Life insurance and Annuity selling by Wade Pfau?

[urn2bfree]

I admire Mr. Pfau's mathematical and scholarly approach, but I sometimes wonder if he is not working hard to prove the value of the stuff he and his colleagues are selling.

Improving Retirement Outcomes with Investments, Life Insurance, and Income Annuities - Forbes

I am trying to see where he missed something, but I can't find it. He ignores the effect of inflation on the annuity but it still seems like his method allows for significantly greater spending early in retirement- when you probably want to spend more..


Sent from my iPad using Early Retirement Forum

 

[ETFs_Rule]

This guy is defending whole life insurance. That's pathetic.

It takes 5-15 years for the typical whole life policy just to break even.
33% of policy holders have dumped their policy after 5 years
50% of policy holders have dumped their policy after 10 years
70% of policy holders have dumped their policy after 20 years
77% of policy holders have dumped their policy after 30 years


Expected returns according to whitecoatinvestor...

Guaranteed cash surrender value at 70 years old: 4.98% return.
Guaranteed cash surrender value at 81.1 years old: 3.02% return.
Guaranteed cash surrender value at 100 years old: 2.83% return.



These are pathetic returns. Basically you have to die young to get higher returns.

 

[meierlde]

After 30 years if you have a participating policy, you can convert it to a paid up policy, to at least pay burial expenses. Of course new participating policies are scarce now since many of the mutual companies decided they needed stock to pay their executives.
But the point I would make is if your in the 15+ year range consider either going paid up or moving to term (both options are likley in the policy document). The old participating policies did pay much better if the company did well. A 45 year old participating policy shows a cash value due to dividends of nearly 5 x the guaranteed value. (Note that holding it as insurance is a way to beat taxes, if paid out after you die no tax is due unless estate tax is due. This policy is paid up next year so I will pay the final 4 premiums, however a 30 year old policy paid up at 85 I will convert to paid up insurance next year.

 

[urn2bfree]

I agree that I have never seen anyone convincingly defend whole life (actually whitecoatinvestor.com had a column today showing the rare exceptions when it might make sense) but I am trying to find the error in Pfau's math and I have not yet found it.


Sent from my iPad using Early Retirement Forum

 

[aja8888]

Start a thread over at Bogleheads touting whole life policies and see what happens....

 

[ETFs_Rule]

Start a thread over at Bogleheads touting whole life policies and see what happens....

+2

 

[gcgang]

I think his math is off in the 401k illustrations. I don't understand how he comes up with a 401k value only 13% less in the whole life example. The annual contribution is $9000 and the non whole life is over 50% more, about $14,500. Why isn't the non whole 401k more than 50% higher?

I guess it comes from being able to invest the 401k more aggressively. I'm not sure I understand exactly what he said, but I think he counted whole life cash value as a bond.

In general, I think his approach is worth considering. On this forum there is much discussion about AA, and withdrawal rates have been beaten to death, but maybe we don't really appreciate what RISK can truly be.


Sent from my iPad using Early Retirement Forum

 

[pb4uski]

Here's what I don't get. Table 1, page 19. The scenario where the 35 yo buys term and invests the difference the annual investment in the 401k is $14,281. Where they buy a whole life policy they invest $9,000 annually. So with BTID they invest in their 401k 159% of what they invest where they buy whole life.

But after 30 years, the 401k balance for BTID is only 124%, 115% and 112% for the 10th percentile, median and 90th percentile outcomes. If you apply the same annual Monte Carlo returns to these different scenarios, shouldn't the resulting value of the 401k in 30 years for the invest $14,281 annually scenario always be 159% of the invest $9,000 annually scenario?

The other nit that the tax rate in retirement isn't lower than the 25% rate used during the working years.

Edited to add: I see gcgang noticed the same thing.

 

[mathjak107]

i think because odds are anyone with enough money to invest , buy the life policy and annuity is likely not be in a lower tax bracket at retirement.

especially if you take their average tax bracket over a 40 year career that spends decades increasing up to the final brackets.

i kind of see his point not reducing the retirement brackets here.
it all boils down to trading market and interest rate risk for longevity risk. what's your poison ?

 

[mathjak107]

This guy is defending whole life insurance. That's pathetic.

