What's your FIREcalc success rate?

[cute fuzzy bunny]

No. I mean if you ask firecalc to determine your success rate for a mix of investments, options and an annual withdrawal amount and you get 100%, theres an option to have it find you a higher annual withdrawal amount that is still survivable.

This option, setting the % to 100.

"FIRECalc will search for settings that will get a success rate of as close to % as possible (usually within 1%) by changing...
how much you spend each year, or"

[garrynky]

My success rate is 100%. I'm out in 97 more working days.

[me_no_boss]

Over 90-100%.
Disclaimer - I am an optimist and calculated risk taker.
I included investment properties sale in future and SS at 62.

[USK Coastie]

Once I got a handle on advanced FIRECALC and ran the numbers I ended up with a 97%- 98% success rate. With that data in hand, promptly started plans to FIRE.
ER'd a week or so now, no tan line from wearing a watch, and starting to have trouble remembering what day of the week it is.

To second may others on this forum "If I'd known how good it was going to be, I'd have left a LOT sooner."

[audreyh1]

When I first started, the success rate was perhaps around 90-95% for a 50 year period. But I didn't worry about it, because I never intended to take yearly inflation adjustments (increases). Instead I decided to stick with a straight 4% withdrawal which has a 100% success rate since the portfolio can theoretically never be depleted if you limit your withdrawal %.

Since then, my portfolio has grown enough that withdrawal needs have lowered to < 3.5%, so I worry about survival even less!

Audrey

[cute fuzzy bunny]

Quote:

Originally Posted by audreyh1 When I first started, the success rate was perhaps around 90-95% for a 50 year period.

Be awfully careful about those super long runs. Firecalc runs dont "wrap around" or substitute extra data for runs that dont have 50 years of data, so anything that ends "short" and didnt fail is considered to be a success.

So for your 50 year run, all of the tests for years 1956/1957-today are shortened by 1-49 years and are therefore successful, whether they actually might have been or not.

So the last "full" test run that succeeded for you is the one starting in 1955/1956.

Its tough to work this problem out for long term retirees. You could do multiple 10 or 20 year concatenated runs with average terminal portfolio sizes, but that screws up year to year correlative returns and thats a bad thing. You're basically mixing the 'true results' approach firecalc takes with monte carlo, in chunks.

Or you could just do 20 year runs and if you succeeded and had an average terminal portfolio growing or holding its ground from your start point, inflation included...decide that making it through all approaches to the great depression and the 64-75 side-slide are satisfactory to assure that you've got enough momentum to make it through anything.

[Nords]

Quote:

Originally Posted by cute fuzzy bunny Be awfully careful about those super long runs. Firecalc runs dont "wrap around" or substitute extra data for runs that dont have 50 years of data, so anything that ends "short" and didnt fail is considered to be a success.

I thought this was fixed in FIRECalc's latest version?

Paging Dory!

[cute fuzzy bunny]

Quote:

Originally Posted by Nords Paging Dory!

I believe he's currently at sea. Hopefully not a three hour tour...

Hmm...how would it be fixed?

By claiming that incomplete runs dont count? Pretty much the same problem...you're collection of 50 year runs would go from 1871-1956...not much modern data there.

By wrapping it around? Not sure that going from steady inflation periods of today to deflationary periods following the civil war works for me as good data.

Filling in funny extrapolations? Fooey.

[cute fuzzy bunny]

And funny bug...go to firecalc support and click on "threads from the beginning". You get no threads other than the one 'sticky'.

So we're currently unable to look at firecalc threads older than one year, the next highest granular option.

[HobbyDave]

I'd suggest in the advanced firecalc, you could probably setup options for how to treat those long time periods:

1. Ignore any time periods where there aren't enough years of data.

2. Wrap around the years

3. Start the wrap around at a certain year (IE, after 2006, make it go to 1972 and continue from there).

4. Inflation growth after that point. So after 2006, you'd just increase by your inflation rate each year.

[Webzter]

Quote:

Originally Posted by cute fuzzy bunny I believe he's currently at sea. Hopefully not a three hour tour... Hmm...how would it be fixed?

If you ask for a 50 year, it stops 50 years ago for it's last starting year.

[Nords]

Quote:

Originally Posted by cute fuzzy bunny I believe he's currently at sea. Hopefully not a three hour tour...

