New Improved 4% SWR

[Midpack]

While there is no foolproof distribution strategy, why isn't this better than the 4% SWR? The 2nd & 3rd columns are simply the straight 4% rule assuming constant returns and inflation - not real, only used to generate the last column. So why not use the equivalent percentage rates applied to the actual remaining "nest egg" balance? It will automatically course correct somewhat whereas 4% SWR as written does not. Needless to say you can use different longevity, returns and inflation to generate your guiding chart/rate. I realize I'm not the first to think of this...

 

[OAG]

This is very close to the method used by the IRS for Traditional IRA RMD's. Of course theirs start at age 70.6 and run out to age 114 (optimistic IRS). It is age based but you could use that one and just adjust the first year to your current age and it would run for 44 years and ignore inflation (CPI). Runs a balance of any amount down to 0. It does start at 3.75% but increases each year at a realistic clip. IRS publication 590 contains the charts.

 

[unclemick]

And in my case courtesy my friends at the IRS - an offer I can't refuse.

heh heh heh - And they said you can pssst Wellesley all you want - after taxes. What a bunch of sweeties!

age 65, ballpark 78% of portfolio is trad IRA.

 

[Rich_by_the_Bay]

This is what Clyatt recommends, what Merriman recommends, and what many have written about. It is also what I plan to do.

You can also set the floor for any given year at 95% of the previous years allotment, and still come out fine with a diverse portfoliio as back-tested by Clyatt. One or two of the studies say that if you choose this "percent of assets" method, you can safely use 4.2 to 4.5% rather than the conventional 4%.

 

[walkinwood]

One or two of the studies say that if you choose this "percent of assets" method, you can safely use 4.2 to 4.5% rather than the conventional 4%.

Just remember to count the expense ratio of your funds as part of the withdrawal percentage.

 

[Rich_by_the_Bay]

Just remember to count the expense ratio of your funds as part of the withdrawal percentage.

I don't quite see it that way. The ER simply gets reflectedd in the value of the funds, upon which the withdrawal amount is calculated. A high ER fund will have a lower end-of-year value than an otherwise identical low ER fund.

But, hey, do whatever works for you.

 

[samclem]

I also plan to take 4% of the annual remaining balance. As Rich points out, Clyatt offers a slightly amended version (higher of either a stated percentage of the year-end balance OR 95% of the amount you took last year, whichever is greater). FIRECalc allows you to select and model this withdrawal approach. This appeals to many retirees who could not handle the higher volatility of a strict "X% of year end total" approach.

Of course, with the simple "X% of year-end total" withdrawal, it's impossible to entirely run out of money. The failure mode is different--an ever-declining standard of living. So, it still can't be a "set and forget" item--inflation still needs to be tracked and charted against the value of the nest egg over time. If the nest egg dips below the inflation-adjusted line, it will result in a decreased standard of living even if the dollar amounts are climbing--which might signal that a decrease in withdrawals is needed.

 

[2B]

This is what Clyatt recommends, what Merriman recommends, and what many have written about. It is also what I plan to do.

You can also set the floor for any given year at 95% of the previous years allotment, and still come out fine with a diverse portfoliio as back-tested by Clyatt. One or two of the studies say that if you choose this "percent of assets" method, you can safely use 4.2 to 4.5% rather than the conventional 4%.

When I plug this into FIRECalc, it generates a much lower SWR than the "regular" SWR. Intuitively, I would think the initial SWR would be higher if you were willing to take a withdrawl reduction if the market tanks. When I change the 95%, it doesn't change the output. I think there's something wrong with how FIRECalc is doing this calculation.

My current plan is to take withdrawls based on Bernicke. This greatly exceeds my basic living expenses. What's left over will be for luxuries and travel. If inflation ramps up where the constant Bernicke withdrawl rate up to age 70 gets closer to my basic living expenses, the luxuries and travel go away.

