4% rule (paper by Sharpe, Scott, and Watson)

Here is a paper criticizing the often-cited "4% rule".
http://papers.ssrn.com/sol3/papers.cfm?abstract_id 05652 Efficient Retirement Financial Strategies
William F. Sharpe, Jason S. Scott, and John G. Watson July 2007
"Unfortunately, the 4% rule represents a fundamental mismatch between a riskless spending rule and a risky investment rule. This mismatch renders the 4% rule inconsistent with expected utility maximization. Either the spending or the investment rule can be a part of an efficient strategy, but together they create either large surpluses or result in a failed spending plan.
,,,
Overall, our findings suggest that it is likely to be more fruitful to clearly specify one's assumptions about a retiree's utility function, then to establish the optimal spending and investment strategy directly. Of course, one should take into account more aspects of the problem than we have addressed in this chapter. Annuities should be considered explicitly, rather than ruled out ex cathedra. Separate utility functions for different personal states (such as "alive" and "dead") could be specified rather than using a weighted average using mortality probabilities, as we have assumed here. Yet our analysis suggests that rules of thumb are likely to be inferior to approaches derived from the first principles of financial economics."
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
snipped-for-privacy@aol.com wrote:

Maybe I need to read the paper a second time, but Sharpe lost me. He criticizes the Monte Carlo type results often used to explain the 4% rule, not liking the idea that one can run out of money (very small chance) or have 2X their starting value. But what I seemed to miss is what method he proposed to use instead. Having one's income swing wildly as they vary withdrawals based on prior year return makes little sense to me, but it would seem some adjustment is in order, the 4% not being cast in stone. JOE www.joetaxpayer.com
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Yeah, the paper didn't seem to be very prescriptive. I do like the point, though, that it's folly to expect to be able to withdraw a fixed amount per year (or an amount that increases only with inflation) for the rest of your life; because the likely outcome is feast or famine.
Indeed, the feast-or-famine situation will apply to any strategy in which the withdrawal amount does not depend on the portfolio's performance.
So here's an alternative. Suppose we pick a percentage and say that at no time can the maximum withdrawal exceed that percentage. We have now changed our main criterion from being a specific dollar withdrawal to one that depends on the portfolio's value.
That change guarantees that we will *never* run out of money. Of course, if the portfolio tanks, we will asymptotically approach running out, and we won't be able to withdraw very much; but that would have happened if a fixed withdrawal had exhausted the portfolio also.
To make things more specific, suppose we say that our desired withdrawal rate is 4%, but that we will never allow it to exceed 6% in any given year. So we start with a 4% withdrawal, and adjust it up or down each year depending on the portfolio's performance. If the portfolio goes down, we will allow ourselves to withdraw up to 6% if we need it, but we will consider this to be a "we're in trouble" situation and do our best to limit spending until the portfolio recovers.
Over the long term, this strategy will always give us more money than we need, so long as the portfolio's long-term performance exceeds inflation + 6%. Moreover, if we can limit our withdrawals to 4%, the likely eventual outcome is that the portfolio will grow to the point where we don't have to worry about it any more.
Of course, if we have several bad years, we may have to scrimp. But that's true whether you're living off your investements or working -- if your job goes away, or you decide to change careers, you may have to scrimp too. The good news is that by matching your spending to your investment performance, you can avoid ever running out of money.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Tell me how this works for the past few years. It's the end of June, and this is when one Jane Smith examines her yearly portfolio returns. For the year ending June 30 2007, Jane's portfolio has skyrocketed. It's up around 25%, because she owns mostly the ETF SPY. It would seem from your proposal, she can start spending way more than previous years. She starts doing so. Then the market dives. We do not know how low it's going. Might it correct to the relative lows of early 2006? It could. Does Jane keep spending her precious principal?

Only with several important assumptions.

Increasingly perhaps the "rainy day plan" is best. Not perfect (no plan is), but "best." To heck with someone critical of the "rainy day plan" because a person who uses it /might/ actually die with money to her name.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

If your withdrawals are bounded by a fixed percentage of each year's total, then the order in which the returns occur doesn't matter.
In other words, suppose that one year your portfolio drops 20% and the next year it rises 30%. Then your balance at the end of that period will be exactly the same as it would have been if it rose 30% and then dropped 20%, even after withdrawals. Of course, your total withdrawals will be less in the first scenario than they would have been in the second, but that's to be expected.
The notion that an early run of bad luck might exhaust your portfolio goes away if you let the portfolio's value determine how much you're willing to withdraw.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Dec 11, 10:59 am, snipped-for-privacy@fractious.net wrote:

I think it is helpful to demonstrate that certain commonly recommended strategies are poor.

