FIRECalc, allocations and expenses

[MichaelB]

It's been a while since I ran scenarios and numbers through FIRECalc, so this morning I looked at the survival rates of of different equity / FI allocations, and also the effect of using short term treasuries instead of the default long term. All modeling a withdrawal rate of 4% over 30 years.

Changing equity allocation from 70% to 40% resulted in a minor decline in survival, from 94.9% to 93.2%. Shortening the bond maturity improved the survival rate, the new range between 96.6% - 95.7%.

This was expected, similar to what I recalled from past runs. Then I changed the investment expense ratio from 0.18 to 0.5, which I've never done in the past, and the portfolio survival rate declined across the board and was significantly lower with the higher fixed income allocation porfolios. With a 40% equity allocation and 60% short term treasuries, the survival rate with the higher expense fell from 95.7% to 87.2%. Ouch.

My takeaway here is I should spend a bit more time looking at my portfolio expense levels and less on the asset allocation.

[braumeister]

Quote:

Originally Posted by MichaelB Changing equity allocation from 70% to 40% resulted in a minor decline in survival, from 94.9% to 93.2%. Shortening the bond maturity improved the survival rate, the new range between 96.6% - 95.7%. ... My takeaway here is I should spend a bit more time looking at my portfolio expense levels and less on the asset allocation.

A very good exercise.

My take on all results in calculators like this is that whatever percentage I get is plus or minus 5%. So I don't really consider those changes significant or meaningful.

But I definitely agree with you on the expense ratios. In the last year I've brought mine down from about 0.49% to the present .08% on my total portfolio. I'm much happier with it this way.

[walkinwood]

I agree. A good exercise.
I sold off my last non-vanguard fund last year. I'm now down to two active funds which are difficult to sell because I have a lot of cap gains. But they're doing well.
My overall expense ratio is now .14%! I can just leave firecalc at its default value.

[audreyh1]

Quote:

Originally Posted by MichaelB It's been a while since I ran scenarios and numbers through FIRECalc, so this morning I looked at the survival rates of of different equity / FI allocations, and also the effect of using short term treasuries instead of the default long term. All modeling a withdrawal rate of 4% over 30 years. Changing equity allocation from 70% to 40% resulted in a minor decline in survival, from 94.9% to 93.2%. Shortening the bond maturity improved the survival rate, the new range between 96.6% - 95.7%. This was expected, similar to what I recalled from past runs. Then I changed the investment expense ratio from 0.18 to 0.5, which I've never done in the past, and the portfolio survival rate declined across the board and was significantly lower with the higher fixed income allocation porfolios. With a 40% equity allocation and 60% short term treasuries, the survival rate with the higher expense fell from 95.7% to 87.2%. Ouch. My takeaway here is I should spend a bit more time looking at my portfolio expense levels and less on the asset allocation.

What short-term did you use? Did you try intermediate = 5 yr treasuries?

I'm not surprised that reducing fixed income durations from long improved things. Long bonds don't respond to inflation nearly as quickly as intermediate to short term.

Yeah - expenses are a drag.

[Ready]

I read John Bogle's entire book on the benefits of low cost indexed funds, but I could have reached the same conclusion from reading your post. That pretty much sums up his years of work. Don't worry about AA, just worry about expense ratios.

[pj.mask]

How did final portfolio value change - some scenarios crater for sure no matter what you fiddle with but if going from 70% to 40% stocks drops your average ending balance significantly that can be something to consider without altering the results you provided

[street]

Interesting observation.

[Dtail]

On a related note, I don't currently include the expense ratio for the fees in my Stable Value fund, which I think has been mentioned in the past on this site. However, I do use it's value in my total invested assets (denominator) to arrive at the composite expense ratio.
Agree/Disagree?

[Dtail]

Just wondering if anyone had a quick answer to my above post?

[SecondCor521]

Quote:

Originally Posted by Dtail On a related note, I don't currently include the expense ratio for the fees in my Stable Value fund, which I think has been mentioned in the past on this site. However, I do use it's value in my total invested assets (denominator) to arrive at the composite expense ratio. Agree/Disagree?

Why would you do it that way? I can't see any reason to include assets and ignore fees other than to give someone a misguided impression of their expense ratio.

[MichaelB]

Quote:

Originally Posted by audreyh1 What short-term did you use? Did you try intermediate = 5 yr treasuries? I'm not surprised that reducing fixed income durations from long improved things. Long bonds don't respond to inflation nearly as quickly as intermediate to short term. Yeah - expenses are a drag.

Yes, not short term, but shorter term 5 year.

Quote:

Originally Posted by SecondCor521 Why would you do it that way? I can't see any reason to include assets and ignore fees other than to give someone a misguided impression of their expense ratio.

Yes, I also wouldn't ignore the fees.

[Dtail]

Quote:

Originally Posted by SecondCor521 Why would you do it that way? I can't see any reason to include assets and ignore fees other than to give someone a misguided impression of their expense ratio.

For regular stocks and bonds, I totally agree.
My thoughts on a Stable Value (401k fund) is that there is a set net yield to the investor which is not totally predicated strickly on market returns in those investments, but a short term yield guarantee (which of course changes over time).
Thus the fee expense embedded in this yield is not totally relevant, but the denominator does include one's SV investment.
Making any sense?

[audreyh1]

Quote:

Originally Posted by MichaelB Yes, not short term, but shorter term 5 year. Yes, I also wouldn't ignore the fees.

OK - so it was intermediate term - which is what I think of 5 year as being.

The FIRECALC models I run use 5 yr treasury because I am mostly invested in intermediate (~5yr) bond funds with a small amount in short-term ~2.5 year bond funds.

[SecondCor521]

Quote:

Originally Posted by Dtail For regular stocks and bonds, I totally agree. My thoughts on a Stable Value (401k fund) is that there is a set net yield to the investor which is not totally predicated strickly on market returns in those investments, but a short term yield guarantee (which of course changes over time). Thus the fee expense embedded in this yield is not totally relevant, but the denominator does include one's SV investment. Making any sense?

Ah. I'm not too familiar with Stable Value funds.

If the return is net of fees, I think that is OK as long as when you're running FIREcalc (or whatever other tools), you use the net return for your investment performance.

In other words, either account for them explicitly as a fee that reduces gross performance, or implicitly by reducing your net performance by fees.

[Dtail]

Quote:

Originally Posted by SecondCor521 Ah. I'm not too familiar with Stable Value funds. If the return is net of fees, I think that is OK as long as when you're running FIREcalc (or whatever other tools), you use the net return for your investment performance. In other words, either account for them explicitly as a fee that reduces gross performance, or implicitly by reducing your net performance by fees.

Agree.