Looking Back ... Were the projections right?

[Steelart99]

I'd thrown in the residence more for an indicator of what I still owed to get this darn thing paid off. We love where we live and plan to stay here. I have a seperate plot of land worth about $100K that I'll sell at some point (I hope) to pay off the remaining mortgage when the remainder is about $100K. That I do consider as an asset. We'd love to not have a mortgage upon retiring (or shortly thereafter).

My post-retirement budget is coming in lower than our current expenditures, but not by a huge amount. I'd added in $5K per year as an emergency fund and about $800/month medical (total for DW and I). We've been fairly frugal other than all the gifts we give to our extended family for birthdays, X-mas, etc. One thing I just found out ... and I feel stupid ... is that SS and Medicare is not charged on withdraws from retirement accounts (This is right ... right?).

My side business was fairly established at one time, but has languished a bit over the last several years. Easily "revived" I believe. Something I enjoyed doing and made a fair income. I should get it moving again before I pull the trigger along with living on my reduced budget. All good points!!! I like the point you made ... what would I do if I had to further reduce my budget by 20-40%. Hmmmmmm .... gonna think on that one.

 

[Katsmeow]

When you say Firecalc is marginal different people consider it differently. There are some people here who would think 99% was marginal. I personally tend to shoot for 95% and I was going over it with DH a couple of months ago and he felt 80% was good enough.

I have found that if I run a place it come out at say 95% I can get it to 100% with very minor changes in the plan. Even getting from 90% to 100% doesn't take a whole lot.

On the other hand if by marginal you mean your plan is at 45% then getting to even 90% is going to be challenging.

So how marginal is marginal in your plan?

 

[FinanceGeek]

So how marginal is marginal in your plan?

I read an interesting perspective once on ER and "success rates". The claim was that looking at history - there is about a 20% chance of a dramatic change in economic & political conditions within a given 30-40 year period. The takeaway being - success rates computed by a calculator based on our present societal conditions have no way of taking that into account and it may be naive to strive for greater than ~80% success rates.

 

[Steelart99]

Katsmeow, I tended to be marginal when I set goals for 100%. Ususally I ended up in the 80s to llow 90s ... yeah, I tend to be a bit conservative and my DW is a cronic worrier. When I first started using FC, I had 100% success in many of the scenarios I built, but then I started using the Manual Entry of Spending Changes to accomodate the out year purchases (cars) and to cover 2-3 years of accelerated mortgage payments (3X principal) along with adding in about $5K per year for emergency expenses.

I obviously have "issues" in the early years of my "planned" ER date before I can hit any of our pensions or SS. I'd be relying on early disbursement of my 401K and my hobby business which I know will bring in some income. DW may also keep working for a few years ... she doesn't think it's "fair" that she can retire long before the time that her parents did. I started this thread to get some insight into how the various programs projected success, knowing all the caveats. I'm getting some great perspectives!!!

 

[MichaelB]

My post-retirement budget is coming in lower than our current expenditures, but not by a huge amount. I'd added in $5K per year as an emergency fund and about $800/month medical (total for DW and I). We've been fairly frugal other than all the gifts we give to our extended family for birthdays, X-mas, etc. One thing I just found out ... and I feel stupid ... is that SS and Medicare is not charged on withdraws from retirement accounts (This is right ... right?).

Correct, payroll taxes were collected on your income when it was earned, even as you allocated it to tax deferred accounts, so your only remaining liability is income tax, both state and federal.

$800 per month for medical looks a bit light. The national average is about twice that amount just for premiums.

 

[Steelart99]

Correct, payroll taxes were collected on your income when it was earned, even as you allocated it to tax deferred accounts, so your only remaining liability is income tax, both state and federal.

$800 per month for medical looks a bit light. The national average is about twice that amount just for premiums.

Interesting ... EhealthInsurance gave me a quote of $516/mo with $1K deductible for myself and family. My company has a retiree health insurance plan, but they won't give me any sort of quote until I'm actually within 90 days of retirement. How helpful is THAT for planning to retire? Here, I thought I was being a bit conservative by putting in $800/mo. I'll have to research this some more.

