FIRECalc default portfolio

[explanade]

It's 75% equities?

What's the best way to enter 50/50?

Do you have to figure out the different classes of equities in your portfolio?

And how do you represent different types of bond funds and cash?

 

[REWahoo]

The AA options available are all on the "Your Portfolio" tab.

 

[explanade]

So just keep Total Market selected and then change 75 to 50 there?

 

[explanade]

Presumably the AA should reflect the portfolio after you RE?

Changed it from 75 to 50 and it didn't change the results much, just lower average portfolio values.

That is, still 100% whether using 75% or 50% stocks.

Kind of curious because I'm trying 3.5% inflation and I had to raise the portfolio starting amount by $100k to go from 1 failed cycle in 40 years to 0 failed cycles.

And then lowering the stock percentage from 75 to 50% still resulted in 100% success.

 

[REWahoo]

Presumably the AA should reflect the portfolio after you RE?

Yes, that will give you a 50/50 AA

Changed it from 75 to 50 and it didn't change the results much, just lower average portfolio values.

You probably won't see much change in failure rates unless you drop your equity allocation to under 35 or increase it to more than 80.

 

[Animorph]

Changed it from 75 to 50 and it didn't change the results much, just lower average portfolio values.

Exactly.

 

[ERD50]

...

Kind of curious because I'm trying 3.5% inflation ...

I wouldn't (and don't) 'try' any inflation number other than the historical ones. The double digit inflation of the 80's is what killed portfolios.

If you are trying to test your portfolio's resistance to failure, 3.5% inflation is a walk in the park with pretty girls on each arm, and a few walking in front to provide scenery. That doesn't tell you much about whether you could crawl up a steep muddy mountainside in a downpour, with big guys with big boots kicking you in the teeth and stepping on your fingers as you cling to roots to hang on.

-ERD50

 

[NW-Bound]

One important thing to look for when changing AA is the minimum value of all runs, or the worst case. You should look at the chart, not the numbers that FIRECalc reports.

Generally, a high-stock AA will have the best good cases, but also worse bad cases, in other words more uncertainty. Do you feel lucky?

 

[ERD50]

....
And then lowering the stock percentage from 75 to 50% still resulted in 100% success.

This called quantization. Pass/fail are binary, so you won't see an output change with small changes in input, unless your inputs are right on the edge of pass/fail.

Two 'solutions', one you already discovered:

1) Look at non-binary output, the average portfolio ending values. You have seen that they change in magnitude with the AA changes.

2) Probably more to your point - increase your spending until you get solidly in the 'fail' range (say 85%). Then you can see some more linear changes in output success % to changes in inputs.

IOW, you are probably in the 150% success rate range (if we could express it that way), so even something way worse still gets counted as "pass", even if it drops from a relative 150% down to 100%. They are both 'pass' - you get no added information. But dropping from 85% to 82% tells you something.

-ERD50

 

[ERD50]

...

Generally, a high-stock AA will have the best good cases, but also worse bad cases, in other words more uncertainty. Do you feel lucky?

I'm not so sure about that.

If you are looking at the negative end-of-portfolio values in FIRECalc, I believe they are misleading. Of course, negative portfolio values are a bit non-nonsensical anyhow, but I think they can be useful to get some relative idea of how bad things are.

The problem is, I suspect that FIRECalc continues to multiply the remaining portfolio times that year's market returns. So a -$100,000 in equities becomes -$130,000 if the market has a good year of PLUS 30% returns! It's as if you shorted the market.

So that could make high EQ AAs appear more negative if the market is turning around in those final years.

A special case should be made for portfolios below zero, just subtract spending from the negative balance and make an inflation adjustment - I think that would keep things fair for comparison purposes.

-ERD50

 

[explanade]

How do you select "historical" inflation?

The choices are PPI, CPI or some number. Assuming you think inflation is going to be hotter than CPI ...

 

[explanade]

This called quantization. Pass/fail are binary, so you won't see an output change with small changes in input, unless your inputs are right on the edge of pass/fail.

Two 'solutions', one you already discovered:

1) Look at non-binary output, the average portfolio ending values. You have seen that they change in magnitude with the AA changes.

2) Probably more to your point - increase your spending until you get solidly in the 'fail' range (say 85%). Then you can see some more linear changes in output success % to changes in inputs.

