = Variable Annuity with a Guaranteed Lifetime Withdrawal Benefit.

Mel,

Thank you. More precisely, I almost guessed the acronym, but I do not understand how it works. I'll do some digging, I am not asking for an explanation. My skepticism says it is one of those insurance products that cannot be explained or understood, but which plays to behavioral fears. Here is what Dr. Pfau says (my bold):

The analysis here is based on realistic low-cost versions for the various income tools available in the marketplace. For systematic withdrawals, I assume that investors use low-cost index funds, and the stock and bond funds are assumed to have a 0.2 percent annual fee. SPIA prices are from Vernon (2012), who obtained them using the Income Solutions platform at the start of April 2012. The 65-year-old couple can buy a 100 percent joint-and-survivors SPIA with a payout rate of 3.875 percent for the inflation-adjusted version and 5.84 percent for the fixed version. Pricing aspects of the VA/GLWB include a payout rate of 4.5 percent, annual fees on the VA contract value of 0.6 percent, an annual guarantee rider fee of 0.95 percent on the high-watermark benefit base, and an annual step-up feature to increase payouts if the contract value reaches a new high watermark. The assumed asset allocation for the VA/GLWB is 70 percent stocks and 30 percent bonds. Using low-cost versions for each tool allows for more direct and meaningful comparisons. Further advisory fees could be added if appropriate.

Huh?

Things like the "annual guarantee rider fee on the high-watermark benefit base" just happen not to be a topic of daily conversation around here.

Keith

[umfundi]

Keith: I could not agree more with your observation that--

"With the demise of defined benefit pensions and more people needing options to generate lifetime income from a defined contribution lump sum, I think this conversation is both timely and necessary."

My sense about the future is that more people are going to have less of an opportunity to retire on their own terms--in large part, due to inherent weaknesses in DC plans (high expenses, poor options, behavioral errors). I've read little evidence that contemporary Americans, on the whole, are adept money managers.

My parents, on the other hand (born in the early 1900s), knew how to save and stretch a buck and still have some left over as an inheritance for the kids.

Lev

Lev,

Yes. And, by the way, no one told the the young people: "The DB pension your parents had is kaput. You personally need to save up to twice as much as they did."

Keith

[Beagler]

Here's the link to VG's VA http://tinyurl.com/3fyyayw

[rj49]

I like Rick's recommendation for balance, perhaps with the basic needs floor comprised of a mix of TIPS and annuities. I think Zvi Bodie has mentioned that as well recently, with the decline in TIPS rates. That also gives you inflation protection above the stock allocation.

Rick is also right about questionable ability of an investor to tolerate the market swings of a high-equity portfolio, even if we believe that it's the best path to higher overall returns (especially with current prospects for fixed income), and even if we have the backup of various pensions and personal capital and house wealth as support for basic needs and inflation protection.

One other cause for concern over Wade's approach is the low current payouts for annuities, so I'd be reluctant to lock myself into a long-term annuity, just as I'd be reluctant to buy long-term bonds or CDs at today's rates. Also, according to this article, insurance companies are reducing annuity benefits and availability due to low interest rates and lower stock market returns, which could also be problematic for finding a secure annuity for the long term.

http://online.barrons.com/article/SB500 ... rticle%3D1

Then there's the usual reasons why retirees tend to dismiss annuities: high expenses, complicated rules, aggressive sales practices by agents, and the traumatic step of surrendering a large chunk of money and losing access to the principal forever (the popularity of dividend strategies shows how traumatic it is for many retirees to touch their principal).

[Cut-Throat]

Then there's the usual reasons why retirees tend to dismiss annuities: high expenses, complicated rules, aggressive sales practices by agents, and the traumatic step of surrendering a large chunk of money and losing access to the principal forever (the popularity of dividend strategies shows how traumatic it is for many retirees to touch their principal).

Add to that.....Is that by the time a person retires, their experiences with Insurance Companies are usually not in the "Plus Column'.

[speedbump101]

Wade is in the forefront of retirement planning research. As his blog entry suggests there is a lot more to be accomplished in the reality of a post DB pension / low yield world:

"I have written about this article before, and I see that the Bogleheads have kindly started a new thread about it which includes some critical comments. This framework is still not complete, as it doesn't include taxes, deferred income annuities, or the ability to build a bond or annuity ladder, but this provides my best efforts to layout a framework for thinking about retirement income. There are so many limitations to safe withdrawal rates, as I discussed in the introduction of the article, and this article provides my best effort thus far to get around those limitations."

http://tinyurl.com/bxn5eac

SB...

[dbr]

Then there's the usual reasons why retirees tend to dismiss annuities: high expenses, complicated rules, aggressive sales practices by agents, and the traumatic step of surrendering a large chunk of money and losing access to the principal forever (the popularity of dividend strategies shows how traumatic it is for many retirees to touch their principal).

Yes, the annuity conundrum, as it is called is often mentioned in academic circles.

One should comment, however, that single premium immediate annuities must be distinguished from all others:

1) Costs are not directly relevant to the purchaser as the payout may be taken at face value and comparison made to alternatives.

2) SPIA's don't have complicated rules.

3) SPIA's generally are not the windfalls to agents that other annuity types are and therefore are not marketed aggressively.

4) Losing principal forever is a reason to eschew an annuity and requires some clear thinking about one's needs and priorities. A lot of that thinking may not be as rational as it should be. As mentioned above, the same issue applies to much of the thinking about alternatives, except perhaps worse.