 

[MichaelB]

Interesting ... EhealthInsurance gave me a quote of $516/mo with $1K deductible for myself and family. My company has a retiree health insurance plan, but they won't give me any sort of quote until I'm actually within 90 days of retirement. How helpful is THAT for planning to retire? Here, I thought I was being a bit conservative by putting in $800/mo. I'll have to research this some more.

Was the eHealth quote underwritten? Usually what they give you is not so much a quote, as in firm offer, but a price that might be if you meet all the underwriting requirements. To get this you must apply.

You can get an idea of healthcare costs using one of the new calculators, here Subsidy Calculator | The Henry J. Kaiser Family Foundation and here National Health Care Calculator . The national average health care premium for large group coverage in 2012 was $5.6K individual, $15.7K family. This number does not include any type of cost sharing, such as co-pays or deductibles.

If I were building a budget or projecting future expenses I would use the national averages, or perhaps a bit more, unless I had a reliable alternative already approved or available.

 

[Steelart99]

Was the eHealth quote underwritten? Usually what they give you is not so much a quote, as in firm offer, but a price that might be if you meet all the underwriting requirements. To get this you must apply.

You can get an idea of healthcare costs using one of the new calculators, here Subsidy Calculator | The Henry J. Kaiser Family Foundation and here National Health Care Calculator . The national average health care premium for large group coverage in 2012 was $5.6K individual, $15.7K family. This number does not include any type of cost sharing, such as co-pays or deductibles.

If I were building a budget or projecting future expenses I would use the national averages, or perhaps a bit more, unless I had a reliable alternative already approved or available.

Excellent links, Thanks. I need to make sure that I have this covered appropriately in my budget. Crazy what the cost of Health Insurance is these days.

 

[Major Tom]

The takeaway being - success rates computed by a calculator based on our present societal conditions have no way of taking that into account and it may be naive to strive for greater than ~80% success rates.

My view is a little different. Given that growth rates may possibly be more moderate over the next few decades than in the past, I tend to think that it may be naive to strive for anything less than a 100% success rate in Firecalc. This is a reflection of my own financially conservative nature.

Of course, Firecalc doesn't predict the future, so we are all free to interpret it's results in the manner that best suits our own individual outlooks.

 

[ERD50]

I read an interesting perspective once on ER and "success rates". The claim was that looking at history - there is about a 20% chance of a dramatic change in economic & political conditions within a given 30-40 year period. The takeaway being - success rates computed by a calculator based on our present societal conditions have no way of taking that into account and it may be naive to strive for greater than ~80% success rates.

My view is a little different. Given that growth rates may possibly be more moderate over the next few decades than in the past, I tend to think that it may be naive to strive for anything less than a 100% success rate in Firecalc. This is a reflection of my own financially conservative nature.

Of course, Firecalc doesn't predict the future, so we are all free to interpret it's results in the manner that best suits our own individual outlooks.

I'm with Major Tom, and I've commented on this before. Yes, you can't think in terms of 85%, 90%, 95%, or 100% as absolute values going forward (and how would we ever know, even after the fact? Like the weather forecast for a 50% chance if rain - 'right' no matter what?). But how can it not be true that a plan that had 100% historical success would not be safer in the unknown of the future than one with 85% historical success?

I am also not comfortable with a plan that I know has failed in past scenarios. Again, that is no guarantee of future success, but given that the worst future scenarios may be worse than the worst past scenarios, why not build in some safety factor?

How is that 'naive'? It seems prudent to me. Others may be fine with the higher risk of a historically 80% safe plan, but it is a higher risk. I don't see how that can be denied.

-ERD50

 

[Major Tom]

Agreed ERD50 - and another reason I'd like to build some extra safety into my plan over what, say, a 90% success rate would give me is that I'd like, as much as possible, to avoid the outcomes that take my portfolio down to a level that might cause me to lose sleep.

I currently have an ~2.5% WR which gives me a much better than 100% success rate over 40 years (is there such a thing as better than 100%? Well, you know what I mean) - and that is not even including my SS, which gives me more of a safety factor.

