Asset allocation in retirement

[walkinwood]

Rick Ferri makes a case for using a 30/70 (equity/bond) allocation in retirment instead of 60/40.

The Center of Gravity for Retirees

I would like to hear his opinion on allocation for longer retirement periods like most of us who ER have. He may participate in a discussion on this at Bogleheads. (I'll add the link if I find a discussion. Please add if you see it first).

[BBQ-Nut]

That's about what I used before Fire, and I don't see on changing it in the near future now that I recently Fire'd.

[Trix55]

Looks to me, he is recommending 70% bond and 30% equity.

Sent from my QMV7B using Early Retirement Forum mobile app

[samclem]

Ferri is arguing for 70% bonds and 30% equities.

Quote:

I believe the center of gravity shifts when a person stops accumulating assets and starts taking income from their assets. The corrected ideal asset allocation for beginning a discussion on asset allocation with a pre-retiree or retiree is 30% stocks and 70% bonds.

It might be good for some folks, but it would not work for us. I think he is confusing/conflating "risK" with "volatility". Some of the most important risks a retiree's account faces have nothing to do with volatility (e.g. failing to keep up with inflation). Reducing volatility can increase some very important types of risk, and I think that's what he is doing.

[walkinwood]

Trix55 - you're right. Thank you
I mistyped. Corrected now.

[walkinwood]

SamClem - can you elaborate on why it isn't right for you?

We use 60/40 (equity/bond). My reasoning is that this AA is more appropriate for our long retirement period. I am 54 now. DW is 51

[Meadbh]

Most of these asset allocation discussions assume that one's entire portfolio is in either equities or bonds. I have some income producing real estate and I think of that as behaving like fixed income. So while my equity to bond ratio is ~60:40, the contribution of equities to my investment portfolio as a whole is only about 40%.

[Dash man]

The author is proposing 70/30 bond/equity position which many will believe is conservative. If using individual bonds it might work. I'd be worried about bond funds with the low interest rate environment. If/when rates do begin to rise, there may be a liquidity crisis with bond funds with many investors rushing to cash out of the funds. So, to me this would be risky. I'm personally using 2-3% CDs in place of bonds in my portfolio with about 60% equity / 20% CDs / 15% real estate / 5% cash. Our equity positions will be downsized shortly when DW exercises some options within the next couple of months, then we'll add to CDs and cash until we have 50% equities.

[flyingaway]

I would think that there will be not much difference between 70/30 or 60/40 or 50/50 (30/70 is too much bond for me) asset allocations in the long run. I will just choose one and be done with it. The key point is to not time the market, i.e., change allocations abruptly based on short market behaviors or rumors.

[scrabbler1]

I have two AAs, one for my taxable accounts which generate the income to cover my expenses, and another for my IRA I don't plan to touch until at least age ~59.5 which is in 8 years from now. For my taxable account it is 65/35 in favor of bonds. For the IRA I once had 55/45 in favor of stocks but I have since changed to 50/50 after I turned 50 in 2013.

[Brooks Saddle]

Rick Ferri is one that I place a lot of confidence in. I read some work that Jim Otar did almost a decade ago that states essentially the same thing. Based upon that writing by Jim Otar coupled with my dreaded fear of sequence risk I scaled back my stock percentage much further than I ever dreamed I would.

[samclem]

Quote:

Originally Posted by walkinwood SamClem - can you elaborate on why it isn't right for you?

1) We can tolerate a lot of volatility in our portfiolio (due to a DB retirement check that meets our "must have" spending)
2) For longer anticipated retirements (30-40 years), higher stock allocations are called for if we stick with WR of 3-4% and want 95% plus success rates.
The below chart from an article by Wade Pfau is applicable. It assumes constant annual inflation-adjusted withdrawals, and since we will be using "% of year-end value" withdrawals, stock allocations near the top of the suggested "window" are appropriate (because withdrawals will be less when portfolio values have taken a hit > portfolio value variability is not as much of a concern )

Anyway, that's my reasoning. His 30% stock recommendation is probably okay for many retirees with a 20 year expected horizon, dependent on constant returns, and taking a constant-dollar annual withdrawal.

Variability does not equal risk.

[Cobra9777]

Excluding real estate and cash, our investment portfolio is 60/40. However, we have 70% of expenses covered by two pensions and rental income, with SS still to come. If I count the PV of our pensions as a bond-equivalent, and include the real estate in equity, we are 50/50. If anything, I'd like to increase the equity a bit. As a 53 year-old retiree, I would not be comfortable at 30/70, especially with one of the two pensions non-COLA.

