What probability of success did you retire at (or plan to retire at)?

What was/is your probability of success for FIRE

  • 100%

    Votes: 115 68.5%
  • 95-99%

    Votes: 37 22.0%
  • 90-94%

    Votes: 9 5.4%
  • 80%

    Votes: 6 3.6%
  • 70%

    Votes: 0 0.0%
  • below 70%

    Votes: 1 0.6%

  • Total voters
    168
Some additional safety margin is gained because not all of your income will actually be taxed at the marginal tax rate, unless you are in the lowest marginal tax bracket.
100% correct. I realized that I left that part out and revised my post accordingly.
 
For those who chose 100%.... are we talking about the bare minimum to reach 100% (based on the number of years, yearly withdrawal and size of the nest egg) or far exceeding it?

Thanks

For me, it wasn't watching "the number" to determine retirement date. Rather, it was the BS bucket that filled up. So, I worked too long some might say but it turned out that my w/d rate was less than 1%.
 
I have a hair over $100,000 more in my check book than i had 5 years ago when i retired. i have everything i want except for my new vette that is paid for already.
 
For those who chose 100%.... are we talking about the bare minimum to reach 100% (based on the number of years, yearly withdrawal and size of the nest egg) or far exceeding it?

Thanks

We are roughly 30-50% over the number we'd need for 100%.
 
I chose a withdrawal method that couldn’t run out of money by definition - thus 100% success guaranteed. This is taking a fixed percent of the portfolio every year. No inflation adjusting on withdrawals. If the portfolio goes down, so does the income. So you live with varying annual income. I figured out that I could live with that.

FIREcalc models this method as %remaining portfolio. You can choose different AAs and even specific asset types look at min, max and average real ending portfolios based on historical data. I had to dig a bit deeper to model drawdown of the portfolio (and thus income) during worst case sequences.
 
Last edited:
FireCalc tells me that historically we would've had a 100% success rate if we had spent 1.5X what our 2019 expenses were - not taking SS or my wife's future pension (which she'll be eligible to receive in 2022) into consideration - at a 70/30 allocation and assuming we lived to 100.

I didn't plan it that way. It just happened.

I left the workforce in December of 2017. My wife still works.
 
This is how I handle expected taxes:

1. Start with actual spending on everything but income taxes.
2. Assume that all money needed to support that spending is taxable ordinary income.
3. Gross up for taxes at the marginal rate.
4. Use the grossed up number in FIRECalc

So, for example, suppose I actually spend $81k per year based on having tracked my spending for many years. That puts me at the bottom of the 22% federal bracket and in the 5.5% state bracket. To gross up, so that after I've paid taxes I have $81k left to spend, I use the following formula: Gross draw = actual spending/(1- marginal rate). In this case $81k/(1 -.275) = $111.72k gross draw. That's what I put in for spending on the first tab of FIRECalc.

Will my actual taxes be that high? No, because I didn't take into account the standard deduction, nor the fact that some of my income will be non-taxable social security and some will be withdrawals from Roth and after tax accounts, nor that some may be long term capital gains. It also assumes taxation of income at the marginal rate instead of the effective rate. But doing it this way ensures that I am being conservative in estimating my chances of success. It is also far simpler and easier to do.

Tax rates may go up in the future, but there is really no way to predict when and by how much, so the best I can do now is just use the current ones and err on the conservative side.

Perhaps one big difference in our situations is that I have no heirs, so I care not one whit about my ending balance, as long as it is above zero. FIRECalc says it will average 8 figures. That's good enough for me.

Your approach is certainly conservative for now. But the puzzle of how to handle taxes extends to the future - if you really end up with 8 figures in your portfolio, your taxes will have less to do with your spending and more to do with the dividends/realized capital gains in taxable accounts and IRA RMDs. Of course if you have 8 figures, you are not worried about having to live in your car, so this is a First World problem.

I think I will test the Firecalc by first reducing my IRA values by my assumed future tax rate.
 
Your approach is certainly conservative for now. But the puzzle of how to handle taxes extends to the future - if you really end up with 8 figures in your portfolio, your taxes will have less to do with your spending and more to do with the dividends/realized capital gains in taxable accounts and IRA RMDs. Of course if you have 8 figures, you are not worried about having to live in your car, so this is a First World problem.