I'm thinking that if you use a WR that gives you an 85% success rate, for example, while your portfolio may well always support your desired WR, you are running a greater risk of it sinking to some rather scary low levels while continuing to support you. I want to avoid those scenarios if at all possible.

 

[FinanceGeek]

Agreed ERD50 - and another reason I'd like to build some extra safety into my plan over what, say, a 90% success rate would give me is that I'd like, as much as possible, to avoid the outcomes that take my portfolio down to a level that might cause me to lose sleep.

I currently have an ~2.5% WR which gives me a much better than 100% success rate over 40 years (is there such a thing as better than 100%? Well, you know what I mean) - and that is not even including my SS, which gives me more of a safety factor.

I'm thinking that if you use a WR that gives you an 85% success rate, for example, while your portfolio may well always support your desired WR, you are running a greater risk of it sinking to some rather scary low levels while continuing to support you. I want to avoid those scenarios if at all possible.

My intent in posting that comment wasn't to say you shouldn't plan conservatively. The idea I mentioned came from Bill Bernstein in a series of articles that he wrote ~10 years ago (before FireCalc existed) where he stated that the retirement calculators and assumptions which were available at that time were too optimistic.

So I think we're on the same page - optimize to the extent you can with the tools you have. Mr. Bernstein just reminds us that at some point your true of risk failure is dominated by factors that are not estimable by simulations like FireCalc.

The Retirement Calculator from Hell, Part III

 

[sengsational]

I read an interesting perspective once on ER and "success rates". The claim was that looking at history - there is about a 20% chance of a dramatic change in economic & political conditions within a given 30-40 year period. The takeaway being - success rates computed by a calculator based on our present societal conditions have no way of taking that into account and it may be naive to strive for greater than ~80% success rates.

So I think we're on the same page - optimize to the extent you can with the tools you have. Mr. Bernstein just reminds us that at some point your true of risk failure is dominated by factors that are not estimable by simulations like FireCalc.

Because most of us here are extremely bought-in to the current system, we don't often stare-down the beast that would be "dramatic change". It's just too uncomfortable. So we spend our time putting finer and finer points on FireCalc inputs when the whole graph could be shifted down by 40% with the stroke of a pen or the sudden loss of faith in creditworthiness. I wish it wasn't a house of cards, but I do think it is. The good news is that societies tend to be pretty resourceful, so even in the face of dramatic change, we can get through it. But I'd be real surprised if my real personal FireCalc line follows any of the badjillion lines it dutifully plots for me.

 

[Major Tom]

My intent in posting that comment wasn't to say you shouldn't plan conservatively. The idea I mentioned came from Bill Bernstein in a series of articles that he wrote ~10 years ago (before FireCalc existed) where he stated that the retirement calculators and assumptions which were available at that time were too optimistic.

Understood now. As is my wont, I was interpreting your comment through the narrow filter of my own experience, as opposed to attempting to understand the context in which you made it.

when the whole graph could be shifted down by 40% with the stroke of a pen or the sudden loss of faith in creditworthiness.

...or as Bernstein puts it, we could be visited by "the ghosts of Hitler, Lenin, and Attila the Hun"

 

[ERD50]

Because most of us here are extremely bought-in to the current system, we don't often stare-down the beast that would be "dramatic change". It's just too uncomfortable. So we spend our time putting finer and finer points on FireCalc inputs when the whole graph could be shifted down by 40% with the stroke of a pen or the sudden loss of faith in creditworthiness. ...

Sure, 'dramatic change' could happen (heck, I've already seen my portfolio down ~ 40%). But wouldn't it still be better to have about a third more after that drop (roughly - the portfolio to support 3% versus 4% WR)?

This is getting to sound a little like saying that I decided not to carry a spare tire, because I could have two flats at the same time, and then that single spare won't help me. But the single spare will help in most cases.

There is no certainty, and we can't plan for every possible future extreme. I've decided to plan for at least what we have seen, and not ignore history.

-ERD50

 

[youbet]

7+ years into ER, I find I'm well within the range of possible outcomes given by FireCalc. My graph is similar to REWahoo's in post #2 above except my dip comes earlier since I retired a year later (a year closer to the 2008 - 09 swoon).

