When Does True SWR Begin?

[marko]

This is an academic question, ok?

My understanding of the traditional SWR 'rule' is that on retirement, one takes one's portfolio balance and then withdraws the same X% each year plus inflation for the next 25-30 years.

But for many people, 'retirement' doesn't start on the day after you leave work; there could be a 10 year transition with varying need for funds.

For example:
One spouse could keep working minimizing the withdrawal
Both or one spouse could continue to work part-time, again minimizing the withdrawal
A long-ish severance or other short term income stream might do the same

Meanwhile over that period of transition the portfolio could grow considerably. The last 10 years many portfolios doubled.

So when do you put that stake in the ground and say: "Ok, now we start our real X% SWR from THIS portfolio balance?

I bring this up because for years I thought this is what I was doing when in reality I learned from another thread (thanks audreyh1) that I was actually using a 'percent of portfolio' approach.

I wonder how many people actually stick to that firm "X% plus inflation" from a portfolio calculation made 8 or 9 years ago.

[Dtail]

Just responding to one aspect.
I don't think there are many folks on this forum who actually use the "X plus inflation" as their WR% for each year.
I believe many use this formula in one form or another pre-retirement to gauge whether they are financially ready for retirement. Even post retirement, some continue to use Firecalc, etc to see if the plan is still working out.
In real life, one's spending doesn't truly work in this manner.
Thus you see many folks spending more when the market is up and vice versa.

[pb4uski]

Quote:

Originally Posted by marko .... I wonder how many people actually stick to that firm "X% plus inflation" from a portfolio calculation made 8 or 9 years ago.

From my reading of posts on this forum, relatively few.

On the first part, that is why I prefer looking at one's "ultimate" WR rather than initial WR. Our initial WR was about 4% when we retired at 56... perhaps even a little north of 4%. However, once my pension started and SS starts our WR will be much lower... less than 2%.

One way to crudely estimate your ultimate WR is to calculate your ultimate gap (spending less pensions and SS or what you will need to withdraw once pensions and SS are online). Then calculate your gap for each year between now and when your ultimate gap starts and sum them up. Subtract that amount from your nestegg... as if you put those funds off to the side and used them in the interim. Then divided your ultimate gap by your adjusted nestegg to get a crude estimate of what your ultimate WR will be.

[audreyh1]

My impression from this forum is that very few use that traditional % of starting portfolio + inflation adjustment withdrawal method. REWahoo is the only one I can think of.

Most seem to use some variant based on current portfolio value, whether it is withdrawing a target % at the start or during the year, or calculating it after the fact by checking spending against the portfolio value to make sure it didn’t exceed what they feel is a safe WR.

Quite a few are living off dividends and interest generated by their stocks/bonds and otherwise don’t touch “principal”.

You could do a poll like this:

Which withdrawal method do you use:

• Traditional initial % of portfolio plus annual inflation adjustment
• % Remaining Portfolio (i.e. % of current portfolio value)
• Clyatt’s 95% Rule
• VWR - Variable Withdrawal Rate
• Withdraw interest and dividend income only for spending
• Check spending against portfolio value annually to determine WR
• Bought Annuity
• Other method - specify

There are some other withdrawal methods whose names I can’t remember at the moment.

[Freedom56]

Quote:

Originally Posted by marko This is an academic question, ok? My understanding of the traditional SWR 'rule' is that on retirement, one takes one's portfolio balance and then withdraws the same X% each year plus inflation for the next 25-30 years. But for many people, 'retirement' doesn't start on the day after you leave work; there could be a 10 year transition with varying need for funds. For example: One spouse could keep working minimizing the withdrawal Both or one spouse could continue to work part-time, again minimizing the withdrawal A long-ish severance or other short term income stream might do the same Meanwhile over that period of transition the portfolio could grow considerably. The last 10 years many portfolios doubled. So when do you put that stake in the ground and say: "Ok, now we start our real X% SWR from THIS portfolio balance? I bring this up because for years I thought this is what I was doing when in reality I learned from another thread (thanks audreyh1) that I was actually using a 'percent of portfolio' approach. I wonder how many people actually stick to that firm "X% plus inflation" from a portfolio calculation made 8 or 9 years ago.

It depends on ones income streams versus expenses. In our case our pension covers all our expenses except the income tax on our taxable investment and savings accounts. The excess from our taxable accounts after taxes is just re-invested. When we planned for early retirement, I didn't look at FIRECALC or any SWR rule. We knew that with zero debt, the relatively generous pension, the amount of capital we had, and the large inheritance that would be coming to us in the future, money would not be an issue at all.

Prior to retirement, we planned our income streams as follows:

Age 55-62 - Pension
Age 62-70.5 Pension, SS, Secondary Pension
Age 70.5 and on Pension, SS, Secondary Pension, Tax deferred Accounts, Taxable Accounts

During my working days, I was surprised at how many of my executive peers did not contribute to a 401K. Now I see why. With a generous executive pension after retirement, it wasn't a necessity. In our case, it is a back-up slush fund to our taxable accounts. As far as the future inheritance goes, we will probably set up a charitable trust with the funds after paying the income taxes due.

