Why So Critical on Annuities? by Allan Roth

[DriftingDudeSC]

Well written.


http://moneywatch.bnet.com/investing/bl ... blog-river

[Allan]

There are some really low cost variable annuities thru Vanguard and Fidelity that to me don't seem to be much different than regular indexed mutual funds. The primary drawback I see is the gains being taxed at ordinary rates, but this is mitigated somewhat if you withdraw during retirement.

I'm not saying they are great but they might have small place in a portfolio, and for business owners and high risk professions they are creditor protected in some states.

Allan

[Mel Lindauer]

Allan is one of the "good guys" who tell it like it is with no sugar coating. He posts here on the forum from time to time, and he'll be with us again this year at Bogleheads 8 in Dallas at the end of this month. So, if you're joining us in Dallas and you have Allan's book, be sure to bring it with you since you'll be able to get it signed during the book-signing session.

[fishndoc]

Roth always makes a lot of sense with few words.

I agree there are uncommon situations when a low-cost VA has its place, especially if you are high tax bracket, and need more tax-deferred space for bonds, REITS, etc.
Also, the protection against law suits could be extremely important.

The downsides are well know, especially costs. Also, as we now know, insurance companies are not bankruptcy-proof.

Wayne

[jegallup]

Unless I missed it, the article criticizes "annuities" without distinguishing between single-premium immediate annuities and variable ones. Aren't they two very different things?

[Mel Lindauer]

Unless I missed it, the article criticizes "annuities" without distinguishing between single-premium immediate annuities and variable ones. Aren't they two very different things?

He says this at the beginning of his column:

These insurance products come in every variation under the sun, with two types being variable annuities and equity indexed annuities. They all claim to have so called “guarantees” against adverse situations.

So it should be clear he's talking about variable annuities and equity indexed annuities.

[jegallup]

Unless I missed it, the article criticizes "annuities" without distinguishing between single-premium immediate annuities and variable ones. Aren't they two very different things?

He says this at the beginning of his column:

These insurance products come in every variation under the sun, with two types being variable annuities and equity indexed annuities. They all claim to have so called “guarantees” against adverse situations.

So it should be clear he's talking about variable annuities and equity indexed annuities.

Read it but not closely enough. Thanks.

[rcshouldis]

I like the article especially the quote about the only way to make money with annuities is to get rid of (sell) them.

[Levett]

Here's Mr. Roth's opening paragraph.

"Anyone who has been a reader of mine will not be at all surprised when I say that I’m not a big fan of annuities as a place to park our nest eggs. These insurance products come in every variation under the sun, with two types being variable annuities and equity indexed annuities. They all claim to have so called “guarantees” against adverse situations."

What do we learn? 1. He doesn't like any kind of annuity in general. 2. He chooses to single out two kinds for the purposes of his (limited) argument.

Once he's made it clear he's not a "big fan of annuities" (clearly implying any kind of annuity), he then goes on to his "market return" argument--never indicating precisely what "market" he has in mind.

Does Mr. Roth want, for example, to compare the 10-year returns, say, of the TIAA Traditional Retirement Annuity with that of the S & P 500 (assuming the latter is what he means by "market return")? I should think not because such a comparison would undermine his argument. VFINX in the last 10 years produced an annual return, according to Vanguard, of -0.86. Just the TIAA Traditional guarantee of 3% (and, yes, that guarantee has held up) might cause Mr. Roth to feel a mite uncomfortable. The ten-year annual return of the Traditional Annuity has been +6.19%, which just might cause an episode of GERD for Mr. Roth if one uses the "market return" comparison.

As for his final sentence, I find it to be malarkey. Here's what he writes:

"There is only one way I know to make money with annuities — sell them!"

Let's think for a minute about the above nonsense. A sixty-year old buys a $100000 SPIA with a 15-year period certain (to protect his beneficaries) and lives to the ripe old age of 90. Did that sixty-year old "make money?" Does Mr. Roth mean to imply that the 30 years of payments were all principal? Oh, dear.

Mr. Roth, I fear, is blinded by his admitted prejudice--he's not a "fan" of any kind of annuity. And, not surprisingly, he doesn't appear to to have considered a lot of responsible research on different kinds of annuities and how to use them.

Somebody introduce Mr. Roth to Dr. Milevsky, Dr. Bodie, Dr. Poterba, and the list could go on. Indeed, send him to Bob's Financial Website. http://bobsfiles.home.att.net/AnnuityLinks.html

Bob U.

