What I'm doing wrong - Firecalc

[retire2020]

It's strange what I see with Firecalc... Two identical scenario with different years(40 vs 50 years of payout) analyzed for 100% success rate to withdraw 120K/year. 50 years period require less portfolio amount, 3.194M vs 3.318M for 40 years...Would not 50 yrs payout period require higher portfolio amount? Is something wrong with Firecalc or I'm making a mistake somewhere?


Spending: 120K
Portfolio: 3M
Years: 40 and then 50
I'm putting 30K/yr Social Security Income starting year 2031
Then go to Investigate tab and choose 2nd from bottom option -- 100% success rate and click on 'starting portfolio value' and hit the 'submit' button

All other parameters are left at 'default' in Firecalc



40 years period says:
FIRECalc Results
Looking for a portfolio value that will result in 100% success rate . . . . . . . . . . . . . . . [done]

A starting portfolio of $3,318,054 provided a success rate of 100.0% (101 total cycles, of which 0 failed).

The highest residual portfolio was $45,953,181. The lowest was $-505, and the average was $11,265,945.

Here is a plot of the success rate of this portfolio value (in the center) and a few that are smaller and larger. The vertical axis shows the success rate.

50 years period says:
FIRECalc Results
Looking for a portfolio value that will result in 100% success rate . . . . . . . . . . . . . . . [done]

A starting portfolio of $3,194,722 provided a success rate of 100.0% (91 total cycles, of which 0 failed).

The highest residual portfolio was $62,333,409. The lowest was $4,042, and the average was $17,223,182.

Here is a plot of the success rate of this portfolio value (in the center) and a few that are smaller and larger. The vertical axis shows the success rate.

 

[Animorph]

FIRECalc uses historical data, in chronological order. It starts at the earliest year, runs out 40 or 50 years and determines your result for that run. Then it starts with the next year and runs 40 or 50 years. When it gets to within 40 or 50 years of its last year of data, that's as far as it can go. It can't calculate what a 50 year retirement starting in 1980 looks like. So with the 40 year period you get 10 more tests, starting a little closer to today's date. Apparently one or more of those starting years that the 50 year period must skip was a portfolio killer.

 

[Midpack]

I've seen that phenomena before as well, with 45-50 year results seeming counterintuitive. It may be as the reply/post above describes, beyond that I don't know either - I can only confirm it's not something you're doing wrong (or we both are).

But planning out to 40-50 years is beyond the scope of most people I'd guess, so I never pursued it. Sorry I don't have a better answer...and FIRECALC support seems to have largely ended with the original author. Few donations = little support, vice versa, repeat, would be my guess why...

I've thought about writing my own FIRECALC, but I have no illusions it would be easy with all the bells & whistles FIRECALC allows.

 

[retire2020]

Thanks...It was not me then. I checked between 30-35-40 and results were as expected - higher initial portfolio value with increased no. of pay out period.

 

[Rustward]

The SS might have something to do with it also.

In the 40 year plan, you will collect SS in 55% of the total plan years.
In the 50 year plan, you will collect SS in 64% of the total plan years.

 

[Katsmeow]

I think that animorph's answer is the one most likely to be correct. The 50 year period necessarily covers fewer examples than the 40 year. The 40 year may have some that were much more negative that don't show up as so negative when adding in another 10 years to be a 50 year period.

 

[youbet]

FIRECalc uses historical data, in chronological order. It starts at the earliest year, runs out 40 or 50 years and determines your result for that run. Then it starts with the next year and runs 40 or 50 years. When it gets to within 40 or 50 years of its last year of data, that's as far as it can go. It can't calculate what a 50 year retirement starting in 1980 looks like. So with the 40 year period you get 10 more tests, starting a little closer to today's date. Apparently one or more of those starting years that the 50 year period must skip was a portfolio killer.

Yep. With the 50 year test you miss out on testing retiring into the high inflation levels that began in the 70's.