I think I will test the Firecalc by first reducing my IRA values by my assumed future tax rate.
For the next 12 years (until the young wife is 72), we plan to do Roth conversions to the top of the 22% bracket, which will help deal with the RMD issue. The amount we currently have in taxable, while sufficient for a few years of expected draws, is a relatively small portion of our overall portfolio (<8%). And it is entirely cash, so I'm not worried about dividend income or capital gains until after RMDs start, which is when taxable will build up rapidly and I'll need to put it in something other than cash. But, as you note, that is a first world problem.
 
Last edited:
I think you just nailed the reason Dory36 declined to include taxes in FIRECalc. :)

No kidding, and my buggy spreadsheet only covers an approximation of my relatively simple tax situation. Some issues that come up - you should deplete taxable accounts first, but to spend $1 you have to sell more than $1 based on the unrealized capital gain. Then, no account balance can go below zero, so once your taxable accounts are gone, you have to tap IRAs, but essentially all of that will be taxed, so you need to withdraw even more to get $1 free cash flow. These extra $ put you in higher tax brackets, which may mean more withdrawal needed, so you have to iterate. You also have to check AMT, and some trigger points like the NIIT are not indexed for inflation. I didn't even bother with IIRMA trigger points, ACA subsidy points, etc, but those can be important too.

Even if you could program the tax code (let alone know the future tax code), who would fill all that out?

Note that for primary intent of the program which I take to be determining if it's safe to retire, portfolios that get very large are safe, so getting taxes just right on them is not the focus. I just wanted to point out to folks not to spend all that future money quite yet.
 
At age 50 in 2003 I created a very simple Excel spreadsheet with inputs for inflation rate and an average rate of return. I started with what we had at that point and made assumptions about future returns and contributions, and then took 4% of the balance at age 65 and looked at the current-dollar value. It was an amount we could live on given our spending patterns. I left out SS, not sure if it would be around.

When I got fed up with the toxic politics at work at age 61, I looked at it again and added in SS, which now looked more certain. I was out the door a week later.

I've kept the withdrawal rate at 3.5% and my advisor (don't flame me for that) has run fancy Monte Carlo simulations that do include taxes and "lumpy" expenses such as a new car every 10 years and they say I have a 98% chance of not outliving my savings without LTC expenses, 95% with LTC.

Right now my net worth has appreciated at an average annual rate of 4% since retirement (AFTER withdrawals) so I think it's sustainable. That 4% is on the high side- usually it hovers around 3%.
 
Uncertainty around taxes was what held us back. All of the calculators significantly over estimated them and don’t get me started on trying to project into the future! I finally came to the conclusion that if taxes started getting significantly higher than I was calculating, it would be because I was on a ‘good’ trajectory and decided to stop worrying about them.
 
Here is how I handled income taxes.


I split up my tax estimate into two parts. The first is a basic tax amount, an amount based on the more predictable monthly bond fund dividends which provide me the regular cash inflows to pay the bills. The second is an excess tax amount , an amount based on the less predictable and more volatile cap gain distributions. I don't really worry about the taxes due on those because I can always take those distributions in cash, especially if they are large, and use them to pay the taxes.


Those large distributions in recent years, however, kept throwing me over the ACA subsidy cliff, so the added taxes due included the subsidy I had to return. With the value of the subsidy rising, I changed my portfolio to greatly reduce these income spikes and the excess taxes they generated.
 
I am planning to retire Mar 2021 @ 55. Right now, we are at 95% if I quit today. If I wait until Mar (which is what I requested), we are at 99%. I have a very large chunk of RSU's that vest when I leave in Mar vs. now, so that is the difference. Plus 3 more months of income.
 
I never used FireCal if that what you are using to get your probability % of success. I went by the 4% rule. Once I reached 25x my annual expenses, I began cutting my hours at work and within about six months, went part time down to about 20. Spent 2 years part time until my net worth reached 33x my annual expenses. Then I retired and currently I am at about 50x expenses.


