4 Percent Rule Can Fail! Here is How it Happens. Video by Ethan S. Braid, CFA

[big-papa]

Didn’t watch, not news…the 4% rule is not a withdrawal method in the first place.

Not only that, but it's not a "rule" in the first place. It's an historical observation.

It's really too bad that it's taken hold like it has. Even today, when a new academic paper shows up to look at one thing or another, they almost always start with SWR as the withdrawal method (whether it's 4% or some other number). Yes, it makes the math easy but that's about it.

Then there are a class of withdrawal methods that start with SWR, then add some "hacks" to it to improve it. Guyton-Klinger, Kitce's ratcheting method, Gummy's sensible withdrawals (there's a blast from the past!), Hebeler's autopilot, and many others.

There's another class of withdrawal methods that use amortization to calculate the withdrawals. These use the same math that is used for loans to calculate a withdrawal. Variable Percentage Withdrawals (aka VPW), ABW (amortization based withdrawals), and TPAW Planner are some examples, all of which are described over on bogleheads. Instead of the risk being one of completely running out of money before you run out of life, these result in individual withdrawals which vary each withdrawal period. In other words, instead of all of the risk being in the very last withdrawal, they spread the risk across all withdrawals and the portfolio itself lasts exactly as long as you planned for it to last. And the individual withdrawal risk can be mitigated by the choice of AA, a floor of fixed income which can include SS/Pension/TIPS Ladder/SPIA as well as simply not withdrawing the full amount calculated whenever it's not needed.

I'm using my own version of an amortization method and I consider the calculated withdrawal to be a "max" and ultimately only withdraw what I need. By not withdrawing the full amount that I could, the extra still invested is available for all future withdrawals. Over time, I would expect the portfolio to grow, giving us additional margin in the future.

Cheers.

 

[2HOTinPHX]

Not only that, but it's not a "rule" in the first place. It's an historical observation.

It's really too bad that it's taken hold like it has. Even today, when a new academic paper shows up to look at one thing or another, they almost always start with SWR as the withdrawal method (whether it's 4% or some other number). Yes, it makes the math easy but that's about it.

Then there are a class of withdrawal methods that start with SWR, then add some "hacks" to it to improve it. Guyton-Klinger, Kitce's ratcheting method, Gummy's sensible withdrawals (there's a blast from the past!), Hebeler's autopilot, and many others.

There's another class of withdrawal methods that use amortization to calculate the withdrawals. These use the same math that is used for loans to calculate a withdrawal. Variable Percentage Withdrawals (aka VPW), ABW (amortization based withdrawals), and TPAW Planner are some examples, all of which are described over on bogleheads. Instead of the risk being one of completely running out of money before you run out of life, these result in individual withdrawals which vary each withdrawal period. In other words, instead of all of the risk being in the very last withdrawal, they spread the risk across all withdrawals and the portfolio itself lasts exactly as long as you planned for it to last. And the individual withdrawal risk can be mitigated by the choice of AA, a floor of fixed income which can include SS/Pension/TIPS Ladder/SPIA as well as simply not withdrawing the full amount calculated whenever it's not needed.

I'm using my own version of an amortization method and I consider the calculated withdrawal to be a "max" and ultimately only withdraw what I need. By not withdrawing the full amount that I could, the extra still invested is available for all future withdrawals. Over time, I would expect the portfolio to grow, giving us additional margin in the future.

Cheers.

Sounds like you have done a lot of research on the various models and still you came up with your own method.
I feel like I am going down a rabbit hole with this stuff. Seems like there is a whole industry of retirement withdrawal planning and it is all based on trying to predict the unknown future.

 

[2HOTinPHX]

I have watched a few of this guys videos and he seems to make a lot of sense. He is low key and pretty clear in his ideas...I think. In this video he discusses the 4% rule and some drawbacks with it. Then he talks about "advanced" guardrail strategies. He notes that an 80% success rate in software programs does not mean a 20% failure rate but 20% chance you need to make adjustments. Wonder if this approach makes more sense?

