FIRE Projection... Thoughts?

[lostmonkey007]

<edited>

[MDM]

Here's a quick look.  See the 'Misc. calcs' tab in the case study spreadsheet to enter different numbers.

See tools such as www.cfiresim.com for more elaborate guesses.

Quick calculation of "Time to FI"
Planned Withdrawal Rate	WR	4.0%
Annual Savings Invested	S	40,000	$/yr
Annual Expenses in Retirement	E	45,000	$/yr
Current Assets Invested	A	130,000	$
Investment return	r_	5.0%
Time to FI	t	14.9	yr
Desired time to FI	t	9.0	yr
Annual Savings Needed	S	83,737	$/yr

[Radagast]

I didn't do no math. Here are a few basic comments.
1. Stock returns can be highly variable over short periods. You should keep in mind your 9 year expected wealth is $500,000 +/- 50%. I am sure you know this, just make sure you know it. If you actually have $300,000 at age 35 that is a perfectly reasonable outcome.

2. IIRC historical real stock returns for the US have been around 6.7%. If you honestly expect 4.5% real than perhaps you should also expect a 3% SWR.

[MDM]

If you honestly expect 4.5% real than perhaps you should also expect a 3% SWR.

A 4% Withdrawal Rate (WR) works (under Trinity study rules - 30 year retirement) with 1.22% real Compound Annual Growth Rate (CAGR).

A 4.5% real CAGR would sustain (again under Trinity study rules - 30 year retirement) a 6% WR.

Change 30 years to 55 years and the above numbers change to
3.35% CAGR
4.9% WR

All assuming that early returns are not bad enough to drive one to a negative balance, regardless of what the CAGR for the asset allocation ends up being at the end of the time period.

[Radagast]

If you honestly expect 4.5% real than perhaps you should also expect a 3% SWR.

A 4% Withdrawal Rate (WR) works (under Trinity study rules - 30 year retirement) with 1.22% real Compound Annual Growth Rate (CAGR).

A 4.5% real CAGR would sustain (again under Trinity study rules - 30 year retirement) a 6% WR.

Change 30 years to 55 years and the above numbers change to
3.35% CAGR
4.9% WR

All assuming that early returns are not bad enough to drive one to a negative balance, regardless of what the CAGR for the asset allocation ends up being at the end of the time period.

For now I'll have to take your word for it because the 4% rule always seemed close enough that I didn't bother getting into debates over semantics. What is your source for these numbers?

[Metric Mouse]

I would say a spouse would dramatically raise the chances of success for your plan - They could easily make up the shortfall in your needed income.

[MDM]

A 4% Withdrawal Rate (WR) works (under Trinity study rules - 30 year retirement) with 1.22% real Compound Annual Growth Rate (CAGR).
A 4.5% real CAGR would sustain (again under Trinity study rules - 30 year retirement) a 6% WR.

Change 30 years to 55 years and the above numbers change to
3.35% CAGR
4.9% WR

All assuming that early returns are not bad enough to drive one to a negative balance, regardless of what the CAGR for the asset allocation ends up being at the end of the time period.

For now I'll have to take your word for it because the 4% rule always seemed close enough that I didn't bother getting into debates over semantics. What is your source for these numbers?

This does assume constant (or at least not-too-bad-in-the-first-few-years) CAGR.

Given that, it's simply a calculation.  See cells F16:G22 on the 'Misc. calcs' tab of the case study spreadsheet, or just use Excel's RATE and PMT functions respectively to get the CAGR and WR numbers.

[Radagast]

A 4% Withdrawal Rate (WR) works (under Trinity study rules - 30 year retirement) with 1.22% real Compound Annual Growth Rate (CAGR).
A 4.5% real CAGR would sustain (again under Trinity study rules - 30 year retirement) a 6% WR.

Change 30 years to 55 years and the above numbers change to
3.35% CAGR
4.9% WR

All assuming that early returns are not bad enough to drive one to a negative balance, regardless of what the CAGR for the asset allocation ends up being at the end of the time period.