It takes 5-15 years for the typical whole life policy just to break even.
33% of policy holders have dumped their policy after 5 years
50% of policy holders have dumped their policy after 10 years
70% of policy holders have dumped their policy after 20 years
77% of policy holders have dumped their policy after 30 years


Expected returns according to whitecoatinvestor...

Guaranteed cash surrender value at 70 years old: 4.98% return.
Guaranteed cash surrender value at 81.1 years old: 3.02% return.
Guaranteed cash surrender value at 100 years old: 2.83% return.



These are pathetic returns. Basically you have to die young to get higher returns.

guarantees cost money . comparing the un-guaranteed returns of investing with the guarantees of insurance products is not apples to apples.

i insure my car and home but i don't expect a return on that money or even what i paid in back.

annuity's and life insurance are all still insurance and not investments, they go along with investments.

 

[pb4uski]

While I'm not sure I agree with his method or his numbers, I think I now understand the lack of symmetry in the 401k balances after 30 years. In the BTID scenario, he assumes a roughly 89/11 AA at age 35 that grades to 48/52 at age 65. In the whole life scenario he is goosing the AA within the 401k assuming the whole life cash value is a bond equivalent so the 401k gradually becomes much higher stock allocation as the cash value grows.

While I agree with this in theory, I'm not sure any 35-65 year old would really act that way, but it would be sensible if they did. I include my whole life CSV as a bond equivalent in my AA, but is it a much smaller percentage of my total.

The other odd thing he does is to gross up the life insurance CSV for taxes. He states that the life insurance CSV at the end of 30 years is $257k, but he grosses it up to $343k ($257k/(1-25%)). The $257k is an IRR of 4.12% based on the premiums paid, which seems sensible. However, I don't yet see that the gross up make sense.

The oddest thing he does is to convert the 401k balances to after-tax balances (presumably by multiplying the actual 401k balances by (1-25%t)) and then he takes 3.5% of those amounts as the withdrawals to support living in retirement. I certainly don't think of my withdrawals that way and while my tax rate was 25% while I was working, it is much lower now.

I think what he is doing with taxes probably biases the result, but I'm still studying it.

Another major flaw in his thinking is that he is focused on age 65 income but the alternatives which are a mix of a SPIA and withdrawals will result in much lower spending power over time that the alternative that is based only on withdrawals, because the withdrawals will increase each year for inflation but the SPIA will not, so 10 years and more into retirement, the couple's spending power will be squeezed more by inflation by having the SPIA.

The other assumption that seems unrealistic is 1.59% ER on the 401k. I know my mega 401k ER was a lot lower.

 

[brewer12345]

The evil Dr. Pfau ("Mr. Bigglesworth!") is an academic who is fascinated with the minimization of risk of all kinds. Of course overpriced products with what appear to be strong guarantees appeal to his machinations! Those of us who live in the real world can safely ignore his ramblings. I think he also ignores the filthy underbelly of the life insurance industry, which is what you would be pressing up against face-first for decades if you hang your hat on the products he is plugging into his models. No, thank you.

 

[jkern]

Another issue is purchasing Term Life from age 35 to 65. As your children grow up and your wealth increases, the need for life insurance may end much earlier.

 

[mathjak107]

i see the logic behind it.

the annuity provides current income and covers essential expenses in good or bad markets but it dies when you do . the life insurance guarantees money for heirs . it makes perfect sense

your own investing covers the non essential expenses ,wants and inflation protection .

pfau , bernstein and ed slott are all in agreement on that very structure producing the greatest success rates and comfort.

don't forget that the sequence risk in retirement is the greatest variable. because of it the same average return according to moishe milwevsky's paper can have a 15 year difference in how long the exact same average return will have your money last.

it can range from zero to more than you started with.

i can see their point because you need to keep a lot less powder dry and un-spent when dealing with sequence risk then when dealing with guarantee money..

by trading some sequence and market risk for longevity risk the use of cash flow can be more efficient ..

think of it this way . if we ruled out the two worst retirement time frames from ever happening the 4% swr would actually been a 6.50% swr. but to get past the data from those 2 time frames requires a whole lot of money kept in reserve.

so what happens at 4% is the income stays constant but the pile left at the end goes up and down .

if you could get gurantees on that pile at the end , and guarantees the income will never go down, sequence risk effect is a whole lot less on your investments so you can spend more than you can if you had to keep so much extra in the bucket to cover all that sequence and market risk..