P.S. I trust Mr. Roth didn't mean to imply in his first paragraph that folks would park "their entire nest egg" in an annuity (or collection of annuities). That would be both downright foolish and irresponsible.

[KyleAAA]

Variable annuities make sense in certain circumstances, like for instance if you have no room in your tax-deferred accounts but still want to own REITs. Their returns come primarily from reinvested dividends (as opposed to capital appreciation), so being taxed as regular income when you withdraw isn't that big of a disadvantage. Ditto with high-yield bonds. But for general retirement investing I don't think they are good vehicles. At least with a 401k you get the upfront tax writeoff.

[sscritic]

Here's Mr. Roth's opening paragraph.

"Anyone who has been a reader of mine will not be at all surprised when I say that I’m not a big fan of annuities as a place to park our nest eggs. These insurance products come in every variation under the sun, with two types being variable annuities and equity indexed annuities. They all claim to have so called “guarantees” against adverse situations."

What do we learn? 1. He doesn't like any kind of annuity in general. 2. He chooses to single out two kinds for the purposes of his (limited) argument.

It depends how you interpret the word "park" (I added the bold to your quote). If park means to set aside for a period of time until you are ready to withdraw funds, then a SPIA is not an annuity for parking funds and is excluded from his discussion. So he is not against all annuities, just "parking" annuities.

[Levett]

Hi sscritic--

Well, since Mr. Roth will be in Dallas perhaps someone will ask him what he means by "park"--among other things. Bob U.

[Mel Lindauer]

Hi sscritic--

Well, since Mr. Roth will be in Dallas perhaps someone will ask him what he means by "park"--among other things. Bob U.

I think it's rather clear that he was talking specifically about variable annuities and indexed annuities, and he specifically said so.

Even if he hadn't said so, two points should make it clear:

1. You don't "park" money in as SIPA.
2. You don't talk about "market returns" when you're talking about as SPIA.

[Levett]

I certainly agree with your points 1 & 2, Mel.

I don't think Mr. Roth is nearly as clear as you are. Bob U.

[VictoriaF]

Here's Mr. Roth's opening paragraph.

"Anyone who has been a reader of mine will not be at all surprised when I say that I’m not a big fan of annuities as a place to park our nest eggs. These insurance products come in every variation under the sun, with two types being variable annuities and equity indexed annuities. They all claim to have so called “guarantees” against adverse situations."

What do we learn? 1. He doesn't like any kind of annuity in general. 2. He chooses to single out two kinds for the purposes of his (limited) argument.

It depends how you interpret the word "park" (I added the bold to your quote). If park means to set aside for a period of time until you are ready to withdraw funds, then a SPIA is not an annuity for parking funds and is excluded from his discussion. So he is not against all annuities, just "parking" annuities.

A park is also a place with trees, and birds, and nests, and nest eggs -- a permanent kind of a place ... like an SPIA

Victoria

[Mel Lindauer]

I certainly agree with your points 1 & 2, Mel.

I don't think Mr. Roth is nearly as clear as you are. Bob U.

I'll be sure to let him know he has to be more careful to define the object(s) of his scorn when I see him in Dallas, Bob. Obviously several folks took it one way (as to include all annuities), while others thought he was clearly taking about a restricted set of annuities which he clearly mentioned at the start of the article.

[HueyLD]

I also read Allan Roth's blog with interest. His blog seems to dismiss the benefits of ALL annuities, not just some annuities. Maybe he forgot to mention the benefits of SPIA that were discussed extensively on the Bogleheads.org forum.

I agree with his assessment of variable and equity indexed annuities.

p.s., Bob U., go Spartans!!

[Levett]

Huey,

Last time I looked them Spartans is losing.

I guess I better count my blessings--er, SPIAs. 8)

Bob U.

[nisiprius]

Unless I missed it, the article criticizes "annuities" without distinguishing between single-premium immediate annuities and variable ones. Aren't they two very different things?

It drives me bananas but it's the way it is. The most common usage always gets shortened. The unqualified word "annuity" now pretty much means "variable annuity."

Similarly the unqualified word "fund" has come to mean "mutual fund" in some contexts. (Think of the statement "I'm adding funds to my brokerage account.")

