Stress-testing the 4% rule: How do you handle worst-case timing?

[Roman]

Hi everyone,

I’ve been wondering how sensitive the classic 4% withdrawal rule is to unfavorable timing under more conservative assumptions (moderate nominal returns, ~2% inflation). By “4% rule” I mean 4% of the initial portfolio, adjusted for inflation each year.

The attached report was generated by a personal retirement planning project I’m developing to stress-test various withdrawal strategies (fictitious numbers): In this specific run, I modeled a ~5-year bond buffer; with an early -30% shock and a later -20% shock, the portfolio runs out after 22 years in this scenario.

How do you account for worst-case timing in your planning? Is probability enough for your "sleep factor", or do you prefer guardrails/more certainty?

Best regards
Roman

 

[Koolau]

I can't say I ever worried too much about it - and I retired a couple of years before the Great Recession.

I was spending well beyond 4% for about 3 years due to 2 rehabs.

 

[keppelbay]

Consider an approach that is not based on selling assets to fund current spending.

(1) Determine how much you (want to) spend. Double (or if you like, triple) that figure to get your cash-flow target.
(2) invest in a mix of assets that provides the requisite cash-flow.
(3) roll the excess income into growing future cash-flow (or into growth assets).

Example: spend 5K/month = 10K target.
PDI 38K shares @ monthly distribution 0.2205 = 8379/mo
PIMIX 30K shares @ monthly distribution 0.055 = 1650/mo
cash flow = ~10K/mo on ~1M invested.
Then do whatever you like with the rest of your assets, and never have to sell anything, unless you want to.

Notes:
- This approach separates funding needs from asset values.

- Obviously, there is no need to go all in on an approach like this. But it can be used to reduce dependence on price performance elsewhere in the portfolio.

- I use the PIMCO funds as an example, because they have been reliably stable in their distribution rates over a changing interest rate landscape. Any other combination of higher-yielding and safer lower-yielding assets can be used to reach a desired cash-flow target (including your pension).

Risks:
- Distribution cuts.
Mitigation: diversify income sources.

- Asset price erosion in CEFs.
Example: PDI NAV is down ~$8.4/share since inception (2012). In that time, the fund has distributed more than $44/share, on an intital cost of $25. Rolling excess cash-flow into new shares mitigates the capital erosion, while increasing future cash-flow. (You can view this like an annuity, but one where the cash-flow is available to your heirs.)

 

[Dtail]

I know it is theoretical, but having a annual stock return of 5.5%, while also having 2 bear market scenarios is not typical at all.
Look at the 1999 retiree. They had 2 crashes in the first 9 years and (so far) have survived the 4% guidance.

 

[marko]

(Involuntarily) Retired in 2005. Severance ran out in January 2008. Ran headlong into the GR.

Never skipped a beat. Portfolio has more than doubled since, despite an average of 6% withdrawals back then. Had no idea what I was doing but thrived in spite of it.

 

[FLSUnFIRE]

The "4%" rule IS based on worst-case timing (looking back using historical data) with most retirement years yielding huge ending balances. There is always the possibility that a worse series of returns will occur but the "4% rule" was never designed to predict the future.

If so inclined, I'd run some Monte-Carlo simulations with pessimistic assumptions if you want to justify a lower WDR. Me, I'm conservative and youngish so try to stay under ~3% and half my expenses are discretionary if I need to adjust.

 

[mrfeh]

The "4%" rule IS based on worst-case timing (looking back using historical data) with most retirement years yielding huge ending balances. There is always the possibility that a worse series of returns will occur but the "4% rule" was never designed to predict the future.

+1

If you believe the future won't differ materially from the past, there is no need to concoct a new worst case scenario.

 

[Aerides]

Having a few years expenses in cash is a great way to avoid worst-case-timing.

 

[Flyfish1]

Consider an approach that is not based on selling assets to fund current spending.

(1) Determine how much you (want to) spend. Double (or if you like, triple) that figure to get your cash-flow target.
(2) invest in a mix of assets that provides the requisite cash-flow.
(3) roll the excess income into growing future cash-flow (or into growth assets).

