Sustainable Withdrawal Rates

[steves]

The problem is that the authors (academic and popular) of these DIY retirement/withdrawal rate strategies are writers. That is, they are addressing their audience passively through the written word.

I can take two individuals with exactly the same age, retirement portfolio, risk profile, and come up with two vastly different strategies they should follow (withdrawal-wise) to manage their investment draw-down.

The reason is simple..... they each have a completely different non-investment data set. One could be paying down a $40K car loan, another might be expecting to realize a large capital gain in 10 years time (selling or downsizing his home). One may have 5 years of rental income coming to him, or a part time job. One has spousal or child support payments for the next 5 years. They might have completely different pension plans (indexed/not/integrated with CPP/not) One might already be taking CPP, the other is stalling taking his CPP. One's portfolio might be RRSP in nature, the other's might be nonreg, One has a 'die-broke/screw the kids' attitude whereas the other may want to ensure a certain estate gets passed on.

There is no way an author such as Jon or Moshe can create a written document/instruction/set of tables which can even remotely satisfy the needs of two identical retirees, each with the same savings, asset mix and risk profile.... it requires too many parameters. It is a computer application.

The only audience for this kind of solution is the guy with the $10M++ nest egg. In this case, the effect of the non-investment cash flows are overshadowed by the large nest egg. Of course, the tax complexion of the nest egg (reg/nonreg/equity/dividends) will effect the outcome differently, even for them.

For the rest of us with our nest eggs in the $100K to $2M range, these sustainable withdrawal rate articles (4%, etc) are virtually useless, IMHO.

[Taggart]

2006 March Issue - Article 6

Decision Rules and Maximum Initial Withdrawal Rates
by Jonathan T. Guyton, CFP®, and William J. Klinger

http://www.fpanet.org/journal/articles/ ... 6-art6.cfm

[steves]

Again.... a complete waste of time.

I can set up two individuals with exactly the same starting RRSP, rates, longevity... etc and run each of them thru a Montecarlo. I will get drastically different results as to safe starting withdrawal rates. The reason is the same as I have stated before... they are completely different from a non-investment perspective. i.e. having/not having .... loans, a future capital gain, spending needs (a new car every 4 years), estate target vs dying broke.....

[Shakespeare]

Don't know if anyone has read this:
"The Calculus of Retirement Income : Financial Models for Pension Annuities and Life Insurance" by Moshe A. Milevsky (Author)

'Millions of baby boomers turning 60 have suddenly awakened to the need to protect their retirement futures. The Calculus of Retirement Income is a useful tool for those devising sensible financial plans and helping manage wealth in the face of capital market and mortality risk. Witty, serious, and entertaining at the same time, the volume will be an invaluable resource for actuarial and financial students, practitioners, and researchers interested in actuarial and financial strategies to avoid ruin.' Olivia S. Mitchell, The Wharton School

[Norbert Schlenker]

Because, when you're withdrawing from a portfolio, volatility is just as important as return.

I want to go back and revisit this remark that I made in February.

A few idle hours on the ferry with a laptop and gummy's Monte Carlo spreadsheet made me think about trying to come up with ballpark estimates that might help people make decisions about appropriate withdrawal rates. In particular, I field a lot of questions along the lines of, "Well, I hear that a 4% inflation adjusted withdrawal rate always worked in the past and should work nine times out of ten over the next 30 years but ..."

• ... what sort of portfolio is that?
• ... past performance doesn't predict future results.
• ... I'm willing to take more risk but I need to know the tradeoff between volatility and return.
• ... I'm retiring early and I think I might need the portfolio to last 40 or even 50 years.
• ... I'll settle for eight times out of ten.

Here's what Monte Carlo taught me.

