cut off point

[aaronc879]

I use 80 which I think is more than necessary since on male member of my family on either side ever lived past 76. I'm tempted to use 70 as an end to my savings and live off SS after that since I live very cheaply.

 

[Rich_by_the_Bay]

Assuming we are talking about intervals of 25 years or more, the consequences of using too large a number are probably pretty small for purposes of FIRECALC type calculations. That is, a 50 year horizon won't fail that much more often than a 35 year calculation. Example (FC default settings): 50 yrs, $1mm, $40k withdrawals = 86% chance of success. Change it to 35 years and it's 92%, not that different considering the widely different assumptions.

So I use a life expectancy of 85 years, but if I live to 90 or even 95 the implications are not that different.

 

[Nords]

Example (FC default settings): 50 yrs, $1mm, $40k withdrawals = 86% chance of success. Change it to 35 years and it's 92%, not that different considering the widely different assumptions.
So I use a life expectancy of 85 years, but if I live to 90 or even 95 the implications are not that different.

I sure hope so. I'd hate to be wrong, especially if the only other alternative is working 10 more years to achieve 33x instead of 25x or 22.3847x.

I guess it's worth re-running Bill Bernstein's existential take on the planning:
http://www.early-retirement.org/for...ment-calculator-from-hell-articles-32828.html

(Heh. Accidental literary joke. "Hell", "existential", get it? Heh-heh.)

 

[BigNick]

I think that everyone needs to factor in Bernicke (discretionary expenditure declining with age, although FIREcalc's interpretation which starts this at age 56 means I have to lie about my age to it!), but also the cost of keeping ourselves in the best quality diapers in the nursing home.

My current planning calls for FIRE at age 54 (DW will be 57), 120% of our pre-retirement expenditure until 62, 100% until 70, 75% until 80, then 40% (sit in front of TV and dribble) until , then a couple of painful years for the inheritance watchers. The key to all that seems to be getting to 80 with enough in the COLA'ed pensions to pay for the dribbling, and 70K or so to pay for the immediate needs annuity when the dam breaks.

 

[Nords]

I think that everyone needs to factor in Bernicke (discretionary expenditure declining with age...

I like his observations and his analysis and I really really want to believe, but he does admit that this doesn't include medical-related end-of-life spending like the Alzheimer's full-time memory-care unit.

 

[fritz]

When calculating your retirement, what age cut off point do you use?

Looking at the TSP calculators I see they go all the way to 115.

I consider 80 the far point.

I really do need to get all finances in order for 5 10 20 years.

Lot of different projections given here - thought I'd give you the U.S. Census Bureau's idea of age projections. FYI - as of 2009 < 2% at +85, project 2035 will be < 3% at +85, and project 2050 just over 4% at +85.

 

[Bikerdude]

So I use a life expectancy of 85 years, but if I live to 90 or even 95 the implications are not that different.

I tend to agree with Rich. However, I plan to skid sideways into the funeral home parking lot at 85, put my last quarter in the parking meter, hop in the casket and close the lid.

 

[walkinwood]

And how low did the WR go? And how much to kick it up to a 98-99% success rate?

-ERD50

There was a thread earlier about just how low the 4%/95% SWR went. The worst case was pretty bad. However, it wasn't "dog food" bad. We have enough slack in our budget.

I don't worry about bumping the probability up from 90+ (I don't remember what Clyatt's exact number) to 98-99%. See the William Bernstein paper that Nords posted.

 

[kyounge1956]

(snip) I guess it's worth re-running Bill Bernstein's existential take on the planning:
http://www.early-retirement.org/for...ment-calculator-from-hell-articles-32828.html (snip)

(snip) I don't worry about bumping the probability up from 90+ (I don't remember what Clyatt's exact number) to 98-99%. See the William Bernstein paper that Nords posted.

Despite having read the Calculator from Hell articles, I've always felt that a portfolio with a success rate of 100% in modeling really is better than one with 80%. I wonder if this is based on a misunderstanding of either the models, the articles or both. As I understand it, Bernstein says any result over 80% is meaningless because he estimates that our way of life itself has only about an 80% chance of survival. But ISTM that the portfolio that survives everything the modeling software throws at it really is better than one that doesn't, because it will almost certainly succeed in the approximately 80% of future timelines which, if Bernstein is right, do not experience "the end of life as we know it", whereas the one with an 80% success rate in modeling has a 1 in 5 chance of failing even if life goes on as it always has in the past.
Perhaps all modeling results should be multiplied by a factor of 0.8 to take this into account?

 

[haha]

throws at it really is better than one that doesn't, because it will almost certainly succeed in the approximately 80% of future timelines which, if Bernstein is right, do not experience "the end of life as we know it", whereas the one with an 80% success rate in modeling has a 1 in 5 chance of failing even if life goes on as it always has in the past.

