Beware the 4% rule

[ERD50]

...

However the "floor rule" allows you to take no less than 95% of what you took the prior year. 95% of $43k = $40,850. You can that amount rather than $35k and still do well long-term at least as far as back-testing shows.

When I look at a FIRECALC run with that '95% spending rule', I see an awful lot of squiggly lines that dip down around what looks to be about 1/2 your original spend rate, and they appear to hang there a looooong time. Even when I start with 4.0% (not the 4.3% you mention). It's hard to follow any one line though, maybe there are better outputs forms for this data?

OTOH, a 3.5% inflation adjusted spend rate provides 100% success.

Personal preference of course, but I'd much rather try to stick to 3.5% on average, than to face years and years (of maybe the best years of my retired life) at something much less. It might even turn out to be a 'false alarm' - just a bad period with a good period to follow and make up for it. Sure, I'd adjust if I see some long term trend, but it's doubtful that I would need to.

edit - I gues FIREdeamer and I were thinking alike!

-ERD50

 

[BigNick]

There are many things that we fund based on one side being fixed and the other variable... gas for one comes to mind... if the price of gas goes up a lot like it did a year or so ago... and my salary does not... I am funding a variable expense based on a fixed income...

That depends on what you mean by "fixed". If gas is a substantial part of your expenditure, you can go quite some way towards compensating for it by having a higher weight of oil company shares, or energy mutual funds, or crude oil futures, in your portfolio. Hedging is not just for big corporations.

 

[DblDoc]

When I look at a FIRECALC run with that '95% spending rule', I see an awful lot of squiggly lines that dip down around what looks to be about 1/2 your original spend rate, and they appear to hang there a looooong time. Even when I start with 4.0% (not the 4.3% you mention). It's hard to follow any one line though, maybe there are better outputs forms for this data?

OTOH, a 3.5% inflation adjusted spend rate provides 100% success.

Personal preference of course, but I'd much rather try to stick to 3.5% on average, than to face years and years (of maybe the best years of my retired life) at something much less. It might even turn out to be a 'false alarm' - just a bad period with a good period to follow and make up for it. Sure, I'd adjust if I see some long term trend, but it's doubtful that I would need to.

edit - I gues FIREdeamer and I were thinking alike!

-ERD50

Did you run that as 4.3% with or without COLA? I think the Clyatt method is non-COLA'd.

DD

 

[ERD50]

Did you run that as 4.3% with or without COLA? I think the Clyatt method is non-COLA'd.

DD

I don't know if the Clyatt method is non-COLA or not, but to to be honest I don't care. No way in heck am I going to plan for a 45 year retirement based on a non-cola 4.3%.

I remember Grampa telling me about nickel lunches with a beer!

-ERD50

 

[HFWR]

I refuse to insist I have an open mind.

My mind was so open it fell out...

 

[FIREd]

Did you run that as 4.3% with or without COLA? I think the Clyatt method is non-COLA'd.

DD

I ran the 4.3% with no COLA.
$1,000,000 portfolio and $43,000 annual income for the first year, WR is 4.3% of remaining portfolio thereafter with "95% rule", 30-year retirement. I used a 50/50 portfolio (50% total market and 50% 5-year treasuries). FIREcalc shows that even the worse cycle would still leave you about $500,000 (2010 $$$) to pass on to your heirs after 30 or even 50 years in retirement. That's great (for your heirs) but, under some scenarios, you would have had to cut your expenses to half the original spending rate (for years at a time) in order to make it happen.

 

[walkinwood]

When I look at a FIRECALC run with that '95% spending rule', I see an awful lot of squiggly lines that dip down around what looks to be about 1/2 your original spend rate, and they appear to hang there a looooong time. Even when I start with 4.0% (not the 4.3% you mention). It's hard to follow any one line though, maybe there are better outputs forms for this data?

-ERD50

That is a definite downside of the 4%/95% rule and I think Bob Clyatt admitted it in one of his posts. Bob has advocated a flexible semi-ER lifestyle where you go out and earn some $s if things are really bad. (otoh, when things are really bad, its hard to get a job!)

I'd like to add two points to the 4%/95% rule that have not yet been called out in this thread :
1. The 4% to 4.3% withdrawal includes all investment expenses too. ie. Management fees.

