Sustainable Withdrawal Rates

[freedom_2008]

What's wrong with dying? I still think we need assisted suicide in this country. At least I know I don't have to save for a retirement home.

newguy

Nothing wrong with dying willingly in old age, be it assisted or not. Looking ahead, the cost of keeping old dying people alive, might be too expensive for our young. When the time comes, we are willing to go into beautiful forest and back to nature. But it shouldn't be forced by lacking of savings.

[Arby]

I'll google the study later just to check. In passing, I find it deplorable that the author gives neither the titles nor the authors of the studies that he relies on.

George, I'd be interested if you find any more recent data.

The concept of decreased spending as retiree's get older appears to have started with a paper on Reality Retirement Planning. A more recent analysis is provided in the paper How Do Spending Needs Evolve During Retirement.

[ghariton]

George, I'd be interested if you find any more recent data.

Best I can do is is this paper from Department of Finance (Canada). It was published in 1999 and uses data for 1994, so 18 years out of date. Still, Chart 1 on page 6 shows Canadians' savings rate by age group.

The first thing to note is that the chart gives three differently defined savings rates. Taking the unadjusted savings rate, it looks pretty flat starting at around age 45. If one adjusts the savings rate to exclude income from pensions, one does get the V-shape suggested by Vettese: it hits a peak at 50 to 55, then declines steeply, going quite negative by age 65 or so, then turns upward again, returning to positive by age 85 or so, and continuing to increase at 90 (the greatest age shown). But it never gets back to the levels of the earning years.

The third definition excludes both pension income and capital gains. It shows much the same pattern as the second definition, only more so.

Again, I stress that the data are 18 years old, so I don't think they are all that relevant any more. But for what it's worth, they don't support Vettese.

George

[brucecohen]

George, you might be interested in this and the 2005 Journal of Financial Planning article to which it links.

[ghariton]

George, you might be interested in this and the 2005 Journal of Financial Planning article to which it links.

Thank you.

Let me quote the last paragraph from Pfau's article:

But what is the best assumption to use: constant inflation-adjusted spending, or decreased spending as one ages more? This is a big question that I think is still not fully resolved. I'd like to find a Ph.D. student willing to dig more into the household survey data and to classify different retirees by their spending patterns over time using surveys that do indeed track the same households over long periods.

I think that's right. We need some longitudinal studies of savings rates or spending patterns, and all we've got is some cross-sectional studies.


BTW I found it ingteresting that Finance in its paper used three different definitions of the savings-to-income ratio. That got me to thinking about the definition of income in other areas. For example, income inequality: When estimating measures of income inequality, to the researchers take into account such things as pensions and benefits when earned (as opposed to when collected)? I bet that would make a huge difference.

George

[johnfranks]

I have skimmed over this thread, and realize that everyone's situation is different and that the answer depends hugely on assumptions about returns, etc.

BUT...
(I'm jumping in with a broad question and will probably regret it...!)

My situation is that I was forced into early retirement in my 40's, and now must rely on savings to last me the rest of my life. That's the bad news. The good news is that I have no debt, and living expenses are fairly modest ($30k-$35k). My savings are a mix of equity ETFs, provincial bonds, Canada real bonds, and cash. It's nothing fancy, but I get a decent dividend and interest stream from it. I know that I can live off of this for at least the next 20 years.

My question is, what about after that? What kind of withdrawal rate can I do to make it last a lifetime, which may be 40 years or even longer if I'm [un]lucky. Based on this thread, it doesn't seem that a constant withdrawal rate is realistic, and that withdrawals are likely to drop off after age 65. Plus, I'll get a reduced CPP at that age, and maybe OAS at age 67 if they don't extend it even further. That alone should drop my withdrawals by at least $10k (adjusted for inflation).

Can I use CPI as a realistic measure of my inflation going forward for the next 40 years? I'm concerned that health costs may push my own inflation higher than CPI. I have no health plan to cover that, other than OHIP of course. Maybe using nominal GDP growth is a better measure of inflation, but that means I'm screwed so I'd rather not use it! What do those of you in your 60's, 70's, and 80's experience for inflation?

Here's what I have been doing for my sustained withdrawal "budget". I take my net worth, and divide it by (85 - my current age). That gives me my yearly budget. It's essentially just dividing my savings by the number of years I have left to live. The upside is that I never run out of money until age 85. The downside is that it's very dependent on how the bond and equity markets do. I'm not sure if anyone else does it this way, but if you do something similar, does it work over the long run? I'm a little concerned I've been over-budget the past 2 years because of unexpected one-time expenses, so this is what makes me wonder about what happens 20 or 30 years from now.

Thank you in advance!

[BRIAN5000]

My question is, what about after that?