• Picking a starting point in the middle of the data, you can be 90% confident that a 4% SWR will last 30 years if the portfolio has a real return of 5% and standard deviation of 11%. This is nothing like the risk/return profile of any portfolio out of the Trinity study or follow-ons using US historical data, so the first lesson is to diversify (international, commodities, something) because you can't get that sort of tradeoff without doing so. (Complaints about distributions will be dismissed with a wave. I am just setting an anchor to hang everything else off.)
• Raising the average return on a volatile portfolio by 1% allows you to increase the withdrawal rate by 0.5%.
• Cutting the return volatility by 1% allows you to increase the withdrawal rate by 0.2%.
• I did these simulations for a wide range of returns and volatility around the anchor and that tradeoff - 1% of extra return is worth about as much as 2.5% of reduced volatility - is pretty robust over a wide interval. Running a multiple linear regression gives an r^2 around 0.96 and significance everywhere.
• IOW, the fellow quoted at the top of this post didn't know what he was talking about.
• If you want an extra 10 years, take 0.5% off the withdrawal rate. Take off about 1% if you want an extra 20 years. (Take off 1.5% and you can have 1000 years!)
• If you will settle for 80% confidence instead of 90%, you can add 0.5% to the withdrawal rate.

[steves]

Am I the only one who sees a major problem with 4% safe withdrawal rate (or 3%/5%....)?

The fact is, that unless you take into account all the other non-investment parameters in addition to just the investments such as where you are in the retirement journey... (50 or 85) what other cash flows are going to enhance/detract from those withdrawals (CPP, OAS, part time income, income tax, loan payments, insurance premiums, a future capital gain, different income targets (dying broke or leaving an $x estate)...... then these rules of thumb are just plain wrong-headed.

Any advisor who tried to answer that 'what withdrawal rate' question without asking for the scale, timing and nature of these non investment cash flows is grossly negligent IMHO.

[NormR]

[*] If you want an extra 10 years, take 0.5% off the withdrawal rate. Take off about 1% if you want an extra 20 years. (Take off 1.5% and you can have 1000 years!)

Er, um 1000 years!?

[adrian2]

Er, um 1000 years!?

Sure. Just watch the movie "Kingdom of Heaven". It happens at the time of the first crusades. Part of the extended version has the main characters (Balian and Saladin) talking about the sustainable withdrawal rates

From The Retirement Calculator from Hell, Part III: Eat, Drink, and Be Merry:

The hard part, of course, is how to interpret this kind of output. Realize that these probabilities are merely an imperfect estimate of the investment risk you are taking. In other words, they assume the continuity of financial and political institutions over the period studied. Consider the implications of the above 97% success rate at a withdrawal of $2,500 per month ($30,000 per year). For this to be a useful estimate of your true chance of not running out of money, the "success rate" of your ambient political, economic, and military environment must be at least 97% over this 40-year period. Do you think that this is likely? Only if you are an historical illiterate (which, I’m afraid, subsumes many finance academics). Let’s examine a small sampling of possible political, economic, and military failure modes:


The mildest scenario is that of catastrophic inflation, as experienced in Germany and Hungary in the 1920s or, more recently, in much of the developing world.

Political failures are slightly worse, since these threaten the basic human motivation to work and produce. The state, for whatever reason, can decide to confiscate your assets or, worse, society’s means of production. Anyone who judges this unlikely should turn on CNN during any G-8 or WTO conference.

Local military action. Probably the lowest-probability item on this list, but something to think about on other continents.

The Big One: Some deranged prime minister or colonel in central Russia, Pyongyang, or South Asia could let loose the four horsemen upon the planet.

So, think about what a 97% 40-year success rate means: the absence of all of the above for approximately the next 1,200 years. (A 97% success rate means a 3% failure rate; those 40 years divided by 0.03 is 1,200 years.) Ignore for a minute the uncertainties of the less-developed world and think only about the winners: Germany—in this century alone, three episodes of military and/or economic disaster, the first two associated with mass starvation. Japan—wartime devastation even worse than Germany’s. England—near brushes with disaster in 1812-1814 and in both world wars. And even the United States—repeated banking failures, civil war, and the near-bankruptcy of the Treasury in the 19th century. The near collapse of the capitalist economy in the 1930s. And oh yes, I almost forgot—the entire globe barely missed mass incineration in October 1962.