2. The strategy leaves the purchasing power of your initial portfolio intact after 40 years some 95% of the time.

 

[Rich_by_the_Bay]

When I plug in the 4.3%/95% scenario in FIREcalc, I find an uncomfortably large number of cycles where the annual income would still have to be progressively cut from $43,000 down to roughly $20,000 at one point or another during a 30-year period. Does it sound right? By the way, I get somewhat similar results using my proposed 3%/90% scenario.

I don't know whether I am allowed to screen capture the FIREcalc results and post them here...

I never quite understood what's going on in FC under the hood in the Clyatt scenarios and haven't found it that useful.

But whatever results you get remember that the model seems to ignore value averaging during up years, which alone might mitigate what appears to be a lot of volatility otherwise. You can be sure that if we hit a great year here and there, I will either take a lower percent that year, or skim some off for future smoothing.

 

[haha]

That is a definite downside of the 4%/95% rule and I think Bob Clyatt admitted it in one of his posts. Bob has advocated a flexible semi-ER lifestyle where you go out and earn some $s if things are really bad. (otoh, when things are really bad, its hard to get a job!)

I'd like to add two points to the 4%/95% rule that have not yet been called out in this thread :
1. The 4% to 4.3% withdrawal includes all investment expenses too. ie. Management fees.

2. The strategy leaves the purchasing power of your initial portfolio intact after 40 years some 95% of the time.

How can this be true if some years you are down to spending half of what you started at? I think it must be hard to accept that there can be no magic. Either your portfolio earns your living, you earn your living (this is called non-retirement), or your liiving standard goes down, or you eventually go broke. Economics is harsh!

We have a tendency to grab onto whatever guru comes out with a new magic formula. It interest me that the group almost always immediately reject magic investing formulas, but hooks right onto magic withdrawal formulas. It's like sports gambling. Money management is surely important, but not near so important as figuring out a way to predict winners.

Ha

 

[FIREd]

I never quite understood what's going on in FC under the hood in the Clyatt scenarios and haven't found it that useful.

But whatever results you get remember that the model seems to ignore value averaging during up years, which alone might mitigate what appears to be a lot of volatility otherwise. You can be sure that if we hit a great year here and there, I will either take a lower percent that year, or skim some off for future smoothing.

That's true. In good years, I plan on skimming some off for future smoothing (FIRECalc probably assumes we spend your entire withdrawal each year). If we get a few good years after retiring, then 4.3% of portfolio value will give us plenty to skim off. That could help ensure that we don't have to cut back too drastically in later years. However, if we hit a rough patch right after retiring and we have to start cutting expenses right away, then I don't see how we could avoid some of the large income cuts predicted by FIREcalc (unless we have ample cash reserves outside of our retirement portfolio to help smooth our income during such early storm).

Perhaps, FIREcalc should have an upper and lower expense cap for the Clyatt scenario: In bad years, take the greater of 4.3% of portfolio or 95% of the previous year's spending and in good years, take the lesser of 4.3% of portfolio or your original spending rate (adjusted for inflation).

 

[DblDoc]

How can this be true if some years you are down to spending half of what you started at? I think it must be hard to accept that there can be no magic. Either your portfolio earns your living, you earn your living (this is called non-retirement), or your liiving standard goes down, or you eventually go broke. Economics is harsh!

We have a tendency to grab onto whatever guru comes out with a new magic formula. It interest me that the group almost always immediately reject magic investing formulas, but hooks right onto magic withdrawal formulas. It's like sports gambling. Money management is surely important, but not near so important as figuring out a way to predict winners.

Ha

But Ha as this thread points most of the posters don't think there is anything magical about the 4% SWR. It is a useful guideline that requires flexibility and some strategies to deal with the "what ifs", it is not a set and forget plan.

DD

 

[ejman]

the

Actually, I do not think that, although I certainly cannot rule it out. The US was a young capitalist country coming into its own with no huge baggage of entitlements, or government spending, and with no lower cost industrial competition.