Admittedly, having the Great Recession come pounding on my door about two years in gave me a few moments of concern. Not only for the condition of my portfolio but for the well being of my extended family and the possibility of needing to help them out if layoffs or other issues became part of their life.

It's worked out OK. The portfolio has fully recovered in real terms. And there were no family financial disasters crying for help.

 

[explanade]

Has there been more than a couple of people, if that, who've expressed regret here for retiring when they did?

Or said that they realized they weren't quite ready?

This is a self-selecting group so what are the chances that people will say ER wasn't the right thing to do, at least at the time they FIRE'd?

 

[sengsational]

Sure, 'dramatic change' could happen (heck, I've already seen my portfolio down ~ 40%). But wouldn't it still be better to have about a third more after that drop (roughly - the portfolio to support 3% versus 4% WR)?

This is getting to sound a little like saying that I decided not to carry a spare tire, because I could have two flats at the same time, and then that single spare won't help me. But the single spare will help in most cases.

There is no certainty, and we can't plan for every possible future extreme. I've decided to plan for at least what we have seen, and not ignore history.

-ERD50

I agree that almost certainly those of us that prepare will be better off in the case of dramatic change, if that should happen. The difference between ant and grasshopper will just be a bit smaller. But my tire analogy would be that we are twiddling inputs to predict that we'll get 9,999 miles out of these tires, but changing assumptions, it could be only 9,950 miles. But we've got 30 years of potholes to dodge, and any one road hazard hit squarely could knock 2,000 miles of life off the tires.

 

[ERD50]

I agree that almost certainly those of us that prepare will be better off in the case of dramatic change, if that should happen. The difference between ant and grasshopper will just be a bit smaller. But my tire analogy would be that we are twiddling inputs to predict that we'll get 9,999 miles out of these tires, but changing assumptions, it could be only 9,950 miles. But we've got 30 years of potholes to dodge, and any one road hazard hit squarely could knock 2,000 miles of life off the tires.

OK, but in your analogy you are grossly underestimating the differences I'm talking about. That minimizes the real effects.

9999/9950 is not even one-half of a % more. I'm talking in the range of 4% WR versus 3.2% WR. The 3.2% WR requires a portfolio that is 25% larger.

A 3.2% WR has been historically 100% safe for over 46 years. A 4% WR fails 5% of the time over 30 years. So to scale that to tire terms, say you were made the following offers:

A) A set of tires for $500 that had a history of blowing out 5% of the time within 30,000 miles.

B) A set of tires for $625 that had a history of zero blow outs and the tests ran to over 46,000 miles.

That sounds like cheap insurance to me, considering how dangerous, or just plain inconvenient a blow out can be.

-ERD50

 

[larrytbm]

I created my own spreadsheet and retired at age 55, that was 10 years ago. I still use that spreadsheet to help my decisions in retirement. I don't spend a lot of time on financial management, I am a retired engineer and spend much of my time doing volunteer work. My point is that I don't live and breathe financial tools or topics. But my experience with FIRECalc has NOT made me a big believer. My approach was to use my actual data starting 10 years ago, including pension and SS info and enter it into FIRECalc. According to the results, the last 10 years of stock market performance would have to rank as one of the top 6 such periods in history for my assets to achieve the growth that has actually happened. I don't think so and I've not seen any headlines proclaiming 'unprecedented 10 year stock market performance', so I have to doubt the significance of FIRECalc. And it doesn't seem that anyone else has tried to benchmark FIRECalc with real world data.

 

[ERD50]

...My point is that I don't live and breathe financial tools or topics. But my experience with FIRECalc has NOT made me a big believer. My approach was to use my actual data starting 10 years ago, including pension and SS info and enter it into FIRECalc. According to the results, the last 10 years of stock market performance would have to rank as one of the top 6 such periods in history for my assets to achieve the growth that has actually happened. I don't think so and I've not seen any headlines proclaiming 'unprecedented 10 year stock market performance', so I have to doubt the significance of FIRECalc. And it doesn't seem that anyone else has tried to benchmark FIRECalc with real world data.

FIRECalc uses 'real world data' - I'm not sure what you are looking for?