History’s best-case scenario was the Roman Empire, which survived more or less intact for about seven centuries (if you ignore the odd sackings of the capital after 200 A.D.).

A wildly optimistic historian might give us another few centuries of economic, political, and military continuity. Back-of-the-envelope, that’s about an 80% survival rate over the next 40 years. Thus, any estimate of long-term financial success greater than about 80% is meaningless.

[Gus]

A wildly optimistic historian might give us another few centuries of economic, political, and military continuity. Back-of-the-envelope, that’s about an 80% survival rate over the next 40 years. Thus, any estimate of long-term financial success greater than about 80% is meaningless.

Adrian's survival rate and Norbert's are entirely independent risk factors. If you want to know if you'll still have a nest-egg before Armageddon happens, then multiply the two numbers together. 90% times 80% gives 72%. The fact that there is a chance of world political collapse before your money runs out means that there is only an 8% chance of living in a peaceful world with no money in 40 years time. I feel much better now.

But Adrian makes a valid point, non-financial catastrophe may be more likely than financial catastrophe, so we need a sense of perspective. Further, purely financial failure can, to a degree, be foreseen and could be forestalled, albeit painfully, by cutting withdrawal rates if, for example, your investments were starting to go to hell.

There is a third independent factor as well, your own personal survival chances over 40 years. Maybe I should buy that damned boat after all...

[Norbert Schlenker]

The fact is, that unless you take into account all the other non-investment parameters in addition to just the investments such as where you are in the retirement journey... (50 or 85) what other cash flows are going to enhance/detract from those withdrawals (CPP, OAS, part time income, income tax, loan payments, insurance premiums, a future capital gain, different income targets (dying broke or leaving an $x estate)...... then these rules of thumb are just plain wrong-headed.

But, Steve!

Yes, all those things you listed are important. No, they have nothing to do with withdrawing from the portfolio. These are separable problems.

Any advisor who tried to answer that 'what withdrawal rate' question without asking for the scale, timing and nature of these non investment cash flows is grossly negligent IMHO.

Quite true but irrelevant.

Er, um 1000 years!?

Er, um, yeah. While Adrian's quote from Bernstein is good to think about, because the real world is undoubtedly more powerful than financial mathematics, the math is at least worth looking at. When it comes to withdrawal rates in a model/virtual/perfect/ever-growing-GDP world, the difference between 30 years and eternity is 1.5%. Don't get too bothered: these are just estimates for partial first derivatives.

only an 8% chance of living in a peaceful world with no money in 40 years time. I feel much better now.

Maybe I should buy that damned boat after all...

You know you're always welcome on Luna. Sorry I missed the race last night.

[steves]

they have nothing to do with withdrawing from the portfolio. These are separable problems.

Yabbut, when you talk about withdrawing from the porfolio, we aren't talking about the color of the stock certificate we are withdrawing or the first letter of it's name, we are talking about the dollar amount of the withdrawal. If we have a 200K portfolio and we determine that a safe withdrawal rate is 4%, then the size of the withdrawal would be $8K. So does that mean that everyone with a 200K portfolio of a certain age should be making that same 4% withdrawal? I don't believe so. Some might be taking a lot more, some a lot less, depending on the other cash flows... or expected cash flows.... A loan which is being paid off over several years, one person might be already in receipt of CPP, another not, one might have a part-time job, another expects to downsize his home in 3 years and will be in massive 'negative withdrawal mode' for that year.

Surely the measurement should not be simply pre tax investment withdrawals, shouldn't they rather be after tax income (lifestyle) from all sources. Many of these non investment cash flows are erratic and discontinous and the investment withdrawals we make should be governed by their relative size and timing. How do you account for the fact that many seniors will be topping up their investment nest eggs at various times.. an inheritance, sale of real estate, etc. Or making adhoc (new car/world cruise...) purchases every 5 years.

