Asset allocation for very early retirement

[accountingsucks]

I have thought about alot of different scenarios as to when I would like to "retire". I think what I've settled on is a compromise being that I'd like to get to a portfolio that supports a 4%SWR AND continue working in a part time or consulting capacity such that I do not have to draw on my portfolio until I'm in my 50's. My biggest beef with working is the inflexibility in hours and/or vacation time. I should be able to achieve this between age 40 and 42 with a portfolio value of about 850K. Currently I am 34 and am about 1/3rd of the way there + my home is paid off.

My question is, given these facts what should my asset allocation be today, at 40 and 50 years old? Should I be more aggressive with my equity allocation or stick to the "age in bonds" allocation that I'm currently trying to get too? Or perhaps I need to be more conservative? At the moment I am about 70% equities, 5% bonds and 25% cash as I try to decipher what the markets are doing.

Does anyone have any book recommendations specifically dealing with asset allocations?

 

[Bikerdude]

Does anyone have any book recommendations specifically dealing with asset allocations?

"The Intelligent Asset Allocator" by William Bernstein

 

[ziggy29]

"The Intelligent Asset Allocator" by William Bernstein

In addition to this one, I'd add "All About Asset Allocation" by Rick Ferri, who is also a contributor on the Bogleheads board.

 

[Rustic23]

Out of curiosity are you taking inflation into consideration? Using 4% of 850,000 you come up with $34,000. If you can live on that today, it will take $46,600 in 8 years to equal the same standard of living, using a 4% inflation rate. Some suspect that will be low over the next 8 years.

 

[Carnage]

Why the 25% in cash? Is this also a rainy day portion of your fund?

 

[utrecht]

Stop trying to decipher what the market is doing.

 

[REWahoo]

Stop trying to decipher what the market is doing.

Yep.

By the time you do it will have already done it. Then you'll have to try to decipher what it's doing next, etc.

The purpose of an asset allocation is to allow us to not seek divine guidance to foresee the future - but let me know if you get a hot tip.

 

[accountingsucks]

Out of curiosity are you taking inflation into consideration? Using 4% of 850,000 you come up with $34,000. If you can live on that today, it will take $46,600 in 8 years to equal the same standard of living, using a 4% inflation rate. Some suspect that will be low over the next 8 years.

I am taking this into consideration. My base expenses are 19K per year, but when I add non recurring car/house/vacation expenses + abit of taxes it gets me to 28K today which would creep up to the 35K-ish I think I would need in about 7 years.

 

[ERD50]

Stop trying to decipher what the market is doing.

Yep.

By the time you do it will have already done it. Then you'll have to try to decipher what it's doing next, etc.

Reminds me of a paper passed around the office (Harvard Business Review, IIRC) - The title was a take off on a Wayne Gresky quote:

A great hockey player plays where the puck is going to be.

Only a few are able to do that. The rest of us Buy & Hold an index fund


-ERD50

 

[DblDoc]

Parts of this paper address the OP's question. http://www.bobsfinancialwebsite.com/pdfs/1996_Bengen_15388_1.pdf

In a nutshell yes, you do need to consider a different equity:bond allocation or reduce your SWR if your retirement time frame is a) longer than 30 years and/or b) you hold a significant amount of your retirement funds in taxable accounts.

DD

 

[Life_is_Good]

I'm an early retiree - was 44 when I left the work force. I started with an asset allocation of about 60/40 to 65/35. I ran with that allocation for my first 5 years including the 2008 crash. On my 5 yr anniversary, I reevaluated and decided I wanted to be a little more conservative. I moved towards 45/55 and positioned money from higher cost (employer 401k) to lower cost funds (VG). I'm sleeping much better.

 

[Rustic23]

I like to read Scott Burns column. He definitely believes in asset allocation, and with all the talk about 'Buy and Hold' being dead, he appears to disagree.

What Investors Wish Most - Registered Investment Advisor

 

[MasterBlaster]

Longer retirements (per Bernstein) require

1) a lower SWR
2) a higher allocation to equities

For a stock-bond portfolio, Bernstein gives an allocation as seen in the chart. Notice for really long duration retirements that the allocation is almost all equities.

 

[FUEGO]

I'll echo the comments from other posters here. Longer retirement period equals more equities.

Since you also plan on drawing an income sufficient to live on and not withdraw from your portfolio for a decade or two, I would say you could support a very high percentage allocation to equities without risking your ER-survivability. As long as you can stomach the risk/volatility of a high equities portfolio.

I plan to ER in my late 30's most likely, and will probably shoot for an equity allocation around the 80%+ range. Maybe some part time work initially, or return to work if the portfolio takes a massive (2008-ish) nosedive in the first 5-10 years of ER.

 

[FIREd]

For a very early retirement, I'd rather focus on keeping my SWR low (3% or less) and use a variable withdrawal rate (no automatic inflation adjustment, expense cuts during bear markets, etc...) than increase my equity allocation beyond 50-60%. But that's only because controlling my expenses feels more attainable than controlling my portfolio's future return.

 

[MasterBlaster]

If I may point out the obvious and the not-so obvious:

Notice from the stock allocation chart that it is nearly linear from retirements from 30 to 60 years. Note also that the slope of this graph over the mentioned range is around 10% less stock allocation over those 30 years (ie. ~85 % at 60 years less ~75% at 30 years).