 

[Exchme]

Sounds like you have done a lot of research on the various models and still you came up with your own method.
I feel like I am going down a rabbit hole with this stuff. Seems like there is a whole industry of retirement withdrawal planning and it is all based on trying to predict the unknown future.

Correct.

There is no escaping from the luck of the draw as to when you accumulated assets and when you retired. The "SWR" method of constant real $ withdrawals gives you an easy ride, unless you have an unlucky time period, then you run out of money! Some folks like to accumulate cash, but that just puts the SORR into the time you pre-sold to accumulate the cash and that may aggravate things since the cash didn't earn much return. With variable withdrawal methods, the bad SORR cases require fairly severe cutbacks for lengthy periods of time when the market is down. Tweaks to variable withdrawal methods to get more money early inherently increase risk of deep cutbacks later.

Plus all prescriptive methods are oversimplified because in real life, your cash flow is highly variable - helping kids or parents, claiming SS and/or pensions, receiving an inheritance, moving/remodeling, long term care, etc. all interrupt the best laid plans.

I'd say the SWR method gives you a general feel for retirement readiness, but as for managing over time, the best you can do is be prepared to be flexible and adjust to reality as you go.

 

[big-papa]

Sounds like you have done a lot of research on the various models and still you came up with your own method.
I feel like I am going down a rabbit hole with this stuff. Seems like there is a whole industry of retirement withdrawal planning and it is all based on trying to predict the unknown future.

I literally spent almost a decade looking at this stuff before I retired. I wouldn't say that I came up with my own method so much as I came up with my own spreadsheet to implement it and I allow for flexibility. Regarding flexibility what I do is somewhat analogous to someone who is subject to RMDs but the RMDs are more than they could possibly spend, so after the forced withdrawal, taxes are paid, and whatever is needed for spending is set aside and they reinvest the remainder. In my case, I never make the full withdrawal if I don't need it and it just stays invested to be used later or to end up as my kiddo's inheritance.

The closest thing to what I'm doing (outside of the flexibility part) is described in this two part series over on bogleheads, which also includes a sample spreadsheet. On top of amortization, it also includes time value of money concepts for future cash flows. For example, that could be a future SS/Pension stream or a future house sale with money left over or an inheritance or whatever. Planning for those sorts of things can help smooth out income over one's remaining lifetime.

Part 1: https://www.bogleheads.org/blog/2019/02 ... -of-money/
Part 2: https://www.bogleheads.org/blog/2019/02 ... ey-part-2/

At any rate, with regards to any withdrawal method you can come up with, there is no escaping Sequence of Returns risk when holding risky assets. You can't escape it, but you do have some say in how it's manifested. With an SWR method, the risk is total portfolio depletion before you're ready. With amortization, total portfolio depletion before you want it to happen is no longer there, but there is always the possibility that any single withdrawal is less than what you need to pay your bills. There are ways to mitigate that and if that failed, frankly, I'd rather be scrambling because some withdrawal in the middle of my retirement is too low than having a $0 balance remaining and otherwise being healthy.

And, yes, even with amortization there are some extreme cases one could envision where the portfolio itself isn't totally depleted but is so low as to be practically $0. Again, this can be at least partially mitigated in any number of ways before that ever happens.

Cheers.

 

[big-papa]

Correct.

There is no escaping from the luck of the draw as to when you accumulated assets and when you retired. The "SWR" method of constant real $ withdrawals gives you an easy ride, unless you have an unlucky time period, then you run out of money! Some folks like to accumulate cash, but that just puts the SORR into the time you pre-sold to accumulate the cash and that may aggravate things since the cash didn't earn much return. With variable withdrawal methods, the bad SORR cases require fairly severe cutbacks for lengthy periods of time when the market is down. Tweaks to variable withdrawal methods to get more money early inherently increase risk of deep cutbacks later.