These two articles may be of interest

Ten Reasons Why the 4% Rule is Too Simplistic for Retirement Planning

http://wpfau.blogspot.ca/search?updated ... results=50

and

from the comments section of another article " It was reassuring to read that Bernstein thinks a 2% withdrawal rate is bulletproof at any age below 65."

http://wpfau.blogspot.jp/2011/11/length ... -safe.html


40.00 years retirement = 57.60% equity, 3.71 % withdrawal is the supposed sweet spot ( I guess it has to be all nontaxable per the Top ten above, etc. ?)

It seems to me equity can be from 35 % to about 75% with no change in longevity just a change in bequest size.

[Shakespeare]

See also Sustainable withdrawal - finiki

and the "Calculators" section.

[freedom_2008]

Thanks for the finiki work. There seems to have been lots of thought put into the topic, very interesting discussion.

But can't one just use the simple way: to live within your means (i.e. income) during retirement? We are also in our early/mid 50s', retired, no pension, living on interest and dividend income only. If we always need less than what comes in, with less than 20% in (dividend paying) equity and about 1/3rd of that in prefs, about 30% in provincial bonds, a bit more than 50% in 5yr GIC ladder and cash in HISA (enough for 5 years living), we haven't felt the need to give much thought about SWR for ourselves. Do we really need to? Have we missed anything? I guess one thing we don't care about is whether or not we die broke.

[johnfranks]

Thanks for the links. I'll check them out.

Yes, living within means is important. But I'm trying to figure out what is my means and what I can expect as living expenses in the future. If I had a pension it would be easy, but without a pension I'm unsure of what my future income stream will look like.

[kcowan]

You have probably already found some ways to save money that you never tried when working. Just focus on reducing your ongoing costs. That is the only thing you can control.

Have you considered living on a cruise ship? They always look for skilled people to offer services to their paying customers and give them free room and board.

[Pickles]

Yes, living within means is important. But I'm trying to figure out what is my means and what I can expect as living expenses in the future. If I had a pension it would be easy, but without a pension I'm unsure of what my future income stream will look like.

I'd consider taking early CPP at 60 since you have so few years of contribution. Otherwise the additional 5 years of non-contribution (between 60 - 65) work against you in calculating youe pension entitlement at age 65. Yes, you will pay a 36% penalty but avoiding the negative effect of five more years of non-contributions adversely affecting your CPP calculation at age 65 will make up some of the difference AND you will have a steady stream of income 5 years sooner.

I am assuming you have some medical condition that prevents you from finding any work in the next 20 years. If not, any work you can pick up will have a significant positive effect on your rate of withdrawal. If you do have a medical condition, you might look into the possibility of the CPP disability pension (which is based on health, not contributions and covers you until age 65).

[johnfranks]

Have you considered living on a cruise ship? They always look for skilled people to offer services to their paying customers and give them free room and board.

If I'm very bad in this life, a cruise ship is where I'm going to end up after I die.

I'd consider taking early CPP at 60 since you have so few years of contribution. Otherwise the additional 5 years of non-contribution (between 60 - 65) work against you in calculating youe pension entitlement at age 65. Yes, you will pay a 36% penalty but avoiding the negative effect of five more years of non-contributions adversely affecting your CPP calculation at age 65 will make up some of the difference AND you will have a steady stream of income 5 years sooner.

Really? That's very interesting. It's a long way off, but I'll definitely keep that in mind.

I am assuming you have some medical condition that prevents you from finding any work in the next 20 years. If not, any work you can pick up will have a significant positive effect on your rate of withdrawal. If you do have a medical condition, you might look into the possibility of the CPP disability pension (which is based on health, not contributions and covers you until age 65).

It's complicated. I'm not sure if I'd quality for disability, but in any case I don't feel comfortable even trying to claim it. I could work, but at this point the kind of work I'm both qualified for and medically capable of doing would not pay much. So I haven't considered that yet. I suppose that's plan B. Plan C would be to retrain for a different career, but that's no guarantee and it's outside of budget. I'm happy with plan A now, it's an uncertain [distant] future that has me concerned. Who knows... I could be dead by then so it won't really matter. My family life expectancy is quite low, so I have that working in my favour... sort of!

[Insomniac]

If I'm very bad in this life, a cruise ship is where I'm going to end up after I die.

That's the way I feel too; going on a cruise is like being stuck on a BC Ferry for days on end! The horror...

[BRIAN5000]

Starts in 15 minutes.

http://wpfau.blogspot.ca/

On Friday, September 14, at 4pm Eastern Time, I will be presenting a webinar called, "Retirement Income With Systematic Withdrawals And Immediate Annuities" at Advisors4Advisors. My understanding is that anyone can freely attend by just filling out the registration form. You are welcome to say either yes or no for the question about starting a trial for the 6-month membership.