With this you are on the right side of Pascal's Wager. If the modern capitalist world falls apart, what could any individual do anyway? If it endures, the more conservative path will have a better shot of making it. It seems always to be better to reasonably defend against not necessarily the more likely outcome, but the most damaging one that can be defended against or mitigated.

Ha

 

[walkinwood]

With this you are on the right side of Pascal's Wager. If the modern capitalist world falls apart, what could any individual do anyway? If it endures, the more conservative path will have a better shot of making it. It seems always to be better to reasonably defend against not necessarily the more likely outcome, but the most damaging one that can be defended against or mitigated.

Ha

You are right. If every one of 5 components of your retirement plan have 99% success rate, your retirement plan success rate is only 93.6% (99%^5). So, it makes sense to maximize the success rate of any one aspect of your retirement plan (SWR in this case)

However, most of the SWR methodologies assume that you will not make any changes to the plan based on current realities (portfolio value, inflation etc.). I think most people (at least on this board) will review their portfolios regularly and will be aware of macro conditions and WILL make changes to their SWR to increase the survivability beyond what the model suggests.

Speaking for myself, I'll take the chance of having to correct mid-way (either reduce expenses in bad years, stash away the excess in good years or work a bit) over working till I can get to some theoretical 100% success rate. I have not counted on SS in my calculations, so there is some fudge factor built in. I'm also hopeful that Bernicke is on to something real.

 

[GregLee]

As I understand it, Bernstein says any result over 80% is meaningless because he estimates that our way of life itself has only about an 80% chance of survival.

It does seem to be a non sequitur. At most, he's shown that a result not taking into account the chance of our way of life enduring must be discounted to something less than or equal to 80%. That's not meaningless.

 

[bentley]

This is one of those datapoints that is difficult to ascertain, and we all must simply make our best guess, add a few years for insurance, and then move along. Otherwise none of us would ever retire early.
My Dad died of lung cancer at age 57 (my age now). (He was a heavy smoker for the many years.) I never have smoked. (Other than his second hand smoke until he stopped smoking when I was 14) My mom is still alive at 84, although she is in a nursing home with parkinsons.

I was recently diagnosed with Prostate Cancer with a gleason score of 9. (not good) I'm on the downhill side of a radical retropubic prostatectomy, and I currently feel pretty good. Last psa was 0.04, and just took my first hormone injection. (Every 4 months for next two years.) I may live another 10 years, or another 20. I suspect my quality of life will not be good during the last 3 to 5 years no matter how long I live.

That being said, I'm using the 20 year number which equals age 77 for me. I'm still w##king three days a week doing some consulting and making over $100k.

Brand new home is paid for. Both vehicles are paid for and have low mileages so they should last for many more years. $1.5mm in ira's and 401k's. We have some paid for rental property which produces $9300 annually. This is about a wash with what our health insurance costs. Health ins is 50% subsidized by ex megacorp employer. (Great insurance, but expensive)

So, at this point, assuming the market doesn't collapse (again), then I feel relatively secure with where I am. Still plan on working another year so that I'm close enough to 59.5 that a 72t will not be required.

It's very interesting to me that many of you are using 90 plus ages in your calculations. Men rarely live that long, although I wish you the best in reaching that milestone.

So, the 64$ question is this; Will the market crash? <again> and if it does, what will the recovery time be? Okay, time to get the crystal balls out.

b

 

[Khan]

Well, that's a different kind of survivor bias than we usually discuss, but it still skews the sample statistics.

You already know how you're going to feel in 10-20 years. Your longevity risk (the "risk" of living longer than your assets) rises slightly with every year you survive from now on. How would you feel if you survived those 10-20 years and realized that you could make it another 20? Do you want to add longevity risk to your list of worries? Or would you rather put a number on your plans that will minimize the risk of outliving your money, allowing you to do more philanthropy today?

I use age 115 or 120, depending on what limit the software allows. Spouse's Ashkenazim lineage has four centenarian grandparents to verify her longevity risk. She says she's going to nag me to stay with her, which will have the same relative time-dilation effect as living to age 115-120.

One option for you would be to see how Social Security at age 62/66/70 affects your longevity risk. Another would be to consider annuitizing what's not already covered.

I'm not worried about outliving my assets.

I know I'm past my software support date.

 

[Emeritus]

If you really believe you have a life expectancy above the average, due to your fantastic genes, clean living, lack of risk taking and wonderful luck, the numbers say you should buy annuities, or at least take the self only choice on a DB pension, and postpone Social security.

 

[ERD50]

There is no magic in making a portfolio last forever.

The 4%/95% methodology has been back tested and proved to keep the purchasing power of the portfolio intact after 40 years 90+% of the time...

And how low did the WR go? And how much to kick it up to a 98-99% success rate?

-ERD50

There was a thread earlier about just how low the 4%/95% SWR went. The worst case was pretty bad. However, it wasn't "dog food" bad. We have enough slack in our budget.

