Are these firecalc results right?

[JD1967]

Hi everyone ... new to the forum ... have run a series of firecalc scenarios and cannot believe some of the numbers I'm getting .. .either i'm doing something wrong or the calculator is off ... if someone could comment/help, i'd very much appreciate ... here goes

Current info ... age 51
current IRA portfolio 1,675,000
current spend 150,000 (highly exaggerated / includes hcare)

Panel 2 (Other Inc Spending) - 12000/yr SSec starting in 2029
- 5400/yr Pension starting in 2032

Panel 3 (Not Retired?) - Retiring in 2026
- not adding anything to portfolio

Panel 4 (Spending Models) - Inflation assumption - CPI
- Constant spending power

Panel 5 (Your portfolio) - Mixed portfolio - 30% LCap Value 70% SCap Value

Panel 6 (Portfolio changes) - none

Panel 7 (Investigate) - using 1970 as the starting retirement year



My results

Because you indicated a future retirement date (2026), the withdrawals won't start until that year. The tested period is 8 years of preretirement plus 32 years of retirement, or 40 years.

FIRECalc looked at the 39 possible 40 year periods in the available data, starting with a portfolio of $1,675,000 and spending your specified amounts each year thereafter.

Here is how your portfolio would have fared in each of the 39 cycles. The lowest and highest portfolio balance at the end of your retirement was $-35,452,228 to $262,525,534, with an average at the end of $85,044,449. (Note: this is looking at all the possible periods; values are in terms of the dollars as of the beginning of the retirement period for each cycle.)

For our purposes, failure means the portfolio was depleted before the end of the 40 years. FIRECalc found that 2 cycles failed, for a success rate of 94.9%.


How can this possible be right ... any feedback would be great ... putting a lot of faith in these numbers

 

[mistermike40]

Why are you using 1970 as the starting retirement year?

 

[REWahoo]

I did a quick run of your numbers and got something significantly different:

FIRECalc Results
Your spending in every year after the first year will be adjusted for inflation, so the spending power is preserved.

Because you indicated a future retirement date (2026), the withdrawals won't start until that year. The tested period is 8 years of preretirement plus 32 years of retirement, or 40 years.

FIRECalc looked at the 107 possible 40 year periods in the available data, starting with a portfolio of $1,675,000 and spending your specified amounts each year thereafter.

Here is how your portfolio would have fared in each of the 107 cycles. The lowest and highest portfolio balance at the end of your retirement was $-7,573,349 to $19,503,572, with an average at the end of $1,248,792. (Note: this is looking at all the possible periods; values are in terms of the dollars as of the beginning of the retirement period for each cycle.)

For our purposes, failure means the portfolio was depleted before the end of the 40 years. FIRECalc found that 53 cycles failed, for a success rate of 50.5%.

Not sure where the discrepancy is but I think my numbers look far more realistic than what you came up with.

 

[USGrant1962]

I ran JD1967's numbers and got exactly the same results. REWahoo, you left the Your Portfolio selection on the default Total Market 75% - that generates exactly your numbers. So the problem appears to be in the U.S. Small Value data. While factor studies have shown historical outperformance for small cap value, AFAIK it is only a couple of percent annually. Seems unlikely to generate a plus-14X, minus-5X variance, even after 40 years.

I gave up using the "mixed portfolio" a long time ago because it has inadequate categories (especially lack of foreign stocks). It looks like there is also some weird data in there.

 

[JD1967]

No particular reason ... it autofilled 1960 I thought using 1970 would make it more accurate ... not sure if that is a good decision

 

[JD1967]

Thank you USGrant and Wahoo ...

What info do you suggest I use for the portfolio panel ... the 70/30 split I entered actually matches my ira

 

[REWahoo]

REWahoo, you left the Your Portfolio selection on the default Total Market 75% - that generates exactly your numbers.

I blame it on CWS - not the College World Series, Calculating While Sleepy.

 

[USGrant1962]

So I looked in Rick Ferri's "All About Asset Allocation" and he has an "efficient frontier" graph showing total stock market versus small cap value. It shows annualized returns from 1975 to 2009 as:

Total Stock ~12% with ~17% standard deviation
SC value: ~18% with ~25% standard deviation

He actually suggests that around 30% SC value could boost returns by 2% with little increase in standard deviation (risk).

So maybe the SC value data in FIRECalc is fine, it is just wilder than I thought. But of course this is historical, and since Fama and French pointed it out this factor is now relatively well known and may be priced in.

The other reason that I now avoid the "mixed portfolio" is that the data only goes back a limited time for many of the categories. From the examples above, it is only going back about 79 years. While that covers the Great Inflation it does not cover the Great Depression.

JD1967, you currently have a very high risk portfolio (with risk defined as variance or standard deviation). Your portfolio could easily be cut in half in the next bear market.
Your FIRECalc results are probably accurate with the data they have used, but again this SC value factor may or may not work going forward.

My suggestion - play around with the Total market portfolio, and the Investigate tab on FIRECalc. For example if you "Investigate changing my allocation" it will default to total stock market (FIRECalc cannot use mixed portfolio for that) and show that you are in the 70% success range with little/no bonds.

Also you might want to pick up "All About Asset Allocation" to get a good overview of risk versus reward in a portfolio.

 

[JD1967]

Thank you USGrant v much

 

[ncbill]

So I looked in Rick Ferri's "All About Asset Allocation" and he has an "efficient frontier" graph showing total stock market versus small cap value. It shows annualized returns from 1975 to 2009 as:

Total Stock ~12% with ~17% standard deviation
SC value: ~18% with ~25% standard deviation

He actually suggests that around 30% SC value could boost returns by 2% with little increase in standard deviation (risk).

Over on the Bogleheads forums the "small cap" premium is much debated.

After a couple of years reading various opinions & analyses presented there, I have come to agree with others over there who feel there is no longer a risk-free premium to small cap (growth or value)

I.e., any "extra" return is now coming from the extra risk inherent with small cap.

 

[Out of Steam]

Even assuming a prudent 20% haircut to current Social Security estimates by 2029, and your taking it at 62, $12K a year seems low.

You probably should get an estimate from SSA.

 

[Rustward]

When posting a firecalc scenario, it is much more accurate to right click on "Link To This Set of Data" on the results page, then click on "Copy Link Address", then paste the link address into your post.

This relieves people from having to reenter your data and possibly get it wrong.
It also shows precisely what you entered.