What I based my statement on was since valuations (PE10) were at all time highs in 1999 and 2000- much higher than the top in 1929- there is no good reason that I can see to be certain or even put a high probability on the idea that the Depression and our wars represent the worse that might befall the stock market. This is not a popular idea, so I am not really prepared to defend it, but it is my working hypothesis. Also, remember that there are not many 30 year periods in the database, and fewer than five thirty year periods with no overlap.

We have the idea that government can pull us out of anything, but IMO it is at least equally possible that their reponses are just digging a deeper hole. My idea is that over an intermediate timespan, inflation is likely to be a bigger problem than deflation. But they are basically Scylla and Charybdis.

Ha

I see, sort of a protracted Japan scenario with an inflation kicker... Agree, not a pretty picture. Hope it doesn't come to pass and that the alternative is not worse.

 

[walkinwood]

How can this be true if some years you are down to spending half of what you started at? I think it must be hard to accept that there can be no magic. Either your portfolio earns your living, you earn your living (this is called non-retirement), or your liiving standard goes down, or you eventually go broke. Economics is harsh!

We have a tendency to grab onto whatever guru comes out with a new magic formula. It interest me that the group almost always immediately reject magic investing formulas, but hooks right onto magic withdrawal formulas. It's like sports gambling. Money management is surely important, but not near so important as figuring out a way to predict winners.

Ha

This thread started with discounting Sharpe because he talked about rigorous adherence to the 4% SWR rule. Why should the 4%/95% rule be any different. I think there is consensus here that all these are guidelines and you need to be flexible to make them work.

People have different and multiple ways of being flexible. Rich talks of one way. For us, ER'ing in our late 40's, going back to work is also a feasible option. We would rather have a few years of ER now, and then work for a while (part-time or even full-time) than continue working till I have a fool-proof plan. (if that is even possible).

 

[REWahoo]

A key element to a successful ER is bending over backwards to be flexible.

 

[FIREd]

A key element to a successful ER is bending over backwards to be flexible.

As long as it doesn't require bending over prison-style...

 

[haha]

This thread started with discounting Sharpe because he talked about rigorous adherence to the 4% SWR rule. Why should the 4%/95% rule be any different. I think there is consensus here that all these are guidelines and you need to be flexible to make them work.

I have no clue what the consensus here might be. One day it appears to be "I'm OK, Firecalc told me so". Next day it's "I'm OK, Bob Clyatt's book says so". Next day it's "Be flexible".

Well, These are not exactly mutually compatible. Of course every human should be flexible, and we may have to in many unattractive ways. But as a group it should be kind of obvious that our flexibility might be rather more limited than one might hope. Read some threads about all the things in the workplace that drive people out.

Anyway, I am not planning on following any of these plans, or am I trying to bust anyone else's plan. It just occurred to me on reading this

2. The strategy leaves the purchasing power of your initial portfolio intact after 40 years some 95% of the time.

that a spending pattern that might undergo a 50% cut for fairly long time, even assuming conditions that are already in the database, should really not be said to keep purchasing power intact, unless it is being claimed that those intermediate years don't count. I base this on what others have said regarding how low the spending can get under this modified plan. I haven't run the scenarios, as I do not plan on using this approach, any more than I plan on using the 4% approach. If I misunderstood what was said about the lean years, just forget my misguided comments.

And as to magical thinking, I think there is plenty of both magical thinking and guru worship too. I see it now, I've seen it before, and I will see it again. It's built into humans. Valid criticism such as the paper by Sharpe is quickly dismissed. Now there may be a lot wrong with his ideas, but I can't see how his basic tenet that it is chancy to fund an essentially fixed expense by liquidating a varying portfolio can possibly be wrong. "Flexibility" is one response to this criticism, in essence saying that the spending need is not fixed, or even bounded except by very loose limits. This may be true for people who are very heavily over-financed, but this is only sometimes the situation, and not a very interesting situation at that.

I am not overfunded, I have been retired for 25 years, I recently began to draw very modest SS, and I pay attention to all this stuff.

Ha

 

[REWahoo]

As long as it doesn't require bending over prison-style...

Then it's probably best you avoid "The Baguette Maneuver"...

 

[NW-Bound]

I agree on the rise and fall of empires... how long did it take for the Roman empire to fall??

It seems to me things are happening at a faster rate now. Why, the end of the cold war and the fall of the Berlin wall feel like just yesterday.

