Firecalc now shows a 5.1% failure rate for the 4% rule

Ask your investment, budget, and other money related questions here
Post Reply
jacob
Site Admin
Posts: 9926
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Firecalc now shows a 5.1% failure rate for the 4% rule

Post by jacob » Sat Jun 09, 2018 12:45 pm

(For the standard 30 year period, which obviously is way too short for extreme early retirement.)

They must have added the last two decades of the 21st century.

For a 60 year period, SWR=4% now results in 17.2% failures.

2Birds1Stone
Posts: 227
Joined: Thu Nov 19, 2015 11:20 am

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by 2Birds1Stone » Sat Jun 09, 2018 12:48 pm

What was the failure rate previously @ 30 years?

jacob
Site Admin
Posts: 9926
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by jacob » Sat Jun 09, 2018 12:51 pm

IIRC it was 0%

User avatar
FBeyer
Posts: 870
Joined: Tue Oct 27, 2015 3:25 am

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by FBeyer » Sat Jun 09, 2018 1:05 pm

The last two decades of the 21st century? Is that 2080 and 2090? Are small cap stocks still the place to be ;)

I take it your lesson is still: Learn to invest, learn to think for yourself, and be flexible?

User avatar
BRUTE
Posts: 3166
Joined: Sat Dec 26, 2015 5:20 pm

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by BRUTE » Sat Jun 09, 2018 1:47 pm

maybe "latest two decades" ;)

jacob
Site Admin
Posts: 9926
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by jacob » Sat Jun 09, 2018 2:31 pm

:P

User avatar
Fish
Posts: 375
Joined: Sun Jun 12, 2016 9:09 am

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by Fish » Sun Jun 10, 2018 1:51 am

@jacob - At first I thought it was the 0.18% expense ratio, but the Wayback Machine confirms that Firecalc has always used the same default assumptions: 750k portfolio, 30k annual spend, 30 years, 75/25 portfolio, 0.18% ER. Reading some PF forums shows that the failure rate has consistently hovered around 5% in recent history:

October 2006: 6 of 106 cycles failed (5.6% failure rate, inferred)
July 2012: 6 of 111 cycles failed (5.4% failure rate)
October 2015: 6 of 115 cycles failed (5.2% failure rate)
June 2018: 6 of 117 cycles failed (5.1% failure rate)

jacob
Site Admin
Posts: 9926
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by jacob » Sun Jun 10, 2018 9:09 am

@Fish - Thanks! Then I recalled incorrectly; probably because I've mostly run it with 3% WR at 60 years.

So a better title would be: Firecalc has shown a ~5% failure rate for the 4% rule since 2006.

User avatar
BlueNote
Posts: 451
Joined: Sat Jun 08, 2013 6:26 pm
Location: Toronto, Canada

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by BlueNote » Thu Jun 14, 2018 6:42 am

It looks like cFireSim is reporting close to the same 5% result too.

As part of my own due diligence I look for an alternate tool in these types of data heavy simulations to see if they produce similar output. In this case the results look substantially the same in both tools.

Other tools that do things like MVO and NPV can produce significantly different outputs when there is only minor changes to inputs and these monte carlo/historical-driven-simulation models are no different. A couple of bad recessions in a row can mean you need to save for a few more years to be safe.

I've always thought 3.5% was closer to the safe mark because that's around where the SWR is , historically, for a cap weighted world portfolio which is what I lean towards.

NPV
Posts: 165
Joined: Fri Dec 06, 2013 9:41 am

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by NPV » Sat Jun 16, 2018 7:11 am

It is very interesting that 3% SWR results in 0% failure rate over both 30 and 60 year period (and over longer periods as well, however much less data is available for longer periods). Seems like 3% is the "magic number" - at least empirically speaking, to a degree that one exists.

jacob
Site Admin
Posts: 9926
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: Firecalc now shows a 5.1% failure rate for the 4% rule

Post by jacob » Sat Jun 16, 2018 7:50 am

One does exist. ~3% has been the real rate of economic return in various civilizations over thousands of years. See e.g. Homer's History of Interest Rates.

Post Reply