 

[Rich_by_the_Bay]

Yeah, FC handles the Clyatt plan oddly in some models. For one thing, it seems to simply ignore anything other than a straight all-out retirement - I plan semiretirement for 4-5 years, for example, which doesn't get factored in correctly.

But with the % of total plan, I feel less reliant on careful modeling. Simply reign in the lifestyle as needed since you will have plenty of warning. Won't stop the ship from sinking but it will give you plenty of time to select your lifeboat.

 

[donheff]

Splain it to me please. The way I read Midpack's post is he ran a standard 4% + inflation of the initial nest egg SWR on a hypothetical $1M portfolio and hypothetical inflation rate using a 30 year horizon. Then instead of running the standard SWR on initial value for real, he will apply the actual %s generated by the hypothetical run to the real portfolio balance each year. For example, in his 30 year scenario he would take out 4% in year 1, 4.5% in year 5, 5.5% in year 10, 9.9% in year 20, and 100% (closing out the portfolio) in year 30. That assures that all funds are used up in a fixed period of time. But you darn well better die on schedule. Did I misunderstand the OP?

Others describe this as what Clyatt proposes but I understood Clyatt to take a standard 4%+inflation of initial nest egg SWR and temper it with reductions (of up to 5% of the last year's withdrawal) in down years.

Naturally 2B tosses in Bernicke to distract me so now I have to go look that up to refresh my flagging memory.

Edit: nevermind the last sentence, I remembered Bernicke -- the cheaper when you are older guy.

 

[Rich_by_the_Bay]

Splain it to me please. The way I read Midpack's post is he ran a standard 4% + inflation of the initial nest egg SWR on a hypothetical $1M portfolio and hypothetical inflation rate using a 30 year horizon. Then instead of running the standard SWR on initial value for real, he will apply the actual %s generated by the hypothetical run to the real portfolio balance each year. For example, in his 30 year scenario he would take out 4% in year 1, 4.5% in year 5, 5.5% in year 10, 9.9% in year 20, and 100% (closing out the portfolio) in year 30. That assures that all funds are used up in a fixed period of time. But you darn well better die on schedule. Did I misunderstand the OP?

I'll let Midpack clarify what he meant but I understood the two situations to be a) the standard SWR at 4% corrected annually for inflation, versus b) 4% of total balance withdrawn annually. I didn't think he was trying to backfill the results of a) into b).

BTW, Merriman suggests kind of a hybrid, but offers no data or testing to support it: a 4% of total annually, kicked up a notch to 4.5% if/when your nestegg grows to 1.25 its original value, then 5% at 1.5x original, etc. (or something along those lines). Seems too arbitrary to me, but there you have it.

I like Bob's plan, am reassured that he modeled it and back-tested, and think we could handle a 5% reduction in income during hard times (in fact we'd probably cut back on our own if it weren't built in to the plan).

 

[Midpack]

Splain it to me please. The way I read Midpack's post is he ran a standard 4% + inflation of the initial nest egg SWR on a hypothetical $1M portfolio and hypothetical inflation rate using a 30 year horizon. Then instead of running the standard SWR on initial value for real, he will apply the actual %s generated by the hypothetical run to the real portfolio balance each year. For example, in his 30 year scenario he would take out 4% in year 1, 4.5% in year 5, 5.5% in year 10, 9.9% in year 20, and 100% (closing out the portfolio) in year 30. That assures that all funds are used up in a fixed period of time. But you darn well better die on schedule. Did I misunderstand the OP?

You read my post correctly, I am asking about taking a fixed but increasing % of total portfolio/assets each year. You may notice that I said you can change the longevity, returns, inflation - I understand the risk of not 'dying on schedule.' However, that risk exists with the 4% SWR (and the Clyatt variation) as well, so my proposal is an improvement in that it course corrects at least somewhat. I should probably have also mentioned that we have no heirs so we'd be happy to die broke - although odds are there will be something left, which would go to our favorite charities.
And actually taking 4% of remaining balance (not what 4% SWR is) would seem to a) ensure a large portfolio when the owner expires and b) a significantly declining standard of living over the term - not at all desirable for us.