Do systematic withdrawals and immediate annuities form the best underlying core in a retirement income strategy?

At this webinar, Wade Pfau, an expert on retirement planning strategies will present the case for each strategy.

Controversial elements of these strategies will be highlighted, such as the reliability of the historical record, biases in advisor recommendations, the role of cognitive decline, and the ability of retirees to follow the assumptions guiding systematic withdrawal rate research. Other issues that will be the focus of this session include:
the role of taxes
framing annuities as investments or risk-reduction tools
helping clients build a legacy
credit risk of annuity providers

Unless you needed the CE credits that was not as good as I had hought it might be.

[nile]

What should my additional withdrawal rate be to be broke at 90?
Any and all thoughts,links etc appreciated
Assume my portfolio is 100% Canadian equity with serious dividend bias.

Also just finished watching homeland season 1 in blu ray.
If anyone wants it I will mail it to them as long as they promise to pass it along once done.

Thanks in advance

Nile.

[StuBee]

What should my additional withdrawal rate be to be broke at 90?
Any and all thoughts,links etc appreciated
Assume my portfolio is 100% Canadian equity with serious dividend bias.

Also just finished watching homeland season 1 in blu ray.
If anyone wants it I will mail it to them as long as they promise to pass it along once done.

Thanks in advance

Nile.

My initial thought was to apply a "safe withdrawal rate rule" of 4% of the initial value of your portfolio and index this by a factor equivalent to inflation. However, ISTM that it is reasonable to believe that your dividend stream growth could easily match (and even outgrow inflation) so that at age 90, your capital would remain (still) untouched.

Since you have not mentioned your age, let me imagine that you are 65 and a neat way would be to apply the RRIF withdrawal formula (without modifying it at age 71).

On January 1st of each year, determine the value of your portfolio and multiply it by: 1/(90-age). This guarantees that you will be "broke at age 90"

StuBee

[dns902]

Of course, decreasing real rates of return will lead to lower SWR. (I say real, because SWR is usually expressed in after-inflation terms. Thus, a 4% SWR is to be interpreted as withdrawing 4% the first year and indexing the amount to inflation thereafter.)

As a quick illustration of the impact of changing real return on the SWF, consider a thirty-year annuity certain, in real terms, with r the real rate of return throughout the thirty years. If r equals zero, pretty close to the situation today, the SWF is obviously 3.33%. (I note that such an annuity completely eliminates market risk, although credit risk remains.)

If r increases to 0.5%, SWR increases to 3.66%

If r increases to 1%, SWR increases to 3.87%

If r increases to 2%, SWR increases to 4.46%

If r increases to 3%, SWR increases to 5.10%

Many of the studies of SWR were performed at a time when real returns were 2% or so. Now, real returns are closer to 0.5%. My simple analogy suggests a drop in SWR of 1.0%, to somewhere around 3%

Such numbers are approximations of course, and don't take into account transaction costs, which would lower the SWF somewhat.

George

As an approximate rule of thumb for retirement calculations, the above figures can be genaralized to include term period n years and real rate of return r via the formula: SWR = 100(1/n + r/2)%

David

[BRIAN5000]

"Perhaps surprisingly, bonds, inflation-adjusted SPIAs, and VA/GLWBs do not serve a useful role in the couple’s optimal retirement income portfolio."
"The third generation clarifies how 4% is only an educated guess based on limited historical data and some rather simplifying assumptions."

An Efficient Frontier for Retirement Income

http://papers.ssrn.com/sol3/papers.cfm? ... id=2151259

[Peculiar_Investor]

We're still a number of years away from worrying about this, however with a tip of the hat to Bogleheads forum, a guest post on Wade Pfau's blog seems worthy of a read, Retirement Researcher Blog: Guest Post: Ken Steiner on how Actuaries Think About Retirement Income.

[Insomniac]

We're still a number of years away from worrying about this, however with a tip of the hat to Bogleheads forum, a guest post on Wade Pfau's blog seems worthy of a read, Retirement Researcher Blog: Guest Post: Ken Steiner on how Actuaries Think About Retirement Income.

Thanks!

I followed the link to Ken's site ( http://howmuchcaniaffordtospendinretirement.webs.com/ ) and downloaded his article and spreadsheet.

[adrian2]

Hat tip to Bogleheads: Is the 4% Rule Folly? A staple of retirement income planning leaves key risks and tradeoffs unaddressed.

Add to that the importance of the first decade in retirement on portfolio failure rate, known as sequence of return risk, and relying on the safety of a 4% rule starts to look like a sucker’s bet. We estimate a failure rate beyond 50% if real yields don’t revert to their historical mean, and a 20% failure rate if yields mean revert after just five years. Unfortunately, there’s no evidence that bond rates have mean-reverted predictably in the past.

