The New Math of Retirement

[Independent]

I'll agree that "most people" reduce their spending as they move through retirement. Note that "most people" are retiring after age 60, which is unusual for this board.

However the Newsweek article uses a lousy statistic - households aged 55-64 vs. households aged 65-74. The first group has more people per household, more workers, more cars, more mortgages, and higher SS taxes. Those differences have nothing to do with four phases of "moving through retirement" in the article.

Similarly, I looked at the 18 planning shortfalls listed in the referenced Society of Actuaries study and didn't notice "declining spending during retirement" on the list.

I do agree that good planning involves identifying types of spending (basic vs. fun, recurring vs. one-time) and planning income to match the spending.

I'm not surprised that a financial adviser wouldn't be a big fan of longevity insurance. But, I'd like to understand the adviser's financial incentives (one time 2% commission on the annuty vs. lifetime 1% on retained assets?) before I'd take his advice.

 

[Independent]

Has anyone seen any studies on the 6% rule that was listed in this article?

Retirement: The New Math of Personal Finance - Newsweek.com

If I were going to analyze a "spend more early" strategy, I'd rather set aside the amount I want to spend on "fun stuff, soon after I retire", then see what the traditional 4% SWR on the remaining money gives me. I feel that I understand that better than doing a Monte Carlo study on not adjusting withdrawals a stochastic rate of inflation.

 

[donheff]

My main hobbies are reading, surfing the internet (including coming places like this) and playing World of Warcraft. The only one of those costs that goes up with retirement is reading and I do think that the book budget will increase.

My reading went up dramatically in retirement but my costs dropped to near zero. I order almost all of my books online from the library and then get an email when they are in. Even small town or county libraries are often networked so you can pull from other counties in the state. I have literally saved thousands of dollars.

 

[MasterBlaster]

Has anyone seen any studies on the 6% rule that was listed in this article?

Retirement: The New Math of Personal Finance - Newsweek.com

There have been a number of studies and lots of case studies and empirical evidence that spending in retirement is not constant year after year. These studies show that the 4% SWR (constant spending) model is probably not realistic of retirees spending patterns.

We have Bernike's Reality retirement model:
Are you saving too much for retirement? - MSN Money

http://www.early-retirement.org/for...pending-decrease-with-advanced-age-25280.html

There has been much written on this topic. For more stuff just google something like "retirement spending models"

Analysts refer to a model as failing if one would go broke during a retirement. Mostly retirements fail due to either a prolonged bear market early in the retirement (burning through the nest egg) or too high a withdrawal rate.

Note also that if you don't retire into a raging bear market that goes on for years and years that the SWR could be significantly higher than the always safe 4%. remember that 4% SWR is the SWR is to get you through those very tough prolonged bear markets. So if you retire during anything but some severe bear markets, then a 6% spending spending model just may work fine.

here is Gummi's model showing that if you retire at various times with a diverse portfolio that you could draw 6% or even significantly higher (depending on the markets). This is kind of the flip side of Bernstein's "retirement Calculator from Hell" series that show depending on when you retire you could indeed draw significantly more each year. Note that for a 30 year retirement, on one of these "good" years to retire you could withdraw 8% or higher and still never go broke.

For the plot the horizontal axis is the years you would spend in retirement. The vertical axis is the maximum fixed withdrwal rate you could use if you retired in that particular year. The green dot example shows if you need a 15 year retirement starting in 1960 that you could take out just under 12% and never go broke. Most of us probably want longer retirements so the far right data of 30 year retirements is more interesting. Note that if (for example) you retired in 1960 that over 30 years you could withdraw around 9% and never go broke. If you were lucky enough to retire in 1950 then (over 30 years) you could take out just over 15% per year and never go broke.