[marko]

OP here.
The responses so far have been what I'd expect.

I do find it interesting that there is so much discussion, clarification and hand-wringing about the 'traditional SWR rule' (initial X% plus inflation) only to realize that few, if any actually use this method.

Could the overall discussions become as academic as my post?

[REWahoo]

I consider the 'traditional SWR rule' to be the point from which everyone (including me) deviates.

[TwoByFour]

Quote:

Originally Posted by marko OP here. The responses so far have been what I'd expect. I do find it interesting that there is so much discussion, clarification and hand-wringing about the 'traditional SWR rule' (initial X% plus inflation) only to realize that few, if any actually use this method. Could the overall discussions become as academic as my post?

Part of what your are pointing out is terminology - SWR is a theoretical maximum withdrawal rate. Someone's actual withdrawal rate is not necessarily either the same rate (4%) or the same technique (inflation adjusted constant dollar amount) as SWR. I think the SWR discussion is academic, but the discussion on actual WR is not.

[eytonxav]

Need to call it FSWR or ASWR for floating or adjustable.

[ERD50]

Quote:

Originally Posted by marko OP here. The responses so far have been what I'd expect. I do find it interesting that there is so much discussion, clarification and hand-wringing about the 'traditional SWR rule' (initial X% plus inflation) only to realize that few, if any actually use this method. Could the overall discussions become as academic as my post?

It's no different than any other general number. It doesn't apply to every case, or every moment in time.

Do you know any households with 2.58 people?

If someone tells you they made a 45 mile trip in one hour, do you assume they drove a constant 45 mph, never stopped, never slowed down, never sped up to pass someone?

etc, etc, etc.

That doesn't mean it isn't useful, it is. But you need to apply the concept to your situation.

-ERD50

[marko]

Quote:

Originally Posted by TwoByFour Part of what your are pointing out is terminology - SWR is a theoretical maximum withdrawal rate. Someone's actual withdrawal rate is not necessarily either the same rate (4%) or the same technique (inflation adjusted constant dollar amount) as SWR. I think the SWR discussion is academic, but the discussion on actual WR is not.

Ok.

My musing is about at what point does one declare themselves 'retired' whereupon they'd say: "Ok, now what is X% of our portfolio that we'll withdraw every year and adjust for inflation for the next 20 years; that number starts today!" even though they might have been withdrawing smaller amounts for years.

It seems that the difference is more about determining what that number is versus actually putting the 'X% plus inflation' to actual practice.

[audreyh1]

Quote:

Originally Posted by marko OP here. The responses so far have been what I'd expect. I do find it interesting that there is so much discussion, clarification and hand-wringing about the 'traditional SWR rule' (initial X% plus inflation) only to realize that few, if any actually use this method. Could the overall discussions become as academic as my post?

Well, some folks have already pointed out that that metric is really used for evaluating whether one have enough accumulated to retire comfortably.

So from that aspect it's useful.

But yes, the initial portfolio value plus inflation adjustment is quickly abandoned after retirement. Folks pretty much think in terms of their current portfolio value or yield.

[TwoByFour]

Quote:

Originally Posted by marko Ok. My musing is about at what point does one declare themselves 'retired' whereupon they'd say: "Ok, now what is X% of our portfolio that we'll withdraw every year and adjust for inflation for the next 20 years; that number starts today!" even though they might have been withdrawing smaller amounts for years. It seems that the difference is more about determining what that number is versus actually putting the 'X% plus inflation' to actual practice.

I think this has been discussed in another thread around here - you can recalculate a withdrawal rate every year of your retirement if you want. It is independent of whatever spending you already did in the past. The important thing is not when you "retired", it is what are your total assets right now and how many years do I want that money to last?

[Alan]

Quote:

Originally Posted by audreyh1 Well, some folks have already pointed out that that metric is really used for evaluating whether one have enough accumulated to retire comfortably. So from that aspect it's useful. But yes, the initial portfolio value plus inflation adjustment is quickly abandoned after retirement. Folks pretty much think in terms of their current portfolio value or yield.

I agree, plus for many there are future income streams to consider, allowing higher % withdrawals in the early years. We retired at age 55 and at age 62 another private pension started coming in. Over the next 7 years we will both start drawing US and UK SS (4 more secure income streams).

[audreyh1]

Quote:

Originally Posted by marko Ok. My musing is about at what point does one declare themselves 'retired' whereupon they'd say: "Ok, now what is X% of our portfolio that we'll withdraw every year and adjust for inflation for the next 20 years; that number starts today!" even though they might have been withdrawing smaller amounts for years. It seems that the difference is more about determining what that number is versus actually putting the 'X% plus inflation' to actual practice.

That's your other question.

Well, the traditional method and its models assume that you retire on Dec 31, withdraw your say 4% annual income on Jan 1, rebalance. The next Jan 1, you withdraw your prior year income adjusted for inflation, rebalance. Repeat annually.

So it doesn't take into account smaller withdraws early in retirement or the concept of someone not starting withdrawals at the beginning. The models all run on an end/beginning of year basis.

