FIRECalc - Negative Accumulation

[success108]

It appears that FIRECalc is accumulating negative balances. For example, if a scenario goes negative in year 15; that negative balance, plus future withdrawals, are then accumulated at the scenario return until the end of the projection.

This doesn't make sense in practice: when your stash is depleted, you wouldn't short sell the market. What's worse is that this likely grossly exaggerates how negative ending balances can be, showing a much scarier picture than what history would have delivered.

Is this the intended behavior? Or is it a simplification - everything is always accumulating instead of it being conditional on a positive balance? Or maybe I'm wrong and not looking at the output accurately.

 

[Gumby]

I don't understand your question. If a FIRECalc scenario ever goes negative during the period you specify, then it counts as a fail. Your success probability is 1 - (fails/total runs). The purpose of FIRECalc is not really to tell you what your final balance may be. It is to give you a probability of success (i.e. - non-failure) based on historical data with the parameters you enter. Because, as you note, once you're out of money, you're out of money.

 

[COcheesehead]

If you are -$1 or -$100,000, you are out of funds. You failed.
You really would like to avoid this situation.

 

[NW-Bound]

This doesn't make sense in practice: when your stash is depleted, you wouldn't short sell the market. What's worse is that this likely grossly exaggerates how negative ending balances can be, showing a much scarier picture than what history would have delivered...

Once you have depleted your stash, in order to continue spending as usual you would have to go into debt, assuming somebody would be willing to loan you money.


That's what the negative balance is. You would not be short selling the market. You just have a negative net worth, and it gets deeper and deeper underground.

 

[CaptTom]

I had the same question as the OP. That negative accumulation looks odd in this context. My portfolio isn't going to see a decreasing negative number.

I like the idea of considering it borrowing.

Either way, it is a clear indication of the magnitude of the failure. Since (hopefully) only some of the lines on the chart go negative, you get a sense of how far from the average both a massive failure and a massive success actually are.

I chalk it up to just a funny quirk of how FIRECalc displays results.

 

[success108]

Thank you all for your replies.

If the sole purpose of FIRECalc is to show historical success rates, regardless of the magnitude of failure; then the question is irrelevant. I would argue that the magnitude of the worst failure does have value: a -$1K worst case can be dismissed, while a -$1M worst case might warrant more consideration.

FIRECalc output shows a range of results, and some might interpret the lower end of the range as how much they would have had to fund by working/borrowing in the worst-case scenario. This negative accumulation, if it really is happening, makes the lower end value potentially misleading (looks much worse than it should).

Based on some of the replies, it sounds like negative accumulation is occurring. If that can be confirmed, then we can safely say the lower end of the output range should be taken with a grain of salt. This in no way invalidates the main purpose of FIRECalc - historical success rate - which it does remarkably well.

 

[NW-Bound]

In real life, if someone was about to deplete his investable assets, he would start thinking about liquidating his house, which is hopefully paid for, and getting into a used motorhome.

And contrary to FIRECalc presumption that a person would continue to spend as nothing had happened, most likely that the person would turn down the spigot to a trickle before it ran dry.

 

[ERD50]

OP is correct. FIRECalc really doesn't handle negative balances in a realistic, or accounting correct way (but it doesn't matter, I agree with OP that it is just a simplification).

To see this: Take the FIRECalc defaults, then increase spending from $30K to $70K to show more failures to make it clearer. In real life, once your portfolio went to zero, the balance should decline linearly at the spend rate of $70K a year to show this annual accumulation of the debit. But they mostly drop far more than that. It appears that a 20% gain in the market would take the balance (a negative number) and multiply it by 1.2 (and subtract the $70K), forcing the balance even more negative. Example:

Year 12, you go to zero. Year 13, you are at -$70K, and the portfolio returns 20%. You are now at -$70K * 1,2 = -$84K. Next year, another $70K spend, and say a 10% gain, so now -$84K-$70K times 1.1 = $-169.4K. And so on, it keeps compounding.

If that wasn't the case, there is no way you could drop to -$6M after ~ 22 years of $70K withdraws, that is ~ negative $1.5M (plus ot minus some inflation?).

Thank you all for your replies.

If the sole purpose of FIRECalc is to show historical success rates, regardless of the magnitude of failure; then the question is irrelevant. I would argue that the magnitude of the worst failure does have value: a -$1K worst case can be dismissed, while a -$1M worst case might warrant more consideration. ...

Yes, but this is reflected in how soon it goes negative. FIRECalc doesn't spit that number out, but you can see it on the graph (squint!).

.....

Based on some of the replies, it sounds like negative accumulation is occurring. If that can be confirmed, then we can safely say the lower end of the output range should be taken with a grain of salt. This in no way invalidates the main purpose of FIRECalc - historical success rate - which it does remarkably well.

Yes, it is kinda confusing, but I can see where the author just didn't want to bother with this case. But maybe it should be spelled out somewhere (and the year of first, or all, failures).

-ERD50

 

[success108]

Thanks ERD50 for confirming. I was doing the same as you to estimate the worst-case ending balance: approximating where the bottom line crossed zero and multiplying remaining years by the desired withdrawal amount.

Thanks again.

 

[ERD50]

Thanks ERD50 for confirming. I was doing the same as you to estimate the worst-case ending balance: approximating where the bottom line crossed zero and multiplying remaining years by the desired withdrawal amount.

Thanks again.

Glad to have helped. I think your approach is the right way to view it. Not sure it really matters, once you've gone to zero, but I suppose someone might figure they could live off CC debt the last few years? .

-ERD50