Observation #1
That's quite a bit less slope than the standard age based "rules of thumb" for age based asset allocation. The age based asset allocation goes something like bond_percentage = 100 -age (or 110 or 120 -age). The Bernstein chart really differs with the age based rules of thumb for long retirements.

Observation #2
For retirements of 20 years and less the age based asset allocation formulas aren't steep enough. For retirements of 20 to 10 years duration the Bernstein chart slope is about 1.5 times the age based rule of thumb.

Observation #3
Therefore the age based asset allocation models just may lead a very early retiree to have too little equities and for older retirees to have too much.

Of course maybe Bernstein is full of beans - things happen, Every market and everyone is different, your milage may vary.

 

[ERD50]

I'll echo the comments from other posters here. Longer retirement period equals more equities.

I was curious just how much of an effect this is. I went to Firecalc, found that a 25 year portfolio provides 95.7% success @ 4.2793% WR; and a 45 year portfolio provides 95.6% success (the closest I could come) @ 3.7866% WR.

The 45 year portfolio seems to 'like' an AA of 75-100% Equities. A 25 year portfolio seems to 'like' 35-75% Equities. Historically, of course - and maybe flawed a bit as there are 20 more data points available for the 25 year scenario.

edit/add - seems similar to what MB just posted, and I'll also add that I used the 'investigate' tab in FC in two steps. One, to get the spend level for a 95% success rate, then entered that and then clicked the "Investigate changing my allocation" for the attached graphs - in case anyone wants to duplicate that for their scenario. Hint - use $1,000,000 for the portfolio, and a decimal point shift is all you need for % SWR.

-ERD50

LEFT - 25 Year Portfolio ---------------------------------- RIGHT - 45 Year Portfolio

 

[FUEGO]

For a very early retirement, I'd rather focus on keeping my SWR low (3% or less) and use a variable withdrawal rate (no automatic inflation adjustment, expense cuts during bear markets, etc...) than increase my equity allocation beyond 50-60%. But that's only because controlling my expenses feels more attainable than controlling my portfolio's future return.

I'll add that this is in my plans too (to maybe a lesser degree than FD).

Roughly 1/2 my withdrawal will be fixed % with inflation adjustment, and the other half tied to a percent of portfolio value (like 1.7% of portfolio + CPI, and 1.7% of annual portfolio value). I call it the "hybrid approach". It should dampen withdrawal volatility year to year, and allow us to tighten up during really bad years while enjoying increases in good years.

 

[DblDoc]

I was curious just how much of an effect this is. I went to Firecalc, found that a 25 year portfolio provides 95.7% success @ 4.2793% WR; and a 45 year portfolio provides 95.6% success (the closest I could come) @ 3.7866% WR.

The 45 year portfolio seems to 'like' an AA of 75-100% Equities. A 25 year portfolio seems to 'like' 35-75% Equities. Historically, of course - and maybe flawed a bit as there are 20 more data points available for the 25 year scenario.

edit/add - seems similar to what MB just posted, and I'll also add that I used the 'investigate' tab in FC in two steps. One, to get the spend level for a 95% success rate, then entered that and then clicked the "Investigate changing my allocation" for the attached graphs - in case anyone wants to duplicate that for their scenario. Hint - use $1,000,000 for the portfolio, and a decimal point shift is all you need for % SWR.

-ERD50

LEFT - 25 Year Portfolio ---------------------------------- RIGHT - 45 Year Portfolio

Similar findings in the Bengen paper I cited above. The bottom line is for longer withdrawal periods or large proportion in taxable accounts you need to either decrease the WR, increase allocation to equities or some combination of both.

DD

 

[jIMOh]

There was a study linked here within last 6 months which showed lowest withdraw rate (3.3%?) had same likelihood of surviving regardless of allocation (100% equity-75/25, 50/50, 25/75) for 40 year withdraw periods (all had >95% success rates).

So for long periods, focus more on keeping withdraw rate down, the allocation has less to do with performance or success (longevity) than the withdraw rate.

 

[FUEGO]

Similar findings in the Bengen paper I cited above. The bottom line is for longer withdrawal periods or large proportion in taxable accounts you need to either decrease the WR, increase allocation to equities or some combination of both.

Regarding the large proportion being in taxable accounts dictating that you need a higher allocation to equities - is this an issue of tax efficiency? The thought being that you cannot afford to pay taxes every year on a large taxable bond component and it would be much cheaper tax-wise to let the cap gains accrue year after year until you sell (and get favorable cap gains tax rates)?

 

[DblDoc]

Regarding the large proportion being in taxable accounts dictating that you need a higher allocation to equities - is this an issue of tax efficiency? The thought being that you cannot afford to pay taxes every year on a large taxable bond component and it would be much cheaper tax-wise to let the cap gains accrue year after year until you sell (and get favorable cap gains tax rates)?

Bengen assumes you are invested in a tax efficient manner in your taxable accounts. He freely admits that this is a gray area as you cannot predict future tax rates/rules and therefore takes a conservative approach. He seems to be more concerned about the risks of inflation long term than equity market risk - this was written in 1996, wonder how he feels now post 2008?

DD