Example: spend 5K/month = 10K target.
PDI 38K shares @ monthly distribution 0.2205 = 8379/mo
PIMIX 30K shares @ monthly distribution 0.055 = 1650/mo
cash flow = ~10K/mo on ~1M invested.
Then do whatever you like with the rest of your assets, and never have to sell anything, unless you want to.

Notes:
- This approach separates funding needs from asset values.

- Obviously, there is no need to go all in on an approach like this. But it can be used to reduce dependence on price performance elsewhere in the portfolio.

- I use the PIMCO funds as an example, because they have been reliably stable in their distribution rates over a changing interest rate landscape. Any other combination of higher-yielding and safer lower-yielding assets can be used to reach a desired cash-flow target (including your pension).

Risks:
- Distribution cuts.
Mitigation: diversify income sources.

- Asset price erosion in CEFs.
Example: PDI NAV is down ~$8.4/share since inception (2012). In that time, the fund has distributed more than $44/share, on an intital cost of $25. Rolling excess cash-flow into new shares mitigates the capital erosion, while increasing future cash-flow. (You can view this like an annuity, but one where the cash-flow is available to your heirs.)

This is the approach I took. Income investing can mitigate the risks and avoid the need for holding three years or so of cash. Just a different approach that's all.

 

[The Cosmic Avenger]

We plan on spending less if there is a recession during our early retirement years. We really only have to make it to 62, and we could always start SS early if there was a deep and prolonged recession, at least for one of us, to minimize the impact on our portfolio.

 

[mrfeh]

This is the approach I took. Income investing can mitigate the risks and avoid the need for holding three years or so of cash. Just a different approach that's all.

There's no reason to avoid selling assets.

Total return is all that matters.

 

[RunningBum]

I don't think the answer is to try to make a 30+ year plan that will handle all (reasonably) possible situations unless you want a lot of buffer. In other words more like a 2% WR than 4%. Wasn't there was a recent thread about how to identify and adjust a plan that is going bad?

 

[Steelpony]

Hi everyone,

I’ve been wondering how sensitive the classic 4% withdrawal rule is to unfavorable timing under more conservative assumptions (moderate nominal returns, ~2% inflation). By “4% rule” I mean 4% of the initial portfolio, adjusted for inflation each year.

The attached report was generated by a personal retirement planning project I’m developing to stress-test various withdrawal strategies (fictitious numbers): In this specific run, I modeled a ~5-year bond buffer; with an early -30% shock and a later -20% shock, the portfolio runs out after 22 years in this scenario.

How do you account for worst-case timing in your planning? Is probability enough for your "sleep factor", or do you prefer guardrails/more certainty?

Best regards
Roman

The easy answer which is always my quick unsubstantiated opinion.

If you’re going to depend on an unknown future value portfolio construction a static “rule” that’s only tied to general inflation and adjusted up each year should in addition be adjusted down sometimes when Mr. Market doesn’t cooperate. So maybe only adjusting to your personal inflation rate as a variable not a static 4% might be better. Forget studies if you’re going to spend down in retirement. It’s not personal enough. You have to deal with your own unknowns.

You seem to like the quest for the holy grail dealing with endless market variables. Some posters including me solved retirement in much different and simpler ways.

 

[Aerides]

We plan on spending less if there is a recession during our early retirement years. We really only have to make it to 62, and we could always start SS early if there was a deep and prolonged recession, at least for one of us, to minimize the impact on our portfolio.

That's the thing though - I won't want to (spend less). Recession? I still want to travel, get my next car if it's time, all that. I wouldn't want to have to cut back and budget less when I'm still in my go-go years. Cash lets me ride it out and keep living as I do.

 

[audreyh1]

I never tried to stress test the 4% rule because that is already stress tested on historical data. I use a different method anyway.

 

[audreyh1]

That's the thing though - I won't want to (spend less). Recession? I still want to travel, get my next car if it's time, all that. I wouldn't want to have to cut back and budget less when I'm still in my go-go years. Cash lets me ride it out and keep living as I do.

Same here, and even through 2000-2002, 2008+, 2022, etc. we’ve never had to cut back as we had adequate cash to cover short term spending. We did cut back spending involuntarily in 2020, but that’s ‘cause you couldn’t do anything!