For now I'll have to take your word for it because the 4% rule always seemed close enough that I didn't bother getting into debates over semantics. What is your source for these numbers?

This does assume constant (or at least not-too-bad-in-the-first-few-years) CAGR.

Given that, it's simply a calculation.  See cells F16:G22 on the 'Misc. calcs' tab of the case study spreadsheet, or just use Excel's RATE and PMT functions respectively to get the CAGR and WR numbers.

Constant returns are probably not a good assumption for this case. I used the monte carlo simulator at https://www.portfoliovisualizer.com/monte-carlo-simulation#analysisResults
Using an annualized 4.5% return (assumed real), 15% annualized standard deviation, and fat tailed distribution, a 4% constant withdrawal rate succeeded 75% of the time at 30 years, and about 50% for 50 years. A 3% withdrawal rate succeeded 89% of the time at 30 years and 71% of the time at 50 years.

Using real annualized returns of 1.22% in the above, the 30 year success rate is about 34%. Using real annualized returns of 3.35% for a 50 year withdrawal results in a 34% success rate. Not 100% safe.

To check the simulation, 6.7% returns at 4% constant withdrawal succeeded about 90% of the time at 30 years and 79% at 50 years. I think that is a little lower success rate than actually happened so the simulation might be slightly biased.

The 4% rule was true over a period with about 6.7% annualized returns. Doing some simple math, 4/6.7, it seems like an appropriate safe withdrawal rate is very approximately 60% of the annualized returns of the period. If we assume 5% real returns then 3% is probably an appropriate constant safe withdrawal rate. In general volatility seems to be more damaging to withdrawal rates than most people assume. It is also more beneficial to accumulation than most people assume.

[MDM]

Constant returns are probably not a good assumption for this case.

Yes, assuming a constant return is almost certainly wrong. :)

I used the monte carlo simulator at https://www.portfoliovisualizer.com/monte-carlo-simulation#analysisResults

And all of these simulations are almost certainly wrong also. ;)

Because...All Models Are Wrong, Some Are Useful.

As noted in Monte Carlo Simulations Pros and Cons - Bogleheads.org, "...had once written a Monte Carlo program for predicting electron beam - solid material interactions. Predictions were dead on because the underlying physics was well understood. But for predictions of investment returns there are only very rough estimates of what the future might look like. So the Monte Carlo predictions will only be very rough estimates."

See also Portfolio Visualizer Questions - Bogleheads.org for pro/con discussion of that tool.

For those unfamiliar with Safe Withdrawal Rates, see Safe withdrawal rates - Bogleheads, Trinity study update - Bogleheads, and links therein.

And there's the 500+ posts in Stop worrying about the 4% rule.

It would be interesting if Portfolio Visualizer noted the CAGR distributions at the end of its 5000 simulations.  One common observation of pure Monte Carlo methods for this purpose is that historical "reversion to the mean" behavior is often violated.  Depending on one's perspective, that can be either a feature or a bug.

[Radagast]

Constant returns are probably not a good assumption for this case.

Yes, assuming a constant return is almost certainly wrong. :)

I used the monte carlo simulator at https://www.portfoliovisualizer.com/monte-carlo-simulation#analysisResults

And all of these simulations are almost certainly wrong also. ;)

Right. So while the monte carlo simulations will not be at all similar to the future (obviously, as I just used numbers from the thread and some vague assumptions) it is somewhat useful. While using constant returns are neither accurate nor useful because they present a single unlikely outcome which is not conservative. The monte carlo simulator presents a range of outcomes which demonstrates that the outcome is unknown, and also that in retirement volatility is more likely do hurt than to help.

[MDM]

While using constant returns are neither accurate nor useful because they present a single unlikely outcome which is not conservative.

Usefulness is in the eye of the beholder.  I find the 1.22% number useful.  Why?  Because it's another way to reinforce the idea that a 4% Safe Withdrawal Rate is based on the worst of recorded history (including the worst sequences of returns).  Thus, to the extent the future is no worse than the worst of the past, there is no need to plan for 3% (or anything less than 4%) withdrawal rates.