New Take On AA: U-Shaped Path

[gerntz]

"The report finds that those who take the opposite approach—by reducing equity exposure right after retirement and then gradually raising it over time—are likely to make their money last longer."

Who'd have thought this?

"According to the research, those who start retirement with 20% to 30% in stocks and end up with 50% to 70% in stocks can withdraw 4% of their portfolio per year and give themselves annual raises to compensate for inflation over 30 years, even in the worst market scenarios. (The authors examined 10,000 simulations and assumed average annual returns of 6.5% for stocks and 2.4% for bonds.)

In contrast, those who keep 60% in stocks throughout retirement or who taper to a 30% equity allocation from 60% are likely to run out of money after 28 years in the 5% of worst-case scenarios"

A New Asset-Allocation Strategy for Investing in Retirement - WSJ.com

 

[walkinwood]

It first seemed unintuitive to me, but I think it makes sense.

The failure tracks are most often those where the withdrawal rate (as a percentage of the portfolio) rises to unsustainable levels early in retirement. One you're in that situation, making a bigger bet on stocks (ie. increasing your equity allocation) may be the only way to dig yourself out of the hole (assuming constant real spending).

I plan to keep my ER AA steady.

 

[Midpack]

First time I saw that approach discussed was in Kotlikoff & Burns book in 2008, and I doubt they invented the idea. It is intriguing...

 

[Singletrack]

I also read that in WSJ this morning. They did some pretty extensive simulation testing, and the results are indeed intriguing. If nothing else, a lower equity allocation at or near retirement can dramatically improve the sleep-at-night factor. Imagine retiring in early 2008 with 60-70% equities....I bet many of those folks had some market-induced insomnia.

I'll be anxious to hear others' thoughts on this.

[Edit....of course, very few assets escaped the 2008-2009 meltdown unscathed....]

 

[Ready]

It certainly is an interesting approach, and one that may give comfort to the newly retired, knowing that their exposure to stocks is dramatically reduced right at the beginning of their retirement phase. If a bear market occurs at the beginning, I guess things work out perfectly.

But if the market does great for a number of years after retirement, and then a bear market comes just as you began to increase your exposure to equities, it would seem you still have some of the sequence of return risks here. Perhaps more muted, but with more risk of missing out on the gains in the early years if they exist as well.

You also have the challenge of knowing how quickly to increase your exposure to equities, since none of us know how long we are ultimately going to live. And for those of us who retired much earlier than the traditional age 65 retirement, there is a much longer time horizon to contemplate in doing this.

But it certainly gives us a very thought provoking concept to ponder as we try to determine what our retirement AA strategy will ultimately be.

 

[ducky911]

This may be true....but I am SURE glad I did not start this way 3 years ago when I started my retirement!

 

[Onward]

It's an interesting and non-intuitive idea. The original paper is here.

IMO the problem with the rising equity glide-path in retirement is that your portfolio gradually gets more volatile just as you, psychologically, are getting less and less tolerant of volatility.

 

[GrayHare]

"The authors... assumed average annual returns of 6.5% for stocks and 2.4% for bonds."

Well, duh, with those assumptions of course a stock-heavy portfolio will perform better than bonds/cash.

 

[Chuckanut]

As I understand, one takes on an AA of say 75%bonds/25%stocks. Then one liquidates the bond portfolio over the first half of retirement while allowing the stocks to ride. Supposedly, the stocks keep increasing in value while the bonds/cash go down, thus resulting in an increasing percentage of stocks over time.

The idea is to insulate oneself against having to sell into a big stock crash at the start of retirement since this seems to be the single biggest threat to the investor's financial health in retirement.

I assume that a big crash later in life does not matter as much since one has built up a huge stock portfolio (remember, one never sells stocks for about the first half of retirement), and one also has fewer years left to need to draw down the powerful.

Isn't simple keeping 3-5 years of needed funds in a safe investment, essentially doing the same thing?

 

[nun]

"The authors... assumed average annual returns of 6.5% for stocks and 2.4% for bonds."

Well, duh, with those assumptions of course a stock-heavy portfolio will perform better than bonds/cash.

.......and a "head in the sand" 4% (with inflation increases) withdrawal rate. Retirement should include a plan to reduce withdrawals in lean times or start off with a withdrawal rate less than 4%.

 

[sengsational]

First time I saw that approach discussed was in Kotlikoff & Burns book in 2008, and I doubt they invented the idea. It is intriguing...

That book, and the program that 'goes with it' would be really good if they were updated to take into consideration the current climate. But yes, I recall reading that idea that the simple taper of equities wasn't optimal....very convincing discussion.

 

[nun]

"According to the research, those who start retirement with 20% to 30% in stocks and end up with 50% to 70% in stocks can withdraw 4% of their portfolio per year and give themselves annual raises to compensate for inflation over 30 years, even in the worst market scenarios. (The authors examined 10,000 simulations and assumed average annual returns of 6.5% for stocks and 2.4% for bonds.)

In contrast, those who keep 60% in stocks throughout retirement or who taper to a 30% equity allocation from 60% are likely to run out of money after 28 years in the 5% of worst-case scenarios"

It's not surprising at all. If you go onto retirement with a high percentage of stocks and the market tumbles in the early years that will greatly reduce your chances of "success", particularly if you keep spending at your planned rate and nibble away at your principal. Having more fixed income will help you ride out that scenario.

The article is an argument for always keeping a cash/pension/fixed income buffer so you don't have to sell equities at a loss to provide retirement income, but we were all planning on doing that anyway.....right? Case in point I'm going into ER with a 60/35/5 AA, but I have enough in cash and a stable value fund to cover 7 years of expenses so I can ride out any market crash.....after all the market looks to be in a bit of a bubble so I am planning for a crash while hoping for more gains.