And actually, the past ten years has been fantastic for the stock market:

SPY Historical Prices | SPDR S&P 500 Stock - Yahoo! Finance

That show that with divs, SPY has about doubled in ten years. That's a good thing for a retiree, no?

163.75 / 83.04 = 1.97x

Anyhow, I'm not sure what that has to do with FIRECalc - if you look there are a lot of lines, I'm sure some look very much like your path. The failures are the worst case scenarios, not the average, not the typical, not the median.

In fact, if you set it for ten years and a $40K/$1M spend/portfolio, here's where history has taken us:

The lowest and highest portfolio balance throughout your retirement was $373,216 to $2,775,650, with an average of $1,272,708. (Note: values are in terms of the dollars as of the beginning of the retirement period for each cycle.)

And the absolute values would be higher, due to inflation. Are you saying your results (adjusted for your spend/income) are radically different?

-ERD50

 

[larrytbm]

And actually, the past ten years has been fantastic for the stock market:


Anyhow, I'm not sure what that has to do with FIRECalc - if you look there are a lot of lines, I'm sure some look very much like your path. The failures are the worst case scenarios, not the average, not the typical, not the median.



And the absolute values would be higher, due to inflation. Are you saying your results (adjusted for your spend/income) are radically different?

-ERD50

If the past 10 year performance of the stock market is one of the top 6 ten year periods out of 131, then everything is great. And if that is the case, then we aren't likely to see our portfolios perform as well in the next few decades as we have just experienced. Or, there is an unexplained 'new norm' that FIRECalc using historical data will not predict.

 

[ERD50]

If the past 10 year performance of the stock market is one of the top 6 ten year periods out of 131, then everything is great.

I'm curious where you are coming up with this 6/131 number. I'm not saying it's wrong, I just don't know how this fits in to the discussion.

And if that is the case, then we aren't likely to see our portfolios perform as well in the next few decades as we have just experienced.

Well, generally the market runs in cycles. I lean towards thinking the next ten years will not perform as well, but anything can happen.

Or, there is an unexplained 'new norm' that FIRECalc using historical data will not predict.

You lost me. FIRECalc does not predict anything.


-ERD50

 

[Fred123]

FIRECalc uses 'real world data' - I'm not sure what you are looking for?

And actually, the past ten years has been fantastic for the stock market:

SPY Historical Prices | SPDR S&P 500 Stock - Yahoo! Finance

That show that with divs, SPY has about doubled in ten years. That's a good thing for a retiree, no?

163.75 / 83.04 = 1.97x

Anyhow, I'm not sure what that has to do with FIRECalc - if you look there are a lot of lines, I'm sure some look very much like your path. The failures are the worst case scenarios, not the average, not the typical, not the median.

-ERD50

I believe the last 10 years have actually been less than average for the SP500 over the period that FIRECALC uses (starting in 1871). Here's a calculator you can play with that shows this. If you plug in the last ten years (or 1/2003 - 12/2012), you get an annualized return of about 7% and an inflation adjusted annualized return of about 4.5%. For the entire period, the values are 8.9% and 6.7%

 

[larrytbm]

I'm curious where you are coming up with this 6/131 number. I'm not saying it's wrong, I just don't know how this fits in to the discussion.-ERD50

When I take my actual assets from 10 yrs ago, plus expenses, pension and SS and run FIRECalc, only 5 of 131 cycles have an end point higher than my current assets. That would make the past 10 years one of the best on record. Does that make sense? Is it really one of the top 6?

Well, generally the market runs in cycles. I lean towards thinking the next ten years will not perform as well, but anything can happen.
-ERD50

It just seems to me that where the past 10 years ranks out of all 131 cycles should be very important. And since FIRECalc is using historical data, it seems that an output showing ranking should be ez. If the past 10 year cycle came in with a ranking of say 60/131 instead of 6/131, then something does not compute.

I also think that recent data 'colors' or shifts our judgement regarding risk, spending, 'should I retire early', etc. If we believe that the past 10 years was an average cycle (#60), or we really don't know, while it was really #6, then based on FIRECalc we will probably make worse decisions.