Similarly unqualified word "chauvinism" now means "male chauvinism," which I think is a real loss, since I thought the original coinage was a brilliant way to capture the nature of male chauvinism. Chauvinism really means excessive and bellicose patriotism, basically a synonym for "jingoism," after a mythical French Napoleonic soldier named Chauvin.

For a while the unqualified word "transistor" was used to denote a transistor radio.

etc. etc. These things affect me like a fingernail scraping on a blackboard, which come to think of it is a dated idiom... they affect me like a dried out marker squeaking on a whiteboard...

[Allan Roth]

Thanks for the great comments. Here is a little clarification.

First, I am talking mainly about variable annuities and equitable indexed annuities.

Second, any comparison of a fixed annuity (such as an equity indexed annuity) to a stock market fund (such as an S&P 500 fund) is an apples to oranges comparison. Compare a fixed annuity to a long-term fixed income investment with similar duration.

Third, I think Vanguard variable annuities are a very good tool to roll over a more expensive annuity via a 1035 exchange.

Fourth, I do think a single premium immediate annuity can be useful but is bought far too often. In most cases, one must be at least 70 and in good health for the mortality credit benefit to outweigh the costs of an indirect investment. Finally, if we do hit hyper inflation, the real payment will decline significantly. I one buys an inflation adjustment, then the payment often becomes too low.

When one buys an investment through an insurance company, they are buying it through an unnecessary intermediary. One can easily disintermediate by investing direct.

I'm familiar with Milevsky and Bodie's arguments. I don't agree in total.

We can debate in Dallas. I'm sure I will learn.

Thanks again.

[Levett]

Thanks, Mr. Roth, for responding to the thread.

Since I'm the guy who mentioned Milevsky and Bodie, let me add a couple of other names: Jim Otar (http://www.retirementoptimizer.com) and William Reichenstein. For the latter, see the following that picks up on a lot of research. http://74.6.239.67/search/cache?ei=UTF-8&p=william+reichenstein%2Fsingle+premium+immediate+annuities&fr=yfp-t-701&u=bobsfiles.home.att.net/Dr_Reichenstein_Study.pdf&w=william+reichenstein+single+premium+immediate+annuities&d=eb82Ct29TdHc&icp=1&.intl=us&sig=XkS6URYmOLSn4v70Kd3VTQ--

Pleasurable confession: I'm a TIAA retiree (in year 10 of retirement) who holds both a SPIA & VIA, in addition to an unannuitized portfolio. Obviously, I have a dog in the hunt.

Best wishes and sorry I'll miss you in Dallas. Bob U.

[kimm16]

I bought a Vanguard Variable Annuity back in 1996 ....there weren't that many 'other' choices at that time...initially $125,000. I used various sub-accounts....this went to $265,000 just before the crash....it went down to $210,000 during the crash....it's now at $249,000.....I used Mel's advice at various points on which sub-accounts to use, which was extremely helpful......I liked the fact that no matter what the market did, the original $125,000 was guaranteed.
If the original $125,000 had been left in an individual account, it would have affected Social Security taxes, and would have done so each year.....no regrets on this one.

Jerry

[Allan Roth]

Jerry,

I'm fairly sure your $125,000 was a death benefit, not a living benefit. By being in a VA, you have shifted this gain from a long-term capital gain to ordinary income. You have also given up your step-up basis upon death.

No argument that Vanguard has the best VAs, however.

[Mel Lindauer]

Jerry,

I'm fairly sure your $125,000 was a death benefit, not a living benefit. By being in a VA, you have shifted this gain from a long-term capital gain to ordinary income. You have also given up your step-up basis upon death.

No argument that Vanguard has the best VAs, however.

I agree with Alan on both counts, Jerry. It's probably just a death benefit which guarantees your heirs will get back no less than your original investment. Tough way to collect, huh? Additionally, these death benefits are way overrated, since, as you've seen, long term, your account should grow, negating the "benefit".

And if you're going to have a variable annuity, you'd be hard pressed to do better than Vanguard. It's a great place to do a 1035 transfer to if you already have an existing high-priced annuity.

[kimm16]

It was my understanding when I first bought an annuity with Vanguard in 1996 that the original principle $124,000 was not taxable upon future withdrawal.....this was confirmed when I called a Vanguard specialist this morning....they told me that the cost basis of $124,000 was made with after tax dollars, and is not subject to 'ordinary' taxes at any point in the future. However, there is no choice when taking a withdrawal, ALL amounts above this must be taken FIRST. As a portion of the total amount will be going to a recognized charity as a Beneficiary, ANY amounts received by the charity will be received tax free.