[NormR]

Er, um 1000 years!?

Er, um, yeah. While Adrian's quote from Bernstein is good to think about, because the real world is undoubtedly more powerful than financial mathematics, the math is at least worth looking at. When it comes to withdrawal rates in a model/virtual/perfect/ever-growing-GDP world, the difference between 30 years and eternity is 1.5%. Don't get too bothered: these are just estimates for partial first derivatives.

Ok, how about including error estimates with your mathematics?

[AltaRed]

I am curious about one specific. The discussion allows for a range of holdings - of course, anywhere from 25% equities on up. There's a single "rule of thumb" for withdrawal - the 4% figure, moderated only by a reduction to 3.3% if you expect to live a long time (and I do!). Surely though, the mix has some effect - a 50% equity portfolio is going to, sooner or later, outperform a 25% - otherwise why bother?

Digging up old debate, but going along with the responses to this Feb 06 post, I became fairly intrigued by the FIRECALC calculator which is populated with US market history. While not totally relevant to Canada, and not a plug for its use, it does deal with the questions on probability of reaching a certain age with assets still in hand, and calculates withdrawal rates accordingly. In most of the 'what ifs' I did using that calculator earlier this year, something less than a 100% equity portfolio provided the best overall result. That is, of course, because of equity volatility as Norbert pointed out.

[Norbert Schlenker]

Don't get too bothered: these are just estimates for partial first derivatives.

Ok, how about including error estimates with your mathematics?

Error estimates on the partial derivatives?

Within ±3% of real return and ±5% of standard deviation, I am confident that the derivatives are

0.5±0.1% vs. real return
-0.2±0.03% vs. standard deviation
-0.055±0.01% vs. years away from 30 that the portfolio needs to last
-0.055±0.01% vs. increased or reduced confidence in the result

So does that mean that everyone with a 200K portfolio of a certain age should be making that same 4% withdrawal?

My underline. Not "should". "Can". Surely you see the distinction.

[Bylo Selhi]

There is a third independent factor as well, your own personal survival chances over 40 years. Maybe I should buy that damned boat after all...

Some depressing reading [my emphasis]:

While it's easy to find articles on reimagining retirement, active retirements, etc., the sorry truth is that millions more people believe they have this option than actually have it.

A recent study at the Center for Retirement Research at Boston College found that individuals in their 50s were likely to experience one or more shocks that could dramatically reduce their retirement security.

Worse, it was more likely that you would experience one of these shocks than not. The study found there was a 69 percent chance that between age 51 and age 61, an individual would experience at least one of these events:

• A major medical condition (41.3 percent).

• A health-related work limitation (33.7 percent).

• Severe disability (6.9 percent).

• Enter a nursing home (3.4 percent).

• Be laid off from job (18.7 percent).

• Be divorced (2.3 percent).

• Be widowed (7.3 percent).

The percentages for each event, by the way, add to more than 69 percent because some individuals experienced more than one of the events. When it rains, it pours.

Only three people in 10 get through their 50s unscathed.

See also: How Secure Are Retirement Nest Eggs?

Note that "major medical condition" includes, "heart problems, cancer, serious lung problems, diabetes, strokes, and psychiatric problems(*)" so that's not necessarily a death sentence.

(*) "Heart ailments are especially common. Over a 10-year period, 16 percent of adults ages 51 to 61 at baseline develop serious heart problems. About 10 percent contract diabetes, 10 percent develop cancer, and 9 percent experience psychiatric problems [that's likely to increase, especially depression, once more people read this study...Bylo ]. Only 5 percent have strokes and 6 percent contract lung disease."

Added: National Retirement Risk Index

The National Retirement Risk Index (NRRI) measures the percentage of working-age households who are at risk of being unable to maintain their pre-retirement standard of living in retirement. It addresses one of the most compelling challenges facing the nation today — ensuring retirement security for an aging population.