[audreyh1]

Kitces et. al showed that starting retirement with 30% stocks but thereafter increasing the stock % gradually over decades until it reaches 60% or 70% can improve portfolio survival, especially in the first 15 years when you mostly spend down the fixed income and let the stocks ride. But leaving it at 30% is not as good for portfolio survival in face of inflation, as a 50/50 approach +/- 10%. This Kitces approach is called the rising glide path.

[Totoro]

Quote:

Originally Posted by audreyh1 Kitces et. al showed that starting retirement with 30% stocks but thereafter increasing the stock % gradually over decades until it reaches 60% or 70% can improve portfolio survival, especially in the first 15 years when you mostly spend down the fixed income and let the stocks ride.

I am taking this exact approach for my mother, also because of some other psychological benefits:

• A big drop in stocks in the beginning isn't a big deal in absolute terms
• A big gain in the beginning means that if there is a drop afterwards the losses are offset somewhat by the initial gain - so you 'dont lose' just go back to square one
• As time goes on the uncertainty in longevity and spending decreases, so you can increase the uncertainty in future returns
• We get used to only buying on a regular basis, and never selling. Basically make the process so boring that the temptation to act and time the market is strongly reduced

It sort of resembles dollar cost averaging starting at your peak uncertainty and risk moment, with is your retirement date.


Idea is to end up at 30/70 (now at 20% or so equities) within a year or two and then wait for a market correction to move to 50/50 somewhere in the next ten years after that. After that leave it alone. Living expenses are taken from the fixed income portion for the foreseeable future.


Note that I don't want the maximum return possible, but a best possible "sleep well" factor for my mother while looking out for her best finanicial interests in the long term.

[ejman]

Quote:

Originally Posted by audreyh1 Kitces et. al showed that starting retirement with 30% stocks but thereafter increasing the stock % gradually over decades until it reaches 60% or 70% can improve portfolio survival, especially in the first 15 years when you mostly spend down the fixed income and let the stocks ride. But leaving it at 30% is not as good for portfolio survival in face of inflation, as a 50/50 approach +/- 10%. This Kitces approach is called the rising glide path.

I've kept my AA at a nominal 50/50 (10% band) since ER'd 12 years ago (currently 64). Although re balancing was an infrequent occurrence with the wide band I use I'm contemplating letting equities "ride" since the equity funds are primarily in my taxable account and I don't want to create taxable events as my rebalancing seems to be usually in the direction of selling equities to buy bonds.

If historical patterns hold the equities % would naturally increase over the years but the research you mentioned gives a measure of comfort in allowing this to happen.

[clifp]

I was looking at AA for my 66 year old sister and her recently retired husband. Their health is decent but not great and I think living another 30 years is unlikely. About 1/2 of their 401K is in cash .

A 4.1% withdrawal rate (what the need/want) has only a 93% success rate with the default 75/25 AA over 30 year and only increases 96.6% over 25 years.

You move the AA down to 35/65 (aka Wellesley AA) and the success moves down to 86% over 30 years but up to 100% over 25 years. The could even move the SWR up to 4.2% and still be 100% ok..

As an aside has anyone looked at Wellesley/Wellington safe withdrawal rates? I am pushing them since they already bought a variable annuity (ugh) to just adopt a W/W portfolio..

[pb4uski]

Firecalc suggests differently. $1m portfolio, $40k spending (4% WR), 30 year time horizon all default assumptions except for AA.

30/70 stock/fixed success rate = 86.8%
60/40 stock/fixed success rate = 95.6%

Extend that time horizon to 45 years for ER, and the success rates become 40.4% and 77.8%, respectively.

[clifp]

I went back and looked at Rick's article. I was disappointed that no mention was made of the retirement length.

Using Firecalc investigate feature
25 years. 95% success rate max SWR
30/70 4.25%
60/40 4.32%

30 years.
30/70 3.83%
60/40 4.02%

35 years
30/70 3.53%
60/40 3.82%

At 25 years almost no difference and you could certainly make a strong case that lower volatility is more important than $700 a year in more spending. At 35 year you need 8% bigger portfolio to achieve the same spending level.

[photoguy]

I can't see Ferri advocating 30s/70b for an early retiree. He mentions in the article that such a portfolio might do 2% better than bonds return alone. Right now it's a struggle to even keep up with inflation, so basically you are looking at a steady drawdown over time.

I think seeing a steadily dwindling portfolio, with little chance of recovery, would make me extremely nervous with my expected lifespan.

If you look at 30 year firecalc runs, a huge number of them trend down toward zero. It isn't looking good for extending it another 20 years or so.