IOW, you are probably in the 150% success rate range (if we could express it that way), so even something way worse still gets counted as "pass", even if it drops from a relative 150% down to 100%. They are both 'pass' - you get no added information. But dropping from 85% to 82% tells you something.

-ERD50

For 1) you're talking about the spreadsheet output?

And for 2) you're looking for more negative ending portfolio values to see how spending changes the results.

Makes sense, spending is the one easiest to control, assuming you padded it with a lot of discretionary items that you can cut back on?

 

[NW-Bound]

I'm not so sure about that.

If you are looking at the negative end-of-portfolio values in FIRECalc, I believe they are misleading. Of course, negative portfolio values are a bit non-nonsensical anyhow, but I think they can be useful to get some relative idea of how bad things are.

No, I do not look at cases where the portfolio value goes negative, only ones where it was skimming the zero line like an Exocet or Harpoon missile.

And the greater dispersion only shows up at higher WRs, if my memory serves.

 

[ERD50]

...

Generally, a high-stock AA will have the best good cases, but also worse bad cases, in other words more uncertainty. Do you feel lucky?

OK, came back with two quick runs, using a 3% WR (30K/1M) to avoid the negative portfolio potential problem.

@ 75/25: 0.55M low; 2.6M avg; 6.6M high
@ 25/75: 0.25M low; 1.1M avg; 4.0M high

So the lower AA gave lower lows, lower averages, and lower highs.

It's not so easy to get interim dip info with FIRECalc, you can only eye up the chart, but even zooming and using a ruler to scale, the dips look to be higher (that is, less bad - above ~ $500K) with the higher AA in this case.

-ERD50

 

[NW-Bound]

It is true that very low stock AA will not do that well.

What I was describing was for stock AA of greater than 50%, and also at a higher WR.

Here are 3 cases starting at $1M, and WR of 3.5%.

90% stock, $-5,215 to $7,429,349, with an average of $2,764,220.

60% stock, $144,278 to $5,036,085, with an average of $1,739,851.

50% stock, $165,501 to $4,604,384, with an average of $1,446,783.

 

[ERD50]

It is true that very low stock AA will not do that well.

What I was describing was for stock AA of greater than 50%, and also at a higher WR.

Here are 3 cases starting at $1M, and WR of 3.5%.

90% stock, $-5,215 to $7,429,349, with an average of $2,764,220.

60% stock, $144,278 to $5,036,085, with an average of $1,739,851.

50% stock, $165,501 to $4,604,384, with an average of $1,446,783.

OK, I extended that down to 40/60AA and got a little higher lows (ending portfolio low of ~ $175K), and it got worse with lower AA, so somewhere ~ 40-50 stock AA seems like a 'sweet spot' for smoothing dips, at least for a 3.5% WR and this data set.

-ERD50

 

[NW-Bound]

Yes.

Now, keep the equity AA around 50%, and bring in more value stocks (use the more detailed mixed portfolio option). You will love the results.

 

[REWahoo]

Now, keep the equity AA around 50%, and bring in more value stocks (use the more detailed mixed portfolio option). You will love the results.

Wellington.

 

[NW-Bound]

Not Wellesley?

 

[REWahoo]

Maybe a little of both...

 

[NW-Bound]

I am still trying to catch (and ride) the top lines in the high-stock AA FIRECalc runs, but when I don't want to do that anymore, might just put 25/25/25/25 into Wellesley/Wellington/Dodge&Cox/Dodge&Cox Balanced, then sit back and watch them race.

 

[REWahoo]

I am still trying to catch (and ride) the top lines in the high-stock AA FIRECalc runs, but when I don't want to do that anymore, might just put 25/25/25/25 into Wellesley/Wellington/Dodge&Cox/Dodge&Cox Balanced, then sit back and watch them race.

+1

I was once a big fan of Dodge & Cox Balanced - it was a core holding during my 401k accumulation period. Then came 08/09 and the fund's performance reflected far too many symptoms of yield chasing - overweight on financial stocks. Although the fund recovered nicely, I lost faith in the management team and moved to Vanguard.

 

[NW-Bound]

I also saw that DODBX was and still is more aggressive than Wellesley/Wellington, but I like to have variety, and maybe even rebalance between them. Not as exciting as holding individual stocks, but enough variation to have something to look at.

I currently have only a token in the above funds, 1 or 2% in each, so that I will reminded to look at their performance and not forget about them.