Key findings in the National Retirement Risk Index show that:
• The retirement landscape is shifting dramatically, making the outlook for retiring Baby Boomers and Generation Xers far less sanguine.
• Over 40% of households are "at risk" of not having enough to maintain their living standard in retirement.
• Saving more and working longer may substantially improve the outlook... [plus links]

[Gus]

Maybe I should buy that damned boat after all...

You know you're always welcome on Luna. Sorry I missed the race last night.

So, Norbert, have recent events forced you to change your estimates of the likelihood of financial shipwreck?

(Private joke, as I nurse a black eye and bruised knee)

[The Wealthy Boomer]

Here's a version of what I just posted to the blog:


Monday’s Advisor Post looked at a new book by American financial planner Bill Bengen: Conserving Client Portfolios During Retirement. Bengen’s 4% “safe” annual withdrawal number was cited by Lee Eisenberg in his bestselling The Number, excerpted in FP Weekend in the spring.
Despite its US$65 price tag, I’ve had several emails from Canadian financial advisors interested in how to get the book. It’s not likely to be stocked in places like Chapters, so online is the best way to get it.
Here’s the URL:

http://www.fpanet.org/member/press/rele ... bengen.cfm


Re: discussion of this, I've also had a number of reader emails pointing out that the "safe" annual withdrawal rate of 4 or 5% is far below the withdrawal rates for Canadian RRIFs once clients get near or past 90. In other words, we are forced to exceed a "SAFEMAX" rate whether we need the income or not.

Any suggested remedies?

[Shakespeare]

far below the withdrawal rates for Canadian RRIFs once clients get near or past 90

But 4% is based on a fairly long retirement period - 25-35 years, depending on the program. How many people at age 90 have 25-35 years left?

[DanH]

RRIF annuitants must understand that at some point, they shouldn't spend all of their RRIF withdrawal if they need their RRIF to kick out a sustainable real income for their lifetime.

[Bylo Selhi]

How many people at age 90 have 25-35 years left?

Optimists who make very conservative financial plans

RRIF annuitants must understand that at some point, they shouldn't spend all of their RRIF withdrawal if they need their RRIF to kick out a sustainable real income for their lifetime.

And this isn't an issue only for those in their 80s and 90s. The minimum RRIF withdrawal exceeds 7% at age 71, so almost anyone who plans based on a 4% or 5% withdrawal rate and who has only a RRIF needs to "save" some of what they pull out.

[DanH]

And this isn't an issue only for those in their 80s and 90s. The minimum RRIF withdrawal exceeds 7% at age 71, so almost anyone who plans based on a 4% or 5% withdrawal rate and who has only a RRIF needs to "save" some of what they pull out.

Probably but the percentages are on the year-end balance whereas the safe withdrawals express a STARTING percentage, where the dollars are indexed thereafter. Illustrations I've done where it's just an income supplement shows that spending all of the withdrawal doesn't hurt their retirement finances. If you compare the RRIF minimums with those starting at 4% (and indexing $ thereafter) you'll see the actual withdrawal % vary quite a bit over the years.

It's those that have relatively little in pensions to fall back on that have to watch their RRIF withdrawals more carefully and, at some point, start putting some of it aside.

[steves]

Hey... it's Thursday, so it must be time for my 'safe withdrawal rant'

Here is how to determine a list of those who would find this book useful. If an individual has any of the following elements in their financial makeup.... this would probably exclude them from finding this book of any practical value......

- 1 or more outstanding loans
- expectation of a future capital gain (windfall, real estate sale...)
- CPP and/or OAS which is not yet being received
- the desire to pass on a sizeable estate
- cash flows which will cease at a future time (support payments, rental income...)
- an anticipated/planned need for extra income (new car every 4 years, once in a lifetime world cruise, planned transfer of wealth)
- finding themselves at the tail end of their lifespan
- having their savings outside rather than inside their RRSP

Eliminating these above individuals from the overall prospect list for the book would leave... I dunno, maybe 32 people.

[brucecohen]

Any suggested remedies?