Someone declares themselves retired when they are no longer working (not including hobby income). It has nothing to do with when someone starts drawing on their retirement portfolio.

At any moment in time someone could decide to start a given withdrawal method. They probably should run the models again, to see what the parameters are - income, survival rate, etc., just like they did when they were planning to retire.

Occasionally there is talk about "Retiring Over and Over Again" - that is really a oddball name for reseting the Traditional method each year as their portfolio rises, but then sticking with the Traditional method thereafter once they hit a down year. Clearly, they've been retired for a while already, but they are pushing the Traditional method into an area of higher probability of failure in order to increase their withdrawals in the early years. Less chance of a large terminal portfolio that way, you just hope you don't push it into one of the handful of cases that fail. But if you picked a withdrawal rate that had 100% chance of success historically, you'd probably be quite safe.

I suppose this variant of the Traditional method should be added to any poll about withdrawal methods.

[youbet]

Quote:

Originally Posted by DFW_M5 Need to call it FSWR or ASWR for floating or adjustable.

No..... That would be an oxymoron. SWR = the max annual (+inflation) amount that can be made for X years which, using historical data, would not have depleted the FIRE portfolio more than 5% of the time. There is absolutely no "floating" or "adjustability" involved.

Then there is WR, a measure of what you're actually doing. Many folks adjust that through time. Many of the comments I see here on the forum that mention SWR really mean WR.

FireCalc makes no pretense that it's testing what you actually do. It backtests a constant WR (+inflation) vs historical investment performance and inflation data to check for survivability. When you do something different, obviously the FireCalc backtesting does not apply to you.

Another area where I see folks mis-using the FireCalc backtesting algorithm and data is portfolio makeup. For example, if you tell FireCalc you are 70% S and P 500 and 25% 5 year treasuries and, in fact, your FIRE portfolio varies from this to any significant degree, the backtesting will no longer be accurate for you.

[youbet]

Quote:

Originally Posted by Alan I agree, plus for many there are future income streams to consider, allowing higher % withdrawals in the early years. We retired at age 55 and at age 62 another private pension started coming in. Over the next 7 years we will both start drawing US and UK SS (4 more secure income streams).

Yes. But more broadly, ALL variations from what you tell FireCalc to include in the backtesting (and which FireCalc is set up to include such as SS, pensions, future portfolio changes, etc.) must be considered as reducing the accuracy of the backtest.

[youbet]

Quote:

Originally Posted by marko Ok. My musing is about at what point does one declare themselves 'retired' whereupon they'd say: "Ok, now what is X% of our portfolio that we'll withdraw every year and adjust for inflation for the next 20 years; that number starts today!" even though they might have been withdrawing smaller amounts for years. It seems that the difference is more about determining what that number is versus actually putting the 'X% plus inflation' to actual practice.

Your issues seem to be:

1. The meaning of the word "retirement." FireCalc means the time at which you begin living the financial parameters you inputted.

2. Since your actual life varies substantially from the simple FireCalc backtesting algorithm, FireCalc clearly doesn't apply to you other than being a hypothetical point of reference to mull over. Try not to let that confuse you.

I find downloading the FireCalc spreadsheets and looking at the numbers and formulas for my projections to be helpful in avoiding misunderstandings regarding what FireCalc does and does not do.

[jebmke]

Quote:

Originally Posted by REWahoo I consider the 'traditional SWR rule' to be the point from which everyone (including me) deviates.

I know a lot of retirees and I don't know any who use any kind of mathematical method to manage their withdrawal. Most of them use either "judgement" or "dividends only + pension/SS if available."


I view the SWR as a curb feeler. This may not be a familiar term to some. Here is an explanation.

Quote:

Curb feelers or curb finders are springs or wires installed on a vehicle which act as "whiskers" to alert drivers when they are at the right distance from the curb while parking. The devices are fitted low on the body, close to the wheels. As the vehicle approaches the curb, the protruding feelers scrape against the curb, making a noise and alerting the driver in time to avoid damaging the wheels or hubcaps. The feelers are manufactured to be flexible and do not break easily.

[Midpack]

Quote:

Originally Posted by marko OP here. The responses so far have been what I'd expect. I do find it interesting that there is so much discussion, clarification and hand-wringing about the 'traditional SWR rule' (initial X% plus inflation) only to realize that few, if any actually use this method. Could the overall discussions become as academic as my post?

Who recommended retirees blindly follow an initial 4% + inflation withdrawal methodology for 30 years?

From the outset the authors of the Trinity study didn’t, they noted at the time: “The word planning is emphasized because of the great uncertainties in the stock and bond markets. Mid-course corrections likely will be required, with the actual dollar amounts withdrawn adjusted downward or upward relative to the plan. The investor needs to keep in mind that selection of a withdrawal rate is not a matter of contract but rather a matter of planning.”

It was an academic study to give a ballpark guideline, to give us all some order of magnitude $ goal to work toward during accumulation, that’s it. Something better than “save a bunch of money, and spend cautiously” or “the stock market returns 10%, so you can spend that every year.”