Jerry

[HueyLD]

However, there is no choice when taking a withdrawal, ALL amounts above this must be taken FIRST.

Jerry,

That statement is not 100% correct. If you annuitize your VA, then your payments will consist of both return of principal and earnings. However, if you make withdrawals w/o annuitization, all withdrawals will come from earnings first.

[kimm16]

Yes, this is true ....they also told me the same thing....if you annuitize, the total amount becomes one package.....I should have added that statement originally.

Jerry

[LH]

Well written.


http://moneywatch.bnet.com/investing/bl ... blog-river

Great article.

It comes down to this:

Costs matter and there is no free lunch.

Only the federal government can "guarentee" performance via abililty to print money. Those costs that get eaten up right off the bat and later, MATTER, they will decrease the performance. The annuity does not really have much to offer in reality. The more they offer, the higher the rate of failure risk in the future of the product. The money has to come from somewhere. These are not like SPIA, where the excess money comes from people who "die early"....... Where does this money come from to make up for all the COSTS??? Nowhere, its imaginary. Its a product that pays the people who sell it NOW, and gives you a private company backed "guarentee" for later.

But like these credit default swaps, CDO, mortgage tranches, that were guarenteed, heh, when the bad times happen, there

will not be the money there to make the guarentee good. (AIG fails).

There is no money fairy, there is no free lunch. these companies invest in stocks and bonds, just like we can in low cost funds---- or place derivatives bets/use hedge funds and expectedly lose relative to the market. They cannot make up the fees lost, they cannot recoup your costs expectantly. They are not guarenteed to survive long term if the bad times, which you are purportedly insuring against, happen.

Yes in markets, there is volatility, if you want to avoid that use nominal bonds or tips. Yes there are periods where any sort of investment, will outperform stocks/bonds.

But you have to follow the money, and here, there are big upfront costs the salesmen use to live off of, that the insurance company pays its shareholders, etc. There is no magic to make up for that.

SPIA are different. The "magic" is people die and lose their money(or pay an extra fee to not have to), and you get some of it.

Costs matter and the only entity that is "guarenteed" long term is the US government.

0)make a complex product, with lots of riders/options to offer that end up paying the insurance company more
1)make guarentee you might not be able to fill - bankruptcy.
2)hide the fee expense
3)take the fees now, go to mexico, put your kid through college, whatever
4)invest the remainder if responsibly, in the same type of investments available to your client without the fees, with the same return, if not, in derivatives and risk blowup aka AIG.

I do not see much else there fundamentally.

[ualdriver]

Jerry,

I'm fairly sure your $125,000 was a death benefit, not a living benefit. By being in a VA, you have shifted this gain from a long-term capital gain to ordinary income. You have also given up your step-up basis upon death.

No argument that Vanguard has the best VAs, however.

Nice series of articles, Allan. I enjoyed reading them and the posts below the articles in the blog.

[Maverick]

The problem with most arguments against variable annuities is that they deal solely with the academic/mathematical perspective but not with the behavioral perspective. Let's say we have 2 hypothetical investors. Both had $100k in late 2007. One of the investors was in a traditional taxable account (Investor 1); the other had a variable annuity with a guaranteed principal option (Investor 2). Let's also assume Investor 2 incurred expenses that were 2% greater than those incurred by Investor 1.

Fast forward to March 2009. Investor 1's balance is now $75,000. Investor 2's balance is $72,500 (as a result of the higher expenses). Investor 1 is convinced that things are going to get worse and finally capitulates. Investor 2 is worried, but knows the account guarantee is still in place, so there are no changes. Now in October, Investor 2's balance is $90,000. Investor 1's balance is $79,000 because he finally got back into the market, but he missed the big increase because he was worried we were going back down. So, who is better off?

The purpose of this is not to argue for or against VAs. It is simply meant to provide a different perspective. While most people on this forum would not capitulate like Investor 1 (and thus wouldn't need, or want, a variable annuity for that purpose) - many investors would (and did, if you look at cash balances in March). There is more to the story than simply looking at expenses - you need to look at behavior and the cost of influencing behavior vs. the cost of not influencing that behavior.

