Mixed Portfolio - what am I doing wrong?

[Japple55]

Hello,
First, a great program. Been looking for a long time for something like this!

Everything seems to work fine except when running using the “mixed portfolio option” the maximum number of cycles comes up at 49; this seems about 9 or 10 years too short.

1927 + 49cycles = 1976 + 30years = 2006. Why not 57 cycles to 2014?

I’m using $1m and 30 year retirement, and a mixed portfolio of 50% LT Corp Bonds and 50% SP500. Any ideas?

Thanks,
Jim

 

[REWahoo]

[-]You aren't doing anything wrong, it is a data availability issue.

I was unsuccessful in my attempt to find a link to an explanation by Dory 36 (creator of FIRECalc), but here is my recollection of why you aren't seeing the number of cycles you expected.

From "How it Works"

FIRECalc's standard model uses the overall US stock market performance.

Dory 36 found this data going back to 1872 and included it in the default calculator. Unlike data on the overall performance of the market, history on the individual components in the mixed portfolio option aren't available back that far. Net result - fewer cycles available for your calculation.[/-]

EDIT: Realized I answered a question not asked. Never mind.

 

[Senator]

If you want to experiment more, use half as much money and go half as long. That will give more cycles, and maybe the same result or more conservative. Actually a longer term horizon should be better in regards to market returns.

 

[Japple55]

Thanks REWahoo and Senator. I was thinking it may have been related to data. However, the disclosure "(based on performance since 1927)" implies that data begins in 1927 for all asset classes. I would have thought the disclosure would show the most recent year for which data existed for the entire set.

I just tested using 100% allocation for each asset class individually and only got 49 cycles for each one, so apparently it only computes for a maximum of 49 cycles regardless of allocation. There's no way to tell what the data limitations are.

I then tested (separately) 100% allocations to SP500 and 1 month Tbills, and the spreadsheet showing performance for each from 1927 appeared to closely match actual performance for those asset classes. I would think if data was missing, it would be early in the data stream, so now I'm not so sure the problem is missing data.

All that leads me to believe that a setting telling Firecalc how many years to compute is incorrect. (IMHO)

 

[ERD50]

[-]I'll make a guess, but the support staff here will need to validate this:

The original author has not been involved for some time ( ~ 2006? coincides with your data observation)[/-]

Good catch!!! OK, here's the answer:

http://www.early-retirement.org/forums/f36/firecalc-updated-for-q4-2008-a-41894.html#post773514

By the way... the only updated data I have is the "standard" market and interest data. So the "mixed portfolio" will not include the new data. ...

So if the current support staff cannot obtain/enter this data, it would be nice to at least add a note on that selection to avoid these surprises. That doesn't fully explain the 2006-2008 gap, but I'd bet it's related.

-ERD50

 

[Japple55]

Thanks ERD50, that explains it. Lacking the most current data is the last thing I would have thought of, but then I just discovered Firecalc, and know nothing of the history. Since that data is readily available, I hope whoever is supporting the site will be able to get it updated or, as you suggested, at least add a note indicating same.