I wish I had kept an old issue of Byte Magazine (circa 1987?) talking about export control on PC hardware to the Sino-Soviet block. Back then, only the 8088 and 8086 (PC-XT class) were allowed for export. The 386's were not out at that time. They could not have the PC-AT class 80286, as these had higher computing power that could be used in weapon systems. That Byte article talked about how that export control was futile and nonsensical, as the Soviet and the Chinese simply bought through a 3rd party in another country.

Now, the Chinese are making all of our consumer electronics!

Also, even if we DID lose the top spot... we would still be second or third... not much farther down... there are a LOT of countries where things would be a lot worse...

I don't think the US will lose the top spot in my lifetime - wishful thinking perhaps. However, the other countries, particularly China, will be catching up fast. Together, they will present a more formidable competition, which will reduce our living standards.

The correlation between foreign equities and that of U.S. is getting higher, however. Adding other asset classes such as precious metals or commodities may help.

Yes. I hold stocks of foreign commodity producers in Brazil, Canada, and Australia. I also like US multi-national companies, or US companies that have a lot of business overseas.

BTW, following a really good market year, you may wish to withdraw less than the allotted 4.3% if you don't need it all, further enhancing the flexibility of this system - kind of a poor man's value cost averaging.

Or in a good year, one satisfies his pent-up demands, buying a new car, spending on home improvements, and indulging in foreign travel. And in bad years, he stays home to surf the Web, and to stain his house exterior with a 25-ft high ladder.

 

[dex]

I don't think the US will lose the top spot in my lifetime - wishful thinking perhaps. However, the other countries, particularly China, will be catching up fast. Together, they will present a more formidable competition, which will reduce our living standards.

It all depends upon the nature of the decline. I don't know of an empire's decline that was smooth. I'm guessing and economic decline crisis will take a large part of the world down with it. Then there will be a period of conflict - as the void left in the world by the USA is filled and fought over by others. So I don't see it as a smooth transition.

The decline of the USA is happing fast. Iraq, Afghanistan, Internet Bubble, Debt Bubble, unfunded entitlement programs all have speed things up. If we can kick this USA & state debt problems and entitlement programs out for about 40 years we ERs might be OK. The problem for us will be that high inflation and a VAT might come towards the end of that 40 years - a real financial risk in our older years.

 

[walkinwood]

...It just occurred to me on reading this that a spending pattern that might undergo a 50% cut for fairly long time, even assuming conditions that are already in the database, should really not be said to keep purchasing power intact, unless it is being claimed that those intermediate years don't count. I base this on what others have said regarding how low the spending can get under this modified plan. I haven't run the scenarios, as I do not plan on using this approach, any more than I plan on using the 4% approach. If I misunderstood what was said about the lean years, just forget my misguided comments.

I agree. Hadn't looked at it this way.

I am not overfunded, I have been retired for 25 years, I recently began to draw very modest SS, and I pay attention to all this stuff.

Haha, 25 years is very impressive. Congratulations! Would you share your strategy for withdrawals with us?

 

[DblDoc]

And as to magical thinking, I think there is plenty of both magical thinking and guru worship too. I see it now, I've seen it before, and I will see it again. It's built into humans. Valid criticism such as the paper by Sharpe is quickly dismissed. Now there may be a lot wrong with his ideas, but I can't see how his basic tenet that it is chancy to fund an essentially fixed expense by liquidating a varying portfolio can possibly be wrong. "Flexibility" is one response to this criticism, in essence saying that the spending need is not fixed, or even bounded except by very loose limits. This may be true for people who are very heavily over-financed, but this is only sometimes the situation, and not a very interesting situation at that.

I am not overfunded, I have been retired for 25 years, I recently began to draw very modest SS, and I pay attention to all this stuff.

Ha

Sharpe raises some excellent points and has some valid criticisms:

1) Funding a nonvolatile spending plan with a volatile investment strategy IS risky. No doubt about it.

Whats the alternative? He proposes several strategies (such as stripped TIPs) none of which are currently available. Maybe some investment company will create and market them. Hmm ... interesting intellectual exercise but nothing actionable.