The most sobering conclusion is that retirees who believed a 4% withdrawal rate was safe in the past should instead rely on a much lower rate—perhaps as low as 2.8%—to achieve the same degree of shortfall security. Such a low rate suggests that workers need to save a lot more and retirees should cut back sharply on their lifestyle to avoid the risk of running out of money. But it doesn’t have to be that way. That’s because there’s something else wrong with shortfall studies.

Longevity Risk

Running out of money is usually at the top of the list of concerns when building a retirement income plan. It should be. The traditional shortfall methodology assumes you spend down your assets without a safety net. What happens when the trapeze artist performs without a safety net? Obviously, they’re going to be more cautious and a lot less fun to watch. You can think the same way about a retirement income plan without a longevity hedge. If you bear that risk, you’re going to be a lot more careful. And careful does not a fun retirement make.

Bearing longevity risk without a hedge is stupid. Many advisors would recoil at the idea of investing a retirement portfolio in the stocks and bonds of one company. The reason? Excessive unsystematic risk means that you get no expected reward for the amount of risk taken. These same advisors, however, don’t think twice about building a retirement portfolio without a longevity risk hedge even though the client gets no reward for the amount of longevity risk they assume.

[...]

The tradeoff of annuitization is reduced liquidity available to meet unexpected expenses, such as a decline in health and a lower bequest amount. But wrapping any investment in an annuity, whether it be the safe bond portion of a portfolio using a SPIA or risky assets in a variable annuity, provides greater annual spending and a safety net that protects against a severe drop in spending if investment assets run out in old age. It is important to note that expenses embedded within each of these options will impact optimal strategies.

So what is the right plan? The answer is that it depends on clients’ risk aversion (and assessing risk aversion isn’t easy) and how much happiness they get from spending their own money versus passing it on to their heirs. The optimal plan for risk-averse retirees leans toward full annuitization. For more risk-tolerant retirees or retirees with a stronger bequest motive, the best option is to annuitize later in life (or buy a deferred annuity). It should be noted that Moshe Milevsky has been making this point for over a decade.

Mitchell and her co-authors estimate that the average retiree could improve expected happiness in retirement by as much as 50% by employing a blended annuitization and investment strategy. They also find that one of the best ways to improve welfare isn’t through just annuitization but flexibility in withdrawal rates. This is the last problem with the 4% strategy: It doesn’t allow retirees to ratchet down spending if they experience bad luck in their investments or increase spending if they do well. This lack of flexibility both increases the likelihood that they’ll run out of money when faced with low investment returns, and leaves more money on the table if markets perform well.

[steves]

Try and spot the word "tax" in that quote. These SWR pros badly over-simplify things. Surely the aim would be to maintain a constant standard of (after tax) living. The retirement income stream surely has to co-exist with a whole lot of other sources of income/outgoe CPP, OAS, loans paid off, windfalls, sale of real estate, different forms of capital (reg, txbl, nontxbl) coming in and out of play and the drastic change in their tax effect over time. Sigh.

[Bylo Selhi]

These SWR pros badly over-simplify things.

No, they don't. They don't claim or try to compete with your software either. Get over it

I had to read all the way to the end to discover that the 4% rule still has value as an investment planning tool even if it isn't some immutable law or iron-clad warranty

So what should we make of a 4% rule that is based on a process that fails to hedge longevity risk, may be overly conservative and spends a fixed amount each year regardless of investment experience? We should recognize that this approach was useful in beginning the conversation about optimal withdrawal rate strategies [my bold], but science is all about constantly improving on old techniques. Best practices mean changing with the times.

Was—and still is. But like any rule, it's only a starting point that remains useful to anyone who's thinking about how much they should plan to set aside in order to fund their retirement.

Does it address all the issues? No. Nor does it claim to.

Does it have flaws? Yes.

Is it useful, nevertheless? You betchya.

When I learned to drive I was taught to leave a car length behind the vehicle ahead for every 10 m/h (~16 km/h) speed. It wasn't meant as an immutable law. Everyone knew it had flaws, e.g. it wasn't enough in wet/icy weather, it wasn't practical in heavy traffic where such other drivers would quickly fill such a large void, etc. But as a general rule in the absence of anything better, e.g. an onboard collision avoidance system, applied with some common sense, it's served me well for several decades—just as 4% SWR has.

YMMV, etc.

[CROCKD]

I maintain that SWR is an oxymoron or at best a misleading term. It should be restated as SSR- sustainable spending rate, net after tax. With my RRIF representing the lion's share of my resources and a requirement to withdraw well in excess of 4% before tax, SWR does not mean anything unless one defines it as a % of the total of ones resources that can be spent in a year without exhausting them.