[Mel Lindauer]

Then there's the usual reasons why retirees tend to dismiss annuities: high expenses, complicated rules, aggressive sales practices by agents, and the traumatic step of surrendering a large chunk of money and losing access to the principal forever (the popularity of dividend strategies shows how traumatic it is for many retirees to touch their principal).

Yes, the annuity conundrum, as it is called is often mentioned in academic circles.

One should comment, however, that single premium immediate annuities must be distinguished from all others:

1) Costs are not directly relevant to the purchaser as the payout may be taken at face value and comparison made to alternatives.

2) SPIA's don't have complicated rules.

3) SPIA's generally are not the windfalls to agents that other annuity types are and therefore are not marketed aggressively.

4) Losing principal forever is a reason to eschew an annuity and requires some clear thinking about one's needs and priorities. A lot of that thinking may not be as rational as it should be. As mentioned above, the same issue applies to much of the thinking about alternatives, except perhaps worse.

Yes, SPIAs can certainly play a role for some investors. Here's a column I did for Forbes on SPIAs.

http://www.forbes.com/2010/07/29/single ... dauer.html

[Levett]

The GAO has put out a lengthy report on the risks and benefits of Guaranteed Lifetime Withdrawal Benefits.

http://www.gao.gov/products/GAO-13-75

Lev

[Peter Foley]

Bobcat2 wrote

One more point that I find troubling about this study is the tremendous amount of sequence of return risk there is in this retirement income strategy. If the investment portfolio is 100% stocks and the first and second years of retirement see a 50% or so drop in the stock market, then given that the retirees are making significant withdrawals from the portfolio, there is very little chance they will ever recover from that sharp market decline. Instead they will be in all likelihood go thru retirement with a much lower living standard than they were anticipating at the onset of their retirement.

BobK

This struck me as the weak point as well. Multiple SPIAs purchased over a 10 year span offer some protection against inflation. However, a bad market early in one's retirement might leave one short of the funds needed to purchase subsequent SPIAs.

[Munir]

The importance of SPIAs for providing steady income in retirement is the main issue here, and I hope that various minor reservations and "what if" arguments will not detract from the value of the ideas advocated by Mr. Pfau. Rick Ferri and Mel's article kept matters in perspective. What percent of the fixed income portfolio should be in SPIAs and when to purchase them can be decided on an individual basis. Some of the counter arguments are mixing up fears and fallacies that relate to other annuities and don't apply to SPIAs.

[mickeyd]

While I can see Wade's logic in offing bonds for annuities, it would make no sense for someone like me. We have 3 pensions as well as 2 SS incomes coming in monthly and that pays for most of our daily/monthly needs. With my 60/40 AA mix, I control my future and have a solid base, some indexed (COLA'd), that already gives me plenty of guaranteed income.

Giving up any flexibility for insured guarantees does not seem to make sense to me.

[555]

"Bobcat2 wrote"

"One more point that I find troubling about this study is the tremendous amount of sequence of return risk there is in this retirement income strategy. If the investment portfolio is 100% stocks and the first and second years of retirement see a 50% or so drop in the stock market, then given that the retirees are making significant withdrawals from the portfolio, there is very little chance they will ever recover from that sharp market decline. Instead they will be in all likelihood go thru retirement with a much lower living standard than they were anticipating at the onset of their retirement.

BobK"

"This struck me as the weak point as well. Multiple SPIAs purchased over a 10 year span offer some protection against inflation. However, a bad market early in one's retirement might leave one short of the funds needed to purchase subsequent SPIAs."

I disagree. The higher payout from SPIAs reduces the percentage you need to withdraw from stocks, so they're better able to withstand market downturns.

Also, if you're planning to buy SPIAs in the future it is prudent to hold the funds in fixed income so the funds are there if you want to buy an SPIA during a stock downturn.

[dbr]

While I can see Wade's logic in offing bonds for annuities, it would make no sense for someone like me. We have 3 pensions as well as 2 SS incomes coming in monthly and that pays for most of our daily/monthly needs. With my 60/40 AA mix, I control my future and have a solid base, some indexed (COLA'd), that already gives me plenty of guaranteed income.

Giving up any flexibility for insured guarantees does not seem to make sense to me.

In the case of a theory that includes annuities in the set of options, pensions and SS count as annuities. In short, you are already there. In this case there is no false accounting as pensions and SS are annuities, absolutely. If one wants to quibble one could engage in the nuances of such things as what the SS guarantee really is, what the respective tax liabilities are, and so on. A real impetus for what Wade is doing is the disappearance of actual pensions as job benefits most places in our economy.

[Rick Ferri]

I am in the same situation as mickeyd. By age 67, I'll have a military pension, two small pensions from corporate America, my Social Security and my wife will have spousal SS income. This is enough annuity income for us.

Rick Ferri

[Levett]

Mickeyd and Rick Ferri have surely identified situations where an SPIA makes no sense whatever, IMHO.

But there are others with entirely different circumstances.

And as time goes by, with the decline of DB pensions and possible changes to the determination of COLAs for SS, the use of SPIAs and inflation-indexed annuities may play a greater part in the construction of a retirement income program.

Lev

[555]

NEWS FLASH!!! SPIAs are somewhat similar to Social Security and pensions. Who knew?