[nisiprius]

The problem with most arguments against variable annuities is that they deal solely with the academic/mathematical perspective but not with the behavioral perspective. Let's say we have 2 hypothetical investors. Both had $100k in late 2007. One of the investors was in a traditional taxable account (Investor 1); the other had a variable annuity with a guaranteed principal option (Investor 2)... Investor 1 is convinced that things are going to get worse and finally capitulates. Investor 2 is worried, but knows the account guarantee is still in place, so there are no changes.... The purpose of this is not to argue for or against VAs. It is simply meant to provide a different perspective.

Your observations are just observations about conservative investments and risk tolerance, and don't apply specifically to VAs. In your hypothetical example, investor 1 chose a portfolio that exceeded his risk tolerance and it turned out poorly.

Investor 2 chose a form of investment that, due to the guaranteed principal option, was more conservative; less risk, and, due to the cost of the option, less reward. Because investor 2 did not exceed his risk tolerance, it turned out better.

If investor 2 has simply used a larger bond allocation... or committed a portion of his portfolio to a pure-insurance SPIA and therefore knew that his minimum retirement expenses were guaranteed... it could well have had the same effect.

The real question is not whether some investors are better served by conservative investing, but whether a VA is good form of conservative investment. It is analogous the familiar argument whole life is good because it forces savings, sidestepping the question of whether there are other forms of forced savings that might be better.

[Maverick]

Your observations are just observations about conservative investments and risk tolerance, and don't apply specifically to VAs. In your hypothetical example, investor 1 chose a portfolio that exceeded his risk tolerance and it turned out poorly.

Investor 2 chose a form of investment that, due to the guaranteed principal option, was more conservative; less risk, and, due to the cost of the option, less reward. Because investor 2 did not exceed his risk tolerance, it turned out better.

If investor 2 has simply used a larger bond allocation... or committed a portion of his portfolio to a pure-insurance SPIA and therefore knew that his minimum retirement expenses were guaranteed... it could well have had the same effect.

The real question is not whether some investors are better served by conservative investing, but whether a VA is good form of conservative investment. It is analogous the familiar argument whole life is good because it forces savings, sidestepping the question of whether there are other forms of forced savings that might be better.

First, a loss of 25% from peak to trough is representative of a balanced portfolio that could reasonably have been expected to have been used an investor that had a fairly low tolerance for risk but that needed to fund a 30-year retirement. We can quibble about a couple percent here or there, but that wasn't really the point.

In the example I provided, both investors had the exact same asset allocation. Your comment is true in that one investor ended up with a more conservative portfolio as a result of the guarantees provided by the VA. The ultimate point was that the guarantee proved to be a behavior modifier, which enabled Investor 2 to maintain discipline in his/her plan. That was really the only point I was trying to make.

All that being said, your presumption is that these hypothetical investors can meet their long-term goals with portfolios that are allocated in a manner that is consistent with their respective risk tolerances. Unfortunately, one's risk tolerance is a function of a number of variables that might change on a daily basis (mood, amount of sleep, etc.), so allocating to a changing target is difficult (risk tolerance is generally too high/low at exactly the wrong times). Also, most people would prefer to actually achieve their goals rather than ratchet down their goals to match a "risk tolerance". That might not be the best approach, but it is reality. At the end of the day, a portfolio needs to be allocated in a manner that will provide the highest probability of an investor meeting his/her goals. If that allocation proves to be too "risky", the investor has to either accept that fact, change his/her goals, change the time frame for achieving his/her goals, or change the savings/withdrawal rate.

With respect to a SPIA - their irrevocable nature proves to be somewhat of a disincentive. True, one can obtain riders so that some amount can be repaid to one's heirs, but then the payout ends up approaching that of a rider on a VA, but the VA provides more flexibility.

Back to the main point - I was simply trying to provide a real world example of the potential benefit of a VA for a certain type of investor. The nature of the benefit defies math/academics, but it is very real in that it has an impact on the behavior pattern of the investor. Once again, this may not be applicable to participants on this forum.

[speedbump101]

Since the other thread on this is locked, I just wanted to point out the responses Allan Roth got to this article... In my mind the answers Allan came up with were a given... We all know there is no free lunch, and Anderson appeared to be providing that... What I did find astounding however was the series of responses that were printed afterwards... Allan handled it very well, however I wouldn't say the same for Mr. Anderson.

SB...