 

[Tortoise]

I'm not retired, yet. My wife and I plan to start retirement in 9 years with a 50/50 AA and then leave it at 50/50 through retirement. I would not be able to relax and enjoy myself with a much higher equity allocation.

 

[Midpack]

First time I saw that approach discussed was in Kotlikoff & Burns book in 2008, and I doubt they invented the idea. It is intriguing...

That book, and the program that 'goes with it' would be really good if they were updated to take into consideration the current climate. But yes, I recall reading that idea that the simple taper of equities wasn't optimal....very convincing discussion.

Here's the AA vs age chart from Spend 'til The End for comparison...

 

[Meadbh]

I'm not retired, yet. My wife and I plan to start retirement in 9 years with a 50/50 AA and then leave it at 50/50 through retirement. I would not be able to relax and enjoy myself with a much higher equity allocation.

I think it depends on your personal risk tolerance and the size of your portfolio. Imagine a situation in which you ERd with a AA of 50% equities, 30% bonds and 20% in cash and CDs, with a WR of 3%. You could live on cash alone for the first 5+ years and your AA would undergo a corresponding shift towards equities, other things being equal. If the markets tanked during that time, you could buy more equities, which would increase the growth of your portfolio over the long term. If the markets did well during that time, your AA to equities would rise more than expected. IMHO this strategy increases the potential to increase wealth during retirement provided that there is wiggle room at the outset and the investor does not have an irrational fear of market risk.

 

[Katsmeow]

Isn't simple keeping 3-5 years of needed funds in a safe investment, essentially doing the same thing?

I wonder that as well. I thought I had read that - in most situations - doing that (having a cash/near cash buffer basically) doesn't work as well as having investing everything with a more typical allocation. Maybe the difference is between having bonds v. cash/near cash.

Philosophically, I never believed in the cash/near cash buffer idea...yet I have one now. Everything depends on one's situation.

We expect to spend about 30% of our portfolio/savings over the next 3 years basically because of 2 kids in college. We have a 55/45 AA but I was worried about a big drop in the next 3 years which would be very damaging given the need for large withdrawals during this period. I just completed revamping my portfolio to set aside a buffer to cover our withdrawals above a 4% withdrawal rate over the next 3 years. About 2/3 of it is in cash or money market (mostly in a taxable account) and the other 1/3 is in VFSUX - Vanguard Short Term Investment Grade bond. The idea is that we spend down this over the next 2-3 years and don't have to worry even if there is an equity decline during that period. We've gotten out of most medium or longer term bonds (moving that portion of our portfolio to a 5 YR CD at PenFed) although we do still have 15% of our portfolio in Wellesley so still have some longer bonds and also have a little bit in the International Bond portfolio.

 

[nun]

I wonder that as well. I thought I had read that - in most situations - doing that (having a cash/near cash buffer basically) doesn't work as well as having investing everything with a more typical allocation. Maybe the difference is between having bonds v. cash/near cash.

I think the key is to have an allocation that works well enough for you rather than one that might perform "better". There are an infinite number of scenarios and having some low risk cash/cash like investments is just insurance against investment loses.

The majority of my fixed income is now in a stable value fund returning 2.6%. I'm sure that in 5 years time I'll be able to look back and come up with a better asset allocation, but right now having enough in cash and stable value to fund 7 years of retirement is where I want to be because the market can tank and I'll still be able to live comfortably and rebalance.

 

[Animorph]

The study essentially just addressed the few worst case Monte Carlo scenarios, not what might be optimum on average. Easy enough to differ with. It also uses a predetermined (though it may vary year by year) AA and withdrawal, regardless of "market performance". So we can expect that a process that includes information about market performance would lead to a better result. But it does suggest that it might be worthwhile to consider being most conservative just after retirement.

One insight from FIRECalc is that really high WR's have the best chance of success with a higher equity allocation than normal or low WR's. Stocks give you a gambling chance, while bonds just don't give the needed growth. So for the worst case scenarios, where the portfolio value drops too much and the WR's become excessive, the "best" way to recover is with lots of equities.

Those worst case scenarios will also take a few years from the start of retirement to drain the portfolio down to dire levels. So no need for high equity allocations to get out of trouble during the initial retirement years. In fact, a more conservative portfolio would be better, for less downside. After the first few years of retirement, since we're following a predetermined AA, the stock allocation should rise, in case we're in trouble.

In the non-worst-case scenarios where the first few years of retirement are OK for the portfolio, the number of retirement years has decreased and hopefully the portfolio real value has increased a little. The year-by-year WR is probably more conservative than FIRECalc would have allowed at the very start of retirement. Increasing the equity allocation should be relatively safer at that point. The portfolio is better able to handle market declines, and there should be negligible impact on success rates from raising the proportion of equities.

So I think for a predetermined year-to-year AA, rising equities with age makes some sense. For the real world, we might want to wait until we know our future portfolio values before we decide our optimum AA's or WR's.

 

[Options]

I think the key is to have an allocation that works well enough for you rather than one that might perform "better". There are an infinite number of scenarios and having some low risk cash/cash like investments is just insurance against investment loses.

+1
Excellent advice.

Just finished a book on probabilities. In a vastly unknowable and uncertain world, I'd rather not engage in return seeking. Too many behavioral pitfalls. What market crash hasn't demonstrated this?

If you've done your homework, done it right, severe market downturns, even poor sequence of returns, should not be a problem.