2) It is inefficient - you have surpluses, could end up leaving money on the table and it has costs to implement.

Well his surplus is my emergency fund for when the roof blows off or I have a medical problem. It would be nice to have a spending plan that was nonvolatile but not going to happen in reality. S@#t happens. I want a surplus.

If I have money left over at the end then good for my heirs or charity of my choice - thats a bonus.

As to the costs to implement again he has no actionable alternative.

The bottom line I guess is you rolls the dice and takes your chances. Give yourself a margin of error, some cushion, pay attention, be flexible and adaptable and enjoy the ride.

DD

 

[chinaco]

You have $1mm on Jan 1, 2000. You take $43k in income, leaving $957k.

Bad year ensues, market down 15%, leaving you with $813k by year's end. According to the basic rule, you would end up taking 4.3% of that $813k for the upcoming year's income = $35k. That's a huge hit.

However the "floor rule" allows you to take no less than 95% of what you took the prior year. 95% of $43k = $40,850. You can that amount rather than $35k and still do well long-term at least as far as back-testing shows.

BTW, following a really good market year, you may wish to withdraw less than the allotted 4.3% if you don't need it all, further enhancing the flexibility of this system - kind of a poor man's value cost averaging.

Hope that helps.

Thanks. It sounds like Clyatt's approach factors in the uncertainty of a sustained downturn by providing a tested guideline for systematically reducing the WR amount (to reduce the risk of prematurely exhausting the portfolio).

Here is an interesting article about research on the matter of sequence risk.

The Dynamic Implications of Sequence Risk on a Distribution Portfolio


It basically says what we all know... one needs to revisit the WR% periodically to see if it still makes sense in light of market realities (as time progresses).

My interpretation: one beings withdrawal with a 4% rate (to set the withdrawal amount with an inflation increase yearly) using a 60/40 mix and it is early in retirement of a 30 year time horizon ) and there is a Bull Market. A Bear Market occurs (with a steep correction) and one continues with the original plan, but the WR% is now about 7% of the portfolio.... this is likely to be ok (assuming the bear is short lived). If, however, the bear persists and the portfolio is significantly reduce... a sustained 7% WR will exhaust the portfolio (since there is still a long time horizon).

 

[teejayevans]

My interpretation: one beings withdrawal with a 4% rate (to set the withdrawal amount with an inflation increase yearly) using a 60/40 mix and it is early in retirement of a 30 year time horizon ) and there is a Bull Market. A Bear Market occurs (with a steep correction) and one continues with the original plan, but the WR% is now about 7% of the portfolio.... this is likely to be ok (assuming the bear is short lived). If, however, the bear persists and the portfolio is significantly reduce... a sustained 7% WR will exhaust the portfolio (since there is still a long time horizon).

Which is why many of us believe follow a buckets method of some type:
stocks, bonds, and some significant amount of safe/cash money to ride
out bear markets.
TJ

 

[HFWR]

I intend to use 4% as a "guideline", as well as use the bucket approach. But then, unless I live to a very ripe old age, my retirement will not likely be 30-40 years; more likely it'll be 20-25.

 

[FIRE'd@51]

My complaint about Sharpe's approach is that he seems to treat 30 years as a perfectly fixed number. Any amount left at the end of 30 years becomes a "surplus", and a source of inefficiency to be squeezed out. But, if he eliminates the possibility of a surplus, then it seems to me he has 100% chance of running out of money if he lives 31 years.

I don't think Sharpe wants to "squeeze out" all surpluses, just that which he refers to as the "wasted surplus". The wasted surplus is that which results from one cutting back on his withdrawal rate only to find out later that it was an unnecessary cutback, i.e. his glidepath was stilll among the 95% successful ones, only he didn't know that (nor could he) when he made his cutback. The two-bucket approach with stripped-TIPS in one bucket and the PV of the maturity payment invested in risky assets in the other, eliminates all possibility of a wasted surplus, since the first bucket will take care of spending needs for 30 years and, by definition, will never require a cutback. The surplus will be the second bucket, which will either be passed on to heirs or used to fund more retirement years. The first bucket matches income to spending regardless of market volatility. The short-term volatility of the risky assets in the second bucket is of less concern since the second bucket need not be touched for 30 years.