That scene is shot from the Ala Wai boat harbor, and every time I see that it takes a couple days to get that $%&^ing jingle out of my head. Which will no doubt follow me long after I'm in geriatric care.

Quote:

Originally Posted by cute fuzzy bunny Hmm...how would it be fixed?

I thought it was fixed by throwing out the incomplete runs, which pretty much eliminates the false optimism of partial runs by replacing it with the false optimism of avoiding all the bad stuff which happened in the last century. It's not as if we could glue together a bunch of 25-year runs, double their failure rate, and proclaim victory.

Just about every other solution for long-term runs defaults to Monte Carlo. Which perhaps isn't such a horrible idea for runs over 40 years, given that we don't have better historical data to use. I know MC has an entire book of its own flaws but... what's better?

Oh, wait, I remember-- living off dividends without touching the principal. Of course by the time most "E"Rs get to a portfolio big enough to support that goal, no one would have to worry about 40-year projections.

Quote:

Originally Posted by cute fuzzy bunny So we're currently unable to look at firecalc threads older than one year, the next highest granular option.

IIRC those have been archived off the public boards. Paging Dory again!

[cute fuzzy bunny]

Ah so you dont get false positives, but you dont get any 'modern' runs either.

Perhaps a better result than telling you 30-40 incomplete runs succeeded when they might not have, but still not a very good tool for determining successful investing in the modern era.

But I'll go back to something thats been said a hundred times, perhaps 85 of them by me...the whole firecalc thing is nice but basically if you made it through the depression and stagflation periods I dont think anything will nick you up any worse than that. Doing runs of 15-20 years seems to thoroughly cover those two bad scenes.

Oh...and...

http://www.webweaverdesign.ca/portfo...sland2ndtv.wav

[hellbender]

100%.

Since my pension is inflation adjusted and I can live on a percentage of it, the model says that I need never make a withdrawal from my portfolio. In fact, if I tell the model that I have only saved $1 in my entire life and never invest a cent in equities, it tells me that my retirement is sound for 40 years and that I will die a millionaire. DB pensions. . . . gotta love 'em.

[lazygood4nothinbum]

i thought i read on an earlier thread that the greatest amount of years for an accurate firecalc run is 38 years. is that not correct?

[Lsbcal]

I voted the 100% success rate -- nothing new there. However, when looking at the FIRECalc results I generally key on the maximum drawdown. Recently when running the numbers the drawdown amounted to a worst case of 36% of the portfolio start value. This turned out to be for retirement in 1966 with the low point at 1982.

So running this tool is all fine and dandy but can you live with the worst case scenario? And that probably is not the worst case for a Monte Carlo simulation. If my portfolio declined to 36% of it's present value I'd be plenty worried. I would also loose a lot of sleep and drive my DW crazy. Would I have the guts to stay the course (investment wise) ? Hopefully I won't be tested on that one .

I have to add this. For years I've felt that our generation (dubbed the boomers by someone) will be tested in a way like our parents were with the great depression and WWII. We can all speculate on how that might come about. What I'm trying to get at is that people should really look at the results of these FIRECalc simulations and look at specifically the worst case drawdown spreadsheet numbers. Try to internalize this so as to prepare yourself psychologically. Does your current portfolio risk characteristics mesh with your ability, willingness, and need to take risks? Only you can answer that one.

Les

[audreyh1]

Quote:

Originally Posted by cute fuzzy bunny Be awfully careful about those super long runs.

I thought I already explained why I don't "sweat it" on those super long runs (or strive for 100% success either)! I don't adjust withdrawals for inflation.

Audrey

[cute fuzzy bunny]

Oh, that'll do nicely...

Mostly just waving the checkered flag for newer members who havent been part of the firecalc/long runs discussions.

[audreyh1]

Fair enough!

Audrey

[Lusitan]

Well I'm planning for a 60 year run, so the long run limitations of FIREcalc are of interest to me. I'll have to factor in some additional uncertainty when I'm playing with these calculators.

One other thing I noticed about FIREcalc is that it doesn't allow for an asset allocation that changes over time. In particular, it doesn't allow for a change between one's asset allocation during the saving/investing years and the asset allocation during the FIRE/withdrawal years. Most of us (relatively) "young dreamers" are heavily invested in equities right now but may not be when we hit FIRE. Just more food for thought when thinking about these results.