It goes without saying that a fixed withdrawal rate higher than 4% is not "safe". However, to be safe under this model - you most certainly will die with tremendous unspent sums in your account. Again remember that the 4% SWR allows for the worst markets early in the retirement. If that horrible market doesn't occur then you will most certainly leave money (perhaps lots and lots of money) unspent when you go. This then gets into all of the variations on non-fixed (variable) withdrawal rates. This topic goes on and on.

for the other side of this issue and to inject some caution here is a link to Bernstein's articles on the Retirement calculator from Hell series: http://www.early-retirement.org/for...ment-calculator-from-hell-articles-32828.html

 

[walkinwood]

Here's a link to Ty Bernicke's paper on "Reality Retirement"
Reality Retirement Planning: A New Paradigm for an Old Science by Ty Bernicke, CFP®

This has turned out to be a great thread!

 

[tuixiu]

This has turned out to be a great thread!

Agreed. It really does do wonders for how long I have to work if I click that "reality retirement spending" dealie in firecalc.

 

[haha]

Agreed. It really does do wonders for how long I have to work if I click that "reality retirement spending" dealie in firecalc.

It's the Big Rock Candy Mountain, Folks

YouTube - Burl Ives - Big Rock Candy Mountain

 

[Independent]

There have been a number of studies and lots of case studies and empirical evidence that spending in retirement is not constant year after year. These studies show that the 4% SWR (constant spending) model is probably not realistic of retirees spending patterns.

We have Bernike's Reality retirement model:
Are you saving too much for retirement? - MSN Money

I'll agree that people generally spend less as they "move through retirement". But, when Bernicke uses BLS statistics, he is missing some important facts.

First, the difference between spending at age 50 and spending at age 70 clearly involves big things like mortgage, children at home, and work expenses (note that FICA is included as an "expense" in the BLS survey).
Most people on this board know that they need to adjust for those things individually in their planning, we aren't comfortable with a simple smooth decrease like Bernicke models.

Then the difference between spending at 65-74 and spending at 75+ also involves family size. Unfortunately, Bernicke picks up the data for "households", without adjusting for the fact that people die between those ages. In his data, the younger group has 1.9 persons per household, while the older has 1.5. Nor does he point out that the first group has more mortgages and pays more FICA taxes. I think we should explicitly plan for number of people, whether we'll be paying a mortgage, and whether we'll be working when we think about our spending after 65.

Now, even when I adjust for those things, there still is a residual drop off in spending. People spend less on transportation, meals away from home, entertainment, clothing, and houshold furnishings. But the overall drop is not the 3% per year that Bernicke models.

I think that for people who have never thought about retirement spending, it's important to challenge the thoughtless recommendation of 70-80% of pre-retirement income, then flat forever. For people who post here, it's better to look at specific expenses and see which will decrease (or increase) as we age.

 

[MasterBlaster]

Independent:

I don't disagree with your post. However I have a couple of comments...

It's a model. Maybe it fits maybe it doesn't. Evidently you don't believe in it. I'm not so sure that I do either. Nonetheless it gets us to think out of the constant spending model box. perhaps we can find some little nugget there.

I suspect that as the original article stated that spending is highest just after retirement. We fix up the house to get it just like we want or buy that perfect retirement home. Maybe we buy new car(s). We travel to wherever we thought that we'd always wanted to go. We buy a bunch of stuff to support our hobbies.

Then we settle into a sustainable spending pattern until our health makes us stick close to the house. Finally we may have some dramatic spending on medical/senior care in the last years.

Oh sure there may be periodic bursts of spending for whatever along the way. but the pattern just may be as described. Note that the constant spending model just doesn't fit too well the described pattern. Perhaps there is a better model.

 

[RonBoyd]

It's the Big Rock Candy Mountain, Folks

Burl Ives does the absolute worst version of this song. He leaves out all the important stuff like "little streams of alcohol flowing down the rocks" and didn't even mention that it's where "they hung the Jerk who invented work."

YouTube - BIG ROCK CANDY MOUNTAINS - Harry Mac McClintock - 1928