 

[donheff]

You read my post correctly, I am asking about taking a fixed but increasing % of total portfolio/assets each year.

If you set a fairly high horizon, this approach seems OK. I still prefer the Clyatt approach. I would hope to live well on a bit less than my withdrawal and would bank the remainder in a "spendthrift" account to tap in down years. But, unlike you, I would like to leave an estate. If you go with this approach I would still recommend taking steps to set aside a cushion to take you further if health care changes end up kicking longevity up a notch further than we expect in the coming decades.

 

[Nords]

Others describe this as what Clyatt proposes but I understood Clyatt to take a standard 4%+inflation of initial nest egg SWR and temper it with reductions (of up to 5% of the last year's withdrawal) in down years.

I had this mistaken impression too.

The Trinity study starts at 4% of the initial portfolio, raises it each year for inflation, and never looks back. Of course it only looks 30 years ahead, too.

Bob's method resets every year to 4% of the portfolio's 1 Jan balance. If that's too much of a cutback in living standard then the ER takes 95% of last year's withdrawal. This means the actual percentage withdrawal that year may be higher than 4%. The other side of Bob's method is that for prolonged downturns an ER is expected to fill in the gaps with part-time work.

Here's some of Bob's earlier posts on the subject:
http://www.early-retirement.org/forums/showpost.php?p=487542&postcount=30
http://www.early-retirement.org/forums/showpost.php?p=467857&postcount=37
http://www.early-retirement.org/forums/showpost.php?p=351092&postcount=2

 

[donheff]

I had this mistaken impression too.

The Trinity study starts at 4% of the initial portfolio, raises it each year for inflation, and never looks back. Of course it only looks 30 years ahead, too.

Bob's method resets every year to 4% of the portfolio's 1 Jan balance. If that's too much of a cutback in living standard then the ER takes 95% of last year's withdrawal.

Thanks for the clarification. That sounds sensible to me. In good years I could still sock away a cushion that would be available in the bad years so I could cut back to even less than 95% (possible to zero) letting the primary portfolio recover more quickly.

 

[old woman]

Knowing how long you will live is the big thing. My mom for example is 82 and has about 200K in the bank in CDs. It just dawned on her that she can spend the money. She moved yesterday selling her house. She will get 1,250 a month from that and not have to pay utilities anymore since she moved in with my brother. She already gets about 30K a year from pensions and interest and SS. Her only bills will be room and board, nobody knows what he will charge her yet, he didn't say and nobody asked. But he and his wife wanted her to come and bought a big house so she had her own suite, they will pay for the yard people and the cleaning lady and buy all the food and pay all the bills so any amount he wants from her will probably be a good deal for her.
Her medical insurance pays 100% with no copays except $5 per prescription and she has car insurance but no other bills. She could take 10K a year from savings the rest of her life or about 20 years if she runs out at 102 she can move to low income housing or go on Medicaid in a nursing home but she will still have about 35K a year in income. She doesn't need to really budget anymore. Yesterday she paid the movers with a generous tip, paid for a 8 X 11 wool rug and bought a wireless router so her computer could be networked with theirs without once whining about how much everything cost. She may live 20 or more years because her mom lived to 98 and her grandma lived to 97.

 

[2B]

She could take 10K a year from savings the rest of her life or about 20 years if she runs out at 102 she can move to low income housing or go on Medicaid in a nursing home but she will still have about 35K a year in income.

The real wildcard for her is nursing/assisted living care. In Texas a nice place runs about $60K/year. Any nursing home is terrible but Medicaid homes (most nice ones don't accept Medicaid patients and will evict those that can't pay) are really, really bad. Medicaid doesn't cover assisted living at all.

I suspect your brother won't charge her or will charge very little. Your mother probably won't be able to spend all of the money she has coming in. Unless the kids are ready to pony up, you should all still be concerned that she has at least $150K sitting around as backup.