I also use a great 401k calculator on Moneychimp.com. It's very concise.

401k Calculator


I would guess that I was at 90% when I first reached FI and now am at 99%. I leave 1% for the unknown which is always probable.
 
This is how I handle expected taxes:

1. Start with actual spending on everything but income taxes.
2. Assume that all money needed to support that spending is taxable ordinary income.
3. Gross up for taxes at the marginal rate.
4. Use the grossed up number in FIRECalc

So, for example, suppose I actually spend $81k per year based on having tracked my spending for many years. That puts me at the bottom of the 22% federal bracket and in the 5.5% state bracket. To gross up, so that after I've paid taxes I have $81k left to spend, I use the following formula: Gross draw = actual spending/(1- marginal rate). In this case $81k/(1 -.275) = $111.72k gross draw. That's what I put in for spending on the first tab of FIRECalc.

Will my actual taxes be that high? No, because I didn't take into account the standard deduction, nor the fact that some of my income will be non-taxable social security and some will be withdrawals from Roth and after tax accounts, nor that some may be long term capital gains. It also assumes taxation of income at the marginal rate instead of the effective rate. But doing it this way ensures that I am being conservative in estimating my chances of success. It is also far simpler and easier to do.

Tax rates may go up in the future, but there is really no way to predict when and by how much, so the best I can do now is just use the current ones and err on the conservative side.

Perhaps one big difference in our situations is that I have no heirs, so I care not one whit about my ending balance, as long as it is above zero. FIRECalc says it will average 8 figures. That's good enough for me.
Dad? Is that you?
 
FIRECALC indicated 100% before I retire.

Actually I am spending less than 3% NW without counting SS, so I'll say 100%+
 
I never used FireCal if that what you are using to get your probability % of success. I went by the 4% rule. Once I reached 25x my annual expenses, I began cutting my hours at work and within about six months, went part time down to about 20. Spent 2 years part time until my net worth reached 33x my annual expenses. Then I retired and currently I am at about 50x expenses.


I also use a great 401k calculator on Moneychimp.com. It's very concise.

401k Calculator


I would guess that I was at 90% when I first reached FI and now am at 99%. I leave 1% for the unknown which is always probable.

Effectively using an historical sequencing concept, Firecalc is calculating in its default format using the 4% WR "rule".
 
Similar for me. 100% in all calculators was a requirement, but also used a spend rate well in excess of what I actually expected. The upshot is that I retired 3-5 years later than I probably could have, but at a much reduced level of worry. This may or may not be a worthwhile trade for many, but I had just recently been through a halving of my assets (and subsequent recovery) during the Great Recession and was ill inclined to take chances.

Same here...100%. My passive income is 3x my living expenses. On top of that, I have a pension I can collect at 60 and then SS is the "fun money" at 62 (or maybe later). I don't suspect I will touch my portfolio...that will just continue to grow.
 
My plan is that I need to have at least (and ideally more) than the amount that FIREcalc says would have given me 100% success rate in the past. I've built into my plan a longer lifespan than I expect (102, 10 years older than my grandfather). I'm omitting the equity in any real estate, etc. I own. I'm also completely omitting any possible inheritance from my plan, as well as the possibility that I might want to get a PT job (retiring at 55, if I make a few thousand a year from a PT job, I can put that in my Roth). I think it's likely that if I'm careful and manage my finances pretty much as I have all my life, my nest egg will continue to grow throughout my retirement even as I draw an income from it.
 
I am 10 years out and coasting towards fire. I have two buckets, one for before my pension and one for after. The one for before I am requiring 100%, the one after is 80%. It's a little easier to feel safe when you have a pension.
 
I used a lot of different tools but hired a fee only Financial Planner and he was the one who convinced us we would be ok.
 
Mine was 100% but I used a more conservative estimate of spending than I am actually living. Despite spending more each of the five years since retiring my portfolio is about a third larger than it was when I retired.
 
I have no idea what success rate I would have been before retirement. I retired some time before I ever knew of this site. I save, in retirement, anywhere from $2,000 to $10,000 per month depending on what bills come in that month including interest on what I have saved. I would guess that I am probably 100%.
 
Back
Top Bottom