Is 3% the new 4% SWR?

[Midpack]

Just another article exploring the merits of the 4% SWR for the future, from a good resource IMO. Retirement Researcher Blog: Retirement Lengths, Withdrawal Rates, and Failure Probabilities

Not debating, just sharing FWIW.

In summary:
Case A based on past history, inflation adjusted 8.6% on S&P and 2.6% on bonds.
Case B based on inflation adjusted 5.5% on stocks and 1.75% on bonds.

Assumption:
40% stock AA, 95% probability of success

Plan|Case A|Case B
30 yrs|4.0%|3.3%
40 yrs|3.0%|2.3%

 

[pb4uski]

As a separate but related issue, I think it is important to focus on your "ultimate" WR rate once SS and pensions are on-line in addition to your initial WR at the time you ER.

In my case, my initial ER WR is much higher than my estimated ultimate WR.

 

[Which Roger]

SWR depends not just on (assumed future) rates of return, but also on (assumed future) volatility. If an asset class is expected to give lower returns than in the past, that implies that it is perceived by investors as being less risky than it was in the past. So for the model in the article to be valid, it should assume less volatility going forward along with the lower returns. The author does not mention whether or not that was done.

This is important because the biggest contributor to portfolio failure is a very bad sequence of returns early in the withdrawal phase. The less volatility, the lower the probability of such a sequence.

 

[GrayHare]

IMO a significant portion of recently shrinking SWRs is due to recency bias.

 

[David1961]

Thanks for sharing. It'd be interesting to see charts for a 50% and 60% stock AA.

 

[mickeyd]

2%, 3%, 4% what the heck! We (if we are as smart as we seem) are all going to adapt our WR to the curremt situation and make the appropriate decsiion based on that.

My 60/40 AA is looking pretty sweet today. Who the hell knows about next week?

 

[Chuckanut]

As a separate but related issue, I think it is important to focus on your "ultimate" WR rate once SS and pensions are on-line in addition to your initial WR at the time you ER.

In my case, my initial ER WR is much higher than my estimated ultimate WR.

+1

 

[siamond]

IMO a significant portion of recently shrinking SWRs is due to recency bias.

+1

 

[Midpack]

Corrected table:

Plan|Case A|Case B
30 yrs|4.0%|3.0%
40 yrs|3.3%|2.3%

 

[photoguy]

Anybody actually use the software that Pfau mentions (MoneyGuidePro)? Does Pfau any relationship with them?

 

[marks]

As a separate but related issue, I think it is important to focus on your "ultimate" WR rate once SS and pensions are on-line in addition to your initial WR at the time you ER.

In my case, my initial ER WR is much higher than my estimated ultimate WR.

To expand on pb4uski's comment, since this is an early retirement site, most people here will have to deal with two-stages of retirement, i.e., most early retirees will have Social Security and/or pensions and possibly other income coming on line in 5-15 years.

There is a lot of focus on a single SWR, when in reality there are two. I have never really seen a methodology to address this situation.

For example, if a couple is targeting $100,000 in income, they can “burn” some capital in the first-stage of retirement in order to meet that need, as long as they preserve enough capital when they hit SS age. This obviously ignores inflation and volatility, but at least it leaves you with an interim goal so that you can adjust spending to hopefully stay somewhat close to the goal at SS age, at which point your WR should be substantially less.

Here is a hypothetical example:

Retirement age = 55 (married couple)
Assets at retirement = $2,500,000
WR @ 4.0% = $100,000 (for 10 years)
Total Income = $100,000

"Burn" $50,000 per year for 10 years

Assets at 65 = $2,000,000
SS Income = $50,000
WR @ 2.5% = $50,000
Total Income = $100,000

You can obviously adjust the initial WR as you go to try to hit the asset target at SS age, as well as the timing of starting to receive SS.

I am sure this seems obvious, but I have not really read anything about this approach. Any comments would be appreciated.

 

[pb4uski]

What I would do in such a situation is subtract the amount needed to fund the ER years from the age 55 assets and then calculate the WR. So assuming 0% real return, $1m on the $2.5m age 55 nestegg would be needed to fund ages 55-64 and beginning at age 65 the withdrawals would be $50k from $1.5m, or a 3.33% WR.

Alternatively, using the 4% "rule" the couple could withdraw $107k a year beginning at age 55 and $57k a year beginning at age 65. IOW, you are splitting the ER age nestegg into two parts, one to fund 55-64 and the other for 65+.

Something like that anyway.

 

[Ronnieboy]

2%, 3%, 4% what the heck! We (if we are as smart as we seem) are all going to adapt our WR to the curremt situation and make the appropriate decsiion based on that.

My 60/40 AA is looking pretty sweet today. Who the hell knows about next week?

+1
This is what I have witnessed reading these forums. If someone was following the SWR, then they take out a set percentage each year plus inflation. This is to happen no matter what the stock/bond market does, it is consistent with the only variable being the amount of inflation for the year.

It seemed (no official numbers) that emotions got in the way for 95% of the posters and they mentioned cutting their withdrawals in 2008/09 downturn, just in case.

I am not retired, so it is easy for me to point this out, but I think I would have done the same thing. 'This time it might be different.'

It seems to boil down to: you adapt to the current situation and comfort level in the moment, when you determine your withdrawal rate.