I think proportion-of-current-capital withdrawal scheme is better than the 4% rule based on initial capital, but as I have written before, the proportion should depend on life expectancy. If annually consuming 4% of capital is optimal for an 80-year-old widower, it almost certainly is not optimal for a healthy couple of 65-year-olds -- the money must last for both of them. There is no simple analytical solution to this problem. What's needed is a computer program that takes inputs such as life expectancy, expected mean returns and volatilities, and risk aversion and generates suggested spending/ investment plans based on these. Many retirees would not be able to use such a program directly, so a financial planner could run it for them. Elle would remind us that the inputs for such a program must be estimated and are not known. Because of this, one would want a plan that is robust to changes in assumptions, to the extent that is possible.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
snipped-for-privacy@aol.com wrote:

So what we'll then have is exquisitely analyzed output based on exquisitely guessed input. The beauty of the 4% rule is that is simple and basically works. Obviously, there are good reasons for adjusting it up or down. My current favorite approach is here: http://www.fpanet.org/journal/articles/2006_Issues/jfp0806-art6.cfm
It has the advantage that it is transparent -- you can see the assumptions you are making and how they are affecting the results. Computerized black boxes worry me because I spent years developing such.
-- Doug
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Nit pick: Not well-known.
I do continue to think the crude guidelines are useful with even the "not well-known" inputs, whether the guidelines derive from a program like B suggests above; the 4% rule; what Douglas suggests from the fpanet article below; living off dividends until D-day (that is, catastrophic medical costs) arrives; or maybe something else.

Indeed, the key words to me being "to the extent that is possible." Time and again I think that, once one realizes a change in assumptions is needed, it may be too late, and she or he may have spent too much of her or his portfolio to avoid financial ruin and living in some gosh-forsaken poorhouse of a nursing home. At least one may be enough out of sorts at this point not to care much, I guess.

I would say the one disadvantage of both proposals is the costs one might pay a financial planner for such services. Maybe these costs are trivial. Maybe not, especially given that, if one is not smart enough to pick and choose a suitable crude gage for withdrawals in retirement, gosh knows how much other stuff said financial planner can sell him/her. Just a little caveat, and not that all financial planners are nefarious.
Lately I wish financial planning academia would pursue a study of what educational approaches work best to keep people invested safely and with a view towards retirement and/or other needs. I wonder at the number of seemingly smart folks I know who take a momentary hit in their stock portfolios and then exit, hook line and sinker, for a few years, missing gains for the long term, and trusting gambles on, say, real estate over stocks.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
snipped-for-privacy@aol.com wrote:

About what? That the 4% rule basically works or my slightly nasty dig at simulations based on estimates of critical parameters such as date of death?

How do you know the results make sense? You need to have some external model to compare your results to.

My skepticism comes at two levels. First, if I were sitting across from a planner and his program spits out "you can spend 4.523% the first year and adjust up for the CPI (Urban) after that." How can I have any confidence?
Second, it ain't that hard a problem. All the important factors are unknown: date of death, market performance, inflation, taxes, health care costs... The benefit of any really detailed analysis is going to be swamped by uncertainty in the input data.
-- Doug
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I'm reminded of a marketing presentation, guy at the front is building a case for a certian high tech product. He stars with a China population of 8 Billion, and ends with a potential sales number. I raise my hand and tell him the world population wasn't even 6 billion at the time, there were likely only 800 million China residents. "well, these are all estimates" he replied, and stuck with his final numbers.
You are right, all of the input isn't known. But there are statistics. I don't know when I will die, but my insurance guy can tell me that in a pool of 10,000 45 year olds in my general health, half will live until X age. The market will not return 20%, nor 0% in the next twenty years. Is 10% right? I don't know, maybe the 8% I've suggested is closer. But isn't that what Monte Carlo and Statistical approaches are all about? Trying to quantify a range of outcomes? And the results are at least better than "I don't know"
JOE
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Suppose a planner said that his simulations had convinced him that for a single man, a reasonable withdrawal rate was X% initially, increased at the rate of inflation thereafter, and that his formula for X was 4 + 0.1*(age - 65), working out to 4% for someone retiring at 65, 5% for someone retiring at 75, etc. I would consider that more sensible than a 4% suggestion for everyone. If that means he suggests a 4.4% withdrawal rate to a 69yo, so what? Are you afraid of decimals?

I'd say that uncertain inputs make the problem harder and make simulations more valuable.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Douglas Johnson wrote:

It seems like you are arguing both sides, so correct me here where I'm wrong. On the one hand you argue for the 4% rule or the other to which you posted a link. These rules were derived from simulations. On the other it seems you are arguing that simulations have no value because the inputs (in this case) are unknown and vary wildly. Given this, why support your two rules? Or are you just worried about black box solutions? In which case why not use B's hypthetical program, but only if the author provides the assumptions?
-Will
william dot trice at ngc dot com
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

When did the 4% rule become something other than the above?
Elizabeth Richardson
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Elizabeth Richardson wrote:

I am not advocating a position, just answering the question:
BWS quote states 4% withdrawal each year, whether it's 1%/qtr, or 4% of prior 12/31 balance, it's a fluctuating number.
I believe Trinity used a $40K withdrawal on a million dollar sum, and each year that $40K was adjusted for inflation regardless of market return or current value of portfolio.
I personally think BWS quote is appealing to those who can break out the 'needed' income from the 'extra' that can be skipped in down years. When my wife and I discuss this, she asks for the number that can keep our lifestyle the same, no skipping anything in a down year. So we shall work a bit longer and go closer to my interpretation of Trinity, the ever increasing withdrawals. JOE www.joetaxpayer.com
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Tables 1 and 2 of the paper are devoted to withdrawals where the annual amount withdrawn is not adjusted for inflation. In these instances, it's truly a fixed income, based strictly on (for a withdrawal rate of say 4%) 4% of the /initial/ portfolio value and not subsequent values of the portfolio. So we see the line after the section titled "What About Inflation?" referring to purchasing power being halved after 25 years of 3% inflation. Table 3 shows the outcome where the annual withdrawal rate is adjusted for inflation and deflation.
The news is bad for those who adust for inflation and start with a high withdrawal rate. But staying at a withdrawal rate of 3-4% still suggests, if history is a guide, that a retiree will be fine.
I presume the much quoted 4% rule uses as its basis Table 3. But the WD rate does not depend on portfolio size; it depends on inflation, at least according to the study's authors. Whether that's good or not remains to be seen. Merriman's proposal (as quoted by Bread with Spam) may be superior.
At the moment, I lean towards the rainy day plan, due in particular to the uncertainty in health care costs. My heirs may be very wealthy, or I'll be in ruins and on Medicaid yada when I die.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Joe, do you expect your retirement portfolio to have average returns of at least 4% plus 3% inflation, or 7%? Were you planning on using the Trinity 4% of the beginning balance or the BWS 4% ever-adjusting balance?
As you know, we have pension income, what amounts to a fixed annuity in this conversation, but what we need/want over and above that is affected by this 4% discussion. But I don't think most of you pre-retirees realize that being on a "fixed" income, isn't really a fixed income. What we're doing is drawing on our IRA to the extent we need to, which is a different amount from month to month. Vanguard IRAs allow you to sell shares whenever, so I'm not stuck deciding in January how much I'll want/need in October (or even February). Periodically, but not on a rigid schedule, I monitor these withdrawals against the portfolio balance to ensure we're not living too high on the hog. So far, so good, as the market has been very kind. At some point, I'll look at starting social security - another factor in the equation.
Elizabeth Richardson
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Elizabeth Richardson wrote:

I was planning to go with my understanding of Trinity, 4% of initial balance, and add inflation each year. FWIW, I am 45, the missus is 51. She plans to retire in 4 years, and I would follow when we hit the number, about 2 years after. SS is not in the equation at all. So that's alot of wiggle room, I'd think. I am open to working longer, if needed, and if the market blows up, a black swan event, I have a few things I can go back to, to make up the difference. JOE
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Guys, do you realize how valuable is the older wife/younger husband scenario? The lower health costs of a married life versus a single/widowed life are just the tip of the iceberg.
Anyway, if I understand this discussion, a projected portfolio return of 7%, would, in fact, provide for your wife's request of keeping your lifestyle the same, would it not?
Elizabeth
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Elizabeth Richardson wrote:

If the portfolio returned precisely 7%, the 4% withdrawal and 3% inflation would work perfectly. The problem is risk/volatility/standard deviation. Trinity was done with data looking back for a return of 10%+ and concluded the 4% was good. That extra 3% covered volatility. Given that I've gone on record that I'd use 8% as a projected market return for the next decade or two, I may be deluding myself that 4% is the right number. That's why I figure if I don't count SS, that when it kicks in for both of us, I am really using a lower rate, maybe 3.5% or less.
FWIW, a fixed 7% return can allow an initial 4.8% withdrawal, and last 40 years. At 90, I doubt my wife will have the energy to powershop or go to her trainer.
JOE
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

You underestimate the power of compounding. It works in life, too! I racewalked a 5k in the heat in August, which nearly got me, and beat an 84 year old woman by only 3 minutes. Get in shape now, stay in shape. At 62, I get my heart rate up to 85% of maximum for an extended time at least once a week, in addition to several days with an hour or more of lower intensity aerobic work, add in some weights for an overall fitness routine. I'm hoping my coach will still be alive and working when I'm 90, 'cause I'm going to still want to train.
Anyway, thanks for the clarification on the extra allowance for volatility.
Elizabeth Richardson
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

BeanSmart.com is a site by and for consumers of financial services and advice. We are not affiliated with any of the banks, financial services or software manufacturers discussed here. All logos and trade names are the property of their respective owners.

Tax and financial advice you come across on this site is freely given by your peers and professionals on their own time and out of the kindness of their hearts. We can guarantee neither accuracy of such advice nor its applicability for your situation. Simply put, you are fully responsible for the results of using information from this site in real life situations.