 

[Dtail]

Same here, and even through 2000-2002, 2008+, 2022, etc. we’ve never had to cut back as we had adequate cash to cover short term spending. We did cut back spending involuntarily in 2020, but that’s ‘cause you couldn’t do anything!

Yes and it works for you knowing you are at 3.3% and how you use the unspent cash reserves.
So theoretically for you, if you spent all your tested 4.35% withdrawal using your remaining portfolio methodology and we are in a bear market currently, would you still spend throughout the year all the funds you withdrew in Jan of that year?
It is just theoretical and I know you wouldn't put yourself in that scenario.

 

[SJhawkins]

That's the thing though - I won't want to (spend less). Recession? I still want to travel, get my next car if it's time, all that. I wouldn't want to have to cut back and budget less when I'm still in my go-go years. Cash lets me ride it out and keep living as I do.

I think this is where a TIPs ladder fits in. Never going to out guess what the market is going to give you.

Starting with a 10 year ladder seemes sane, if lucky to get a few good years in the market you roll a few rungs and you may get a good 12+ years out of it. Obviously this has to fit into one's plan, total AA etc.

 

[FREE866]

There's no reason to avoid selling assets.

Total return is all that matters.

Agree. I think there is such an aversion to selling assets and spending down principal. I also think people get caught up in "income producing assets" which aren't as tax efficient as liquidating equities and provide a much greater total return, especially over decades.

 

[RetiredHappy]

I think this is where a TIPs ladder fits in. Never going to out guess what the market is going to give you.

Starting with a 10 year ladder seemes sane, if lucky to get a few good years in the market you roll a few rungs and you may get a good 12+ years out of it. Obviously this has to fit into one's plan, total AA etc.

I like MYGAs better.

 

[SJhawkins]

I like MYGAs better.

Absolutely, many roads to Dublin as they say.

 

[USGrant1962]

Big Ern at earlyretirementnow.com has done a truly massive amount of work around stress testing SWR strategies, guardrails, and the like. I suggest OP take a look there for tools and analysis.

The Safe Withdrawal Rate Series - Early Retirement Now

Discover the Safe Withdrawal Rate Series: 4% Rule, Personalized strategies, sequence risk, FIRE planning, free Google Sheet, 2025 updates.

earlyretirementnow.com

 

[SnowballCamper]

OP, you should probably read one of the original sources, like this one https://kyestates.com/wp-content/uploads/2015/02/Bengen1.pdf
If you're not using historical data, the simulation is just an exercise in tweaking parameters to get results that you like. That can be useful to understand the process, but IMHO doesn't give reliable answers.

 

[pb4uski]

A 30% decline followed shortly thereafter by a 20% decline and then otherwise normal returns is unnecessarily pessimistic. Has never happened. Typically, large downturns are subsequently followed by better than long-term average returns. Besides that, when in the history of man had there ever been a 30% downturn shortly followed by a 20% downturn?

If there was a scenario like that then I suspect that thee would be bargains galore, likely deflation rather than inflation built into your assumptions so your withdrawals would decline and extend the 22 year drawdown.

Keep working.

 

[engineernerd]

Hi everyone,

I’ve been wondering how sensitive the classic 4% withdrawal rule is to unfavorable timing under more conservative assumptions (moderate nominal returns, ~2% inflation). By “4% rule” I mean 4% of the initial portfolio, adjusted for inflation each year.

The attached report was generated by a personal retirement planning project I’m developing to stress-test various withdrawal strategies (fictitious numbers): In this specific run, I modeled a ~5-year bond buffer; with an early -30% shock and a later -20% shock, the portfolio runs out after 22 years in this scenario.

How do you account for worst-case timing in your planning? Is probability enough for your "sleep factor", or do you prefer guardrails/more certainty?

Best regards
Roman

First of all, are you sure 2% inflation is a conservative assumption? I personally would disagree.

I have placed my faith in the historical approach embedded in FIRECalc and FI Calc. I see that you are not in the USA. Theoretically I guess you could buy US investments but then of course currency risk, and (I hate that this is true) political risk, would affect your results.

Are historical data regarding inflation and/or investment returns relevant to you available?