 

[samclem]

If a retiree makes it past the first few years of retirement with an intact portfolio (thus avoiding a "sequence of return" failure), then it seems the three major risks remaining are:
1) Inflation (failure of the portfolio value-- and available withdrawals-- to keep up with it). Buying a non-inflation indexed annuity (as Pfau suggests) doesn't help with this.
2) General market failure/asteroid/societal breakdown: Insurance companies are not immune to these failures.
3) Longevity risk: Not a major risk for ER types, since we aren't counting on spending down our portfolios significantly for decades. The difference in historic failure rates/sustainable withdrawal rates between a 30 year horizon and a 40 year horizon falls well within the uncertainty range of the whole calculation. And, I prefer delaying SS as a lower cost means of buying insurance against longevity risk.

 

[mathjak107]

key word always being "IF " when there are no guarantees in place.

 

[samclem]

key word always being "IF " when there are no guarantees in place.

Are you referring to the "If the insurance company remains solvent?" And about those guarantees . . . .A belly laugh.

 

[mathjak107]

nope , that is the smallest "IF " of them all. even 2008 did not effect one annuity or life policy. in fact agg's insurance business was unscathed.

then you have various state guarantees . states require insurers to just absorb failed companies.

not reallty a worry. it may be a concern but never a worry like markets and rates would be. failed insurers likely accounted for a tiny tiny part of failed retirements if even at all.

 

[pb4uski]

While I don't like whole life as an "investment" as used in the paper, the risk of the insurer NOT performing on either the life insurance or the SPIA is so negligible that it can be ignored.

 

[MichaelB]

If a retiree makes it past the first few years of retirement with an intact portfolio (thus avoiding a "sequence of return" failure), then it seems the three major risks remaining are:
1) Inflation (failure of the portfolio value-- and available withdrawals-- to keep up with it). Buying a non-inflation indexed annuity (as Pfau suggests) doesn't help with this.
2) General market failure/asteroid/societal breakdown: Insurance companies are not immune to these failures.
3) Longevity risk: Not a major risk for ER types, since we aren't counting on spending down our portfolios significantly for decades. The difference in historic failure rates/sustainable withdrawal rates between a 30 year horizon and a 40 year horizon falls well within the uncertainty range of the whole calculation. And, I prefer delaying SS as a lower cost means of buying insurance against longevity risk.

+1

Another risk - making a bad choice (doing something dumb). With a portfolio, that is selling out at the bottom and watching a market recover from the sidelines, suffering a permanent loss of capital.

With insurance and annuities, the focused sales effort to change the policy, shifting into one that generates new fees and profit opportunities for the issuer. This is common, well covered on the media, and I would imagine a risk that grows as one ages.

 

[mathjak107]

for sure , it is almost non existent and certainly 2008-2009 would have wiped out plenty if it could.

 

[samclem]

then you have various state guarantees . states require insurers to just absorb failed companies.

That works in the case of an isolated company that fails for an isolated reason (malfeasance, uniquely poor decisions, etc). It does nothing but spread the damage and force other teetering companies over the edge in the case of a systemic problem. The weight of the increasing number of companies falling into the pit assures even the strongest companies will be drug under. 2008 is far from as bad as things can get, I'm sure most people here are structuring their affairs to weather such a storm. And yet, during that time, insurance companies were not sanguine about their ability to withstand the gale, instead approaching the government for help. It's all "solid as a rock" when selling policies to people, some companies (and their trade groups) didn't sound that way behind the scenes when push came to shove. We can all choose which version to believe today.

 

[mathjak107]

you have a whole lot more to worry about than insurers.

i think you are dwelling on something very remote.

 

[samclem]

you have a whole lot more to worry about than insurers.

i think you are dwelling on something very remote.

And I think you are brushing it aside. "Remote risks" are what insurance is >supposed< to reduce. Yet, by "betting" our future--decades of future-- on one company (or industry) instead of a more diversified approach, I'm not sure we reduce those remote risks.

Annuities have a place. If I were 80, in great health, had a portfolio that might not be able to support my basic needs for the next 10-20 years, I might buy one and hope for the best.

 

[mathjak107]

our entire system of safe withdrawals is based on remote chance. the entire reason we don't spend 5-6% is because there were 2 remote times, otherwise the average rate is 6.50% .

but we don't just because of remote chances and that is why we have insurance against certain things. guarantees will always cost .