 

[Gotadimple]

BTW, Merriman suggests kind of a hybrid, but offers no data or testing to support it: a 4% of total annually, kicked up a notch to 4.5% if/when your nestegg grows to 1.25 its original value, then 5% at 1.5x original, etc. (or something along those lines). Seems too arbitrary to me, but there you have it.
quote]

Rich:

Merriman's flexible withdrawal plan is similar to Gummy's plan. Gummy says if you made more in your portfolio than your withdrawal amount last year, then you can take more for your good money management. If you made less, think about a tighter budget.

Merriman says be more mechanical in the process.
Start with 4% of the current portfolio balance for your year 1 withdrawal. In each subsequent year for each 1% gain or loss in portfolio value you can withdraw an additional (or reduce by) 2 basis points (.02%).

So the example he uses most frequently is if your portfolio year end balance is 25% higher than last year's balance, then the withdrawal rate moves 50 basis points to 4.5%.

On the flip side, however, if you lost 21% of your portfolio balance, your withdrawal should be reduced .42%, or in year 2 that would be 3.58%.

I think one needs to take a look at these schemes based on the portfolio allocation. Both Clyatt and Merriman construct portfolios with value tilts which generally allows increasing amounts of withdrawals (in terms of dollar amount withdrawn) because of lower volatility and dividend yields (in a normal market!!). If one's portfolio didn't have the value tilt, I'm not so sure a flexible withdrawal plan would have lower volatility.

However, in this environment, I plan to use the Gummy method -- that is I'm taking what I need when I rebalance in January (which should be about 4%), then we'll see from there.

-- Rita

 

[W2R]

How about this method?

***********************************************
Q. At the beginning of the year, has the portfolio grown by more than inflation?

1. If answer is yes, then withdraw 4% of the new portfolio size, less any left unspent from last year's withdrawal.

2. If answer is no, then withdraw 3% of the new portfolio size, less any left unspent from last year's withdrawal.
***********************************************

I am thinking of something along these lines, except lowing the withdrawals in both cases by 1% due to my conservative (45:55) AA and long lived ancestors.

I am not sure where I got this method but it sounds appealing. It should ensure that the portfolio grows over time, or at least does not shrink. As the portfolio grows, so will the withdrawal allowance.

 

[unclemick]

Check your 12/31 balance. Wet your finger and check your navel while squinting one eye into the distance/or crystal ball if you have one.

Take a number(in $) between your SEC yield(the ole pssst Wellesley trick) and 5% variable(do ya feel lucky) and deduct to MM (your expenses for one year).

I did this in Feb - 2008 so I had all of Jan to whine and prognosticate the future.

Note that with certain portfolio stock/bond mixes things can flip - 5% variable being the lower no.

heh heh heh - with all the gloom and doom I'm looking forward to my lower no.(SEC yield) right now. We'll see.

Target 2015 = 3.81% this weekend.

 

[W2R]

I like your method, UncleMick! Especially the "pssst... Wellesley" SEC yield part.

 

[walkinwood]

I don't quite see it that way. The ER simply gets reflectedd in the value of the funds, upon which the withdrawal amount is calculated. A high ER fund will have a lower end-of-year value than an otherwise identical low ER fund.

But, hey, do whatever works for you.

Rich,
I'm simply re-stating what Bob Clyatt says in his book. He tested the 4%/95% scheme using index values - not real funds. So, he says you need to include your fund expenses in the amount you withdraw.

 

[Rich_by_the_Bay]

Rich,
I'm simply re-stating what Bob Clyatt says in his book. He tested the 4%/95% scheme using index values - not real funds. So, he says you need to include your fund expenses in the amount you withdraw.

Gotcha. Semantics - my planning routine looks at returns after expenses for mutual funds including index funds. So if I think the total stock index fund will earn 7% annually, that would be after expenses, lagging a bit off the actual index holdings themselves.

I could see where back-testing would use actual index returns since expenses vary so much among funds.