Plus all prescriptive methods are oversimplified because in real life, your cash flow is highly variable - helping kids or parents, claiming SS and/or pensions, receiving an inheritance, moving/remodeling, long term care, etc. all interrupt the best laid plans.

I'd say the SWR method gives you a general feel for retirement readiness, but as for managing over time, the best you can do is be prepared to be flexible and adjust to reality as you go.

If I were to look at way to determine retirement readiness, SWR wouldn't be it. Nor would some multiple of expenses like 25X since 25X is a direct result of the 4% so-called rule (1/4% = 25) or any other % SWR.

The funded ratio always seemed to me like a better approach. One thing I personally like about it is it has a direct link to amortization based withdrawal methods. I personally believe that using long TIPS as the discount rate is overly conservative, but I understand why it came about. And, at any rate, it's a fairly easy calculation to perform with plenty of example spreadsheets available on the internet.

What I actually did for several years before I retired was to actually calculate my withdrawal and did a rough tax calculation after that. Once the withdrawals passed what I knew I wanted as after tax spending, along with some margin, I knew I was ready and gave notice. I was lucky-ish because I worked for another 15 months or so before actually leaving resulting in another 15 months of salary, 2 record corporate bonuses and a bunch of RSU's that vested when stock was at a record high which, right at the finish line, put us in the "more than enough" category.

You make a good point above what SORR can do. Amortization inherently relies on an estimate of future returns. VPW uses a fixed "rate" for this based on long term worldwide trends for stock returns and around a 2% fixed real return for bonds. If you use this, you'll find that the same worst case SORR that sets the worst case historical SWR resulted in a pretty deep dip in withdrawals, in real terms, through most of the 1970's before recovering.

In my case, my bonds are all in a TIPS ladder, so there's as close to 0 risk in future cash flows as there can be. I also include SS streams in my cash flow calculations (including adding a future haircut) so that leaves stocks. There is no perfect predictor of stocks, nor would you actually want that since it would probably remove any equity risk premium that might exist. For simplicity, I simply use 1/CAPE as a crude real returns estimate for stock and update it each withdrawal period. With my stock/TIPS split and including SS, the dip during the 1970's, if it were to happen today, would be greatly mitigated and tolerable.

Cheers.

 

[Chuckanut]

In this video we are shown how a balanced portfolio can be safer that an all stock or all bond portfolio. It looks good in the video presentation but he picked the year 2000 as a starting point. The all stock portfolio gets hit right away by sequence of return risk. Is this a fair comparison? Wouldn't they have done somewhat better if they had a recommended three to five years cash reserve? To bad the comments are turned off for the video. Is this video a fail or does it still make a good case for a balanced fund?

It’s time to reread the book How to Lie with Statics. One way to deceive people is to choose time periods that support the writer’s narrative. That’s why we have tools like FireCalc that look at the success of investments over a hundred periods of time. <- No guarantees, but much safer.

 

[2HOTinPHX]

It’s time to reread the book How to Lie with Statics. One way to deceive people is to choose time periods that support the writer’s narrative. That’s why we have tools like FireCalc that look at the success of investments over a hundred periods of time. <- No guarantees, but much safer.

Good point.....
Check out the other video I posted above. I think he represents a better overall view perhaps?

 

[Koolau]

I have watched a few of this guys videos and he seems to make a lot of sense. He is low key and pretty clear in his ideas...I think. In this video he discusses the 4% rule and some drawbacks with it. Then he talks about "advanced" guardrail strategies. He notes that an 80% success rate in software programs does not mean a 20% failure rate but 20% chance you need to make adjustments. Wonder if this approach makes more sense?

Yeah, I don't think any rational person (and that includes MOST of us here ) would fail to make adjustments to spending, etc. if we saw that "failure" was becoming more likely.

IIRC we've had one person claim "failure" on the forum.