I don't worry about bumping the probability up from 90+ (I don't remember what Clyatt's exact number) to 98-99%. See the William Bernstein paper that Nords posted.

Others have chimed in with much the same, but I'll respond directly -

I say it isn't 'magic' because it is just a choice. One can take a fixed X.X% WR, or a somewhat higher initial (but variable) WR. Personally, I prefer to stick with a fixed/lower WR, rather than tell DW & myself, "the market was down this year, so we are cutting our budget by 5%, and then tell her the same thing for the next 2, 3 or X years". Or, to run into a short 2 year drop followed by big gains, and have to say "Hmmm, I guess we really didn't need to cut back at all the past two years - sorry." No right/wrong, no magic, just my choice and yours may be different for good reasons. The higher initial WR has it's upside also.

As far as the Bernstein paper, I just don't think it has any relevance to the success rates that FIRECALC presents. It is a different view of the issue, as kyoung1956 points out. While a reported 95% success rate in FIRECALC may not have much absolute accuracy going forward, I have zero doubt that someone with a 100% success WR will be better able to handle bad times than someone who chooses a 90% success WR. How could that not be true? It has nothing to do with the "accuracy" of 95%.

I'm a big fan of Bernstein, I just think this is a misapplication of what he says.

-ERD50

 

[Emeritus]

. Personally, I prefer to stick with a fixed/lower WR, rather than tell DW & myself, "the market was down this year, so we are cutting our budget by 5%, and then tell her the same thing for the next 2, 3 or X years". Or, to run into a short 2 year drop followed by big gains, and have to say "Hmmm, I guess we really didn't need to cut back at all the past two years - sorry." No right/wrong, no magic, just my choice and yours may be different for good reasons. The higher initial WR has it's upside also.
-ERD50

Obviously "you pays your money and you makes your choice" No criticism of any kind of your lifestyle choice is suggested. However due to the uncertainty of life anyone may have to cut the budget for X this year due to the need to consume more of Y. Taxes also change the budget. So IMHO the concept of "stability" in consumption is a phantom.

Can I propose a slightly different approach?

By whatever means you split your consumption into required and variable . You use different investment strategies for the two components.

Required consumption you generate from the most secure inflation protected vehicle possible. Low risk, low return. Puts food on the table and a roof over your heads.

Variable consumption is fueled by high risk high return investments. This does expand and cut back with economic conditions and personal needs.

Required can obviously either be "bare bones" or close to your current lifestyle. Owning your house outright put it down as an investment to support required consumption. For many social security and pensions can fill out the remaining required consumption. Insurance can also play a part both health and long term care.

Beyond that there is absolutely no reason to worry about a stable long term SWR. SWR is not a concept critical to variable consumption. whether the variable is a medical emergency, a child in need of help or a trip to Tahiti its simply a different calculus.

In our personal case we found that it takes about 2/3 of our assets, considered as an income stream to fund our comfortable required lifestyle.
That is funded by secure "bedrock" investments. The rest is in broad based market matching stock funds, which we will liquidate as desired. .

 

[ERD50]

Obviously "you pays your money and you makes your choice"
...

Can I propose a slightly different approach?

By whatever means you split your consumption into required and variable . You use different investment strategies for the two components.

I think that's reasonable, and most people probably do this, with varying degrees of consciousness/formality. And some might do very little of it if most of their income was COLA pensions - but even they need to look at is because their personal COLA may outstrip official COLA.

For that reason, I'd question being able to fund the "required" category from secure inflation protected, low risk, low return vehicles. We may very well need something above inflation returns to fund our personal inflation rate. So now we face draw-down in bear markets.

And using a combination of your methodology and the 95/5% approach, I think it means that if your investments assigned to the "variable" category fell by 10% in one year, you would cut your variable spending by 10% (or maybe just 5% to do some smoothing). That seems more reasonable to me, with the inflation caveat (which throws a pretty major wrench into this). And I think this focuses my concerns with the 95/5% approach - if someone's "variable" component is a small % of their total expenses (a common tendency for for LBYM types), the 95/5% approach could have you sitting at home denying yourself of most/all 'fun' activities for many years - maybe irretrievable years.

I hate doing micro-budgets, so I've taken pretty much a "40,000 foot view" of this. If projections show my SWR rising a bit, I might decide to get a few more years out the cars, delay a remodel project, fix versus replace, or whatever. OTOH, market downturns can be the best time to buy these things. It's a balancing act. Long term, if I see trouble ahead it would mean downsizing and hopefully pulling net equity out my house.

-ERD50

 

[kaudrey]

I use 95. It's not that out of the realm of possibility - my grandmothers died at 94 and 87, and my grandfather died at 86.

I do agree that using 90 or 95 doesn't make that much difference to me. There is a little history of Alzheimer's in my family, so I need LTC by then anyway.