1. In your 70s use most of the RRIF money to buy a life annuity. Keep some in the RRIF to create your own inflation indexing if you can't get a satisfactory quote on an indexed annuity.

2. I believe Dan alluded to this in suggesting that RRIFers should not spend all of their withdrawal, but to flesh it out a bit.....

A) If 4% is the target withdrawal rate, you're telling me that's an adequate amount for Joe...

B) But the law says Joe must take out, say, 9%....

C) So, Joe should take what's left after tax of the gap between 4% and 9%, and invest it outside the RRIF. While this would not prevent the erosion caused by the 9% withdrawal, it would reduce it.

3. Consider whether this really is a problem. What's the worst that can happen? The combination of high withdrawal rates and poor investment performance could wipe out Joe's RRIF money. But he would still have fully-indexed CPP and fully-indexed OAS PLUS he would now be eligible for GIS that's not only fully-indexed but also tax-free. What's all this worth? Just back to all the columns you wrote, quoting Malcolm Hamilton's analysis that it doesn't pay to save when govt is so generous.

4.Why don't you call Finance and ask for the rationale used to work out the current minimum withdrawal schedule? Also ask how many RRIFers only pull out the minimum; they used to track that and may still. You could ask if they've done any work on gauging financial hardship caused by RRIFers running out of money, but they'd likely tell you they don't consider that a meaningful concern because of the annuitization option. But ask.

[John H.]

Always an interesting topic with various views, so here are mine. Firstly I do not subscribe to the "safe" wthdrawal, I believe that the withdrawal should be at the level required to meet your needs be it 4% or 15% or whatever with the only caveat being that you understand the implications of your withdrawals.
I also believe that any withdrawals should be made in the most tax efficient manner possible, in that if you only need a 5% withdrawal to meet your needs but can take a 15% withdrawal within the same tax bracket, take the 15% and reinvest the "not needed" amount in after tax dollars, and where possible, as these are after tax dollars, invest in a joint account to take advantage of future tax defferals and income splitting opportunities.
In the present interest rate environment I also do not subscribe to annuitizing your RRIF or any part thereof as the only ones assured of making money are the insurance company and the insurance agent. I believe that a tax efficient withdrawal and reinvestment program will meet the needs of the majority of the people most of the time.
In the intrests of full disclosure I am a " Financial Advisor" who has the ability to provoide a full range of investment and insurance products to my clients.
I placed the phrase "Financial Advisor" in quotation marks as I know this board to be a DIY centre wth a certain distain for the so called "professionals" in the industry but I would like to think that we each have the ability to lean from the other with the end result being that we are all more successful at what we do, and let's face it, we all have the same goal of increasing wealth be it of you, me or my clients.

[steves]

I concur with your view on taking just enough from your RRSP/RRIF to keep below the first tax bracket (i.e. keeping just below the basic personal limit), but when you try and extrapolate this to include each successive tax bracket, I have never been able to verify this.

The problem with tax efficiency and the preferential juggling of reg and non-reg capital is that it involves very complex time value of money math... you can't look at it using a simplistic spreadsheet methodology. Also, life expectancy has to be considered…. there is an estate planning issue here as well.

The problem is that the determination of tax efficiency is to measure the present value of all those future tax payments, and further, the tax formulation itself is not static... the brackets are indexed to inflation. Those large future RRIF withdrawals are not so scary tax-wise when you consider that the effective tax rate goes down in later years and that those future taxes have to be looked at in present value terms..

My major complaint with the safe withdrawal concept is a cash flow one..... if you are living in a world where there was nothing except your investment nest egg (no expectation of entitlement nor future lump sum income) no cash flow income or expense which had a discontinuous nature, (support payments, loan payments..) then the concept of a safe withdrawal rate makes some sense.

But, if we accept the idea that these withdrawals are meant to fund our (constant) lifestyle, then the withdrawal schedule will jump all over the map as these various 'non-lifestyle' cash flows come in and out of play over time.