Withdrawal Rate Example

[David1961]

There's a lot of talk about withdrawal rates (WR) and to my knowledge, there is no agreed-upon definition. Let me give a hypothetical example and I'll give my calculation of what the WR would be.

Investment portfolio (current value) = $1.6 mil
Investment portfolio (12 months ago) = $1.5 mil
Value of house, is primary residence, no mortgage = $300k
Pensions and/or SS received during past yr = $30k
Dividends and other distributions from portfolio during past yr = $50k
Expenses for last 12 months (includes all taxes and healthcare) = $65k

Assume all distributions from the portfolio ($50k) are all taken as cash.
Assume there is no income from rental property.


The question is what is this person's WR for the past year

I'd calculate it as ($65k - $30k)/ $1.5 mil = 2.3%

Note that I do not add the value of the house in the denominator. And also, since I'm looking at the WR for the past year, I use the portfolio value of a year ago instead of the value today in the denominator. Although income exceeds expenses, and he/she did not have to sell any of the portfolio during the past year, but still had to take $35 k from the portfolio.

How would others calculate the WR?

 

[Alan]

There's a lot of talk about withdrawal rates (WR) and to my knowledge, there is no agreed-upon definition. Let me give a hypothetical example and I'll give my calculation of what the WR would be.

Investment portfolio (current value) = $1.6 mil
Investment portfolio (12 months ago) = $1.5 mil
Value of house, is primary residence, no mortgage = $300k
Pensions and/or SS received during past yr = $30k
Dividends and other distributions from portfolio during past yr = $50k
Expenses for last 12 months (includes all taxes and healthcare) = $65k

Assume all distributions from the portfolio ($50k) are all taken as cash.
Assume there is no income from rental property.


The question is what is this person's WR for the past year

I'd calculate it as ($65k - $30k)/ $1.5 mil = 2.3%

Note that I do not add the value of the house in the denominator. And also, since I'm looking at the WR for the past year, I use the portfolio value of a year ago instead of the value today in the denominator. Although income exceeds expenses, and he/she did not have to sell any of the portfolio during the past year, but still had to take $35 k from the portfolio.

How would others calculate the WR?

Assuming the excess dividends (in cash) went back into your retirement account then this is exactly how I calculate my WR each year. It works for me but I'm sure others will have equally valid ways to do the calculation that will work for them.

 

[haha]

To the extent that I do this at all, what I do is take my gross expenses including income tax, subtract my gross SS, and divide the remainder by 12/31/ year-1 portfolio value.

There are actually many ways to do this, and what matters is that the studies you are basing the desirability or survivability of your scheme on uses this same way of looking it.

If you look around this site, there is not much more agreement about this than there is about anything else.

Ha

 

[Animorph]

Assuming the excess dividends (in cash) went back into your retirement account then this is exactly how I calculate my WR each year. It works for me but I'm sure others will have equally valid ways to do the calculation that will work for them.

I pretty much agree with that, but with the additional condition that the divisor needs to be at least roughly the portfolio value at the time of the withdrawal. If the portfolio has a value of $1M at the start of the year, falls to $0.5M by mid year, and you sell $40k of equities for expenses at the end of the year your withdrawal rate is not 4%! There's probably some simple calculus to figure out what the effective withdrawal rate is when withdrawals occur throughout the year.

The previous year value works fine if you take your expenses out of the portfolio at the start of the year, like most studies and many here.

 

[REWahoo]

The correct way to calculate your withdrawal rate - like the correct definition of retirement, the accurate way of determining the true value of gold, and the right way to measure the real rate of inflation - is how I do it. If you do it some other way you're just wrong...

 

[David1961]

Assuming the excess dividends (in cash) went back into your retirement account then this is exactly how I calculate my WR each year.

Yes, I also assume the dividends that are not spent go back into the portfolio.

 

[David1961]

I pretty much agree with that, but with the additional condition that the divisor needs to be at least roughly the portfolio value at the time of the withdrawal. If the portfolio has a value of $1M at the start of the year, falls to $0.5M by mid year, and you sell $40k of equities for expenses at the end of the year your withdrawal rate is not 4%! There's probably some simple calculus to figure out what the effective withdrawal rate is when withdrawals occur throughout the year.

I toyed with the idea of calculating a net WR for each month, based on the portfolio value at the beginning of the month and any sales during the month, but this was too much analyzing in my opinion.

 

[heeyy_joe]

I agree with your WR calculation.

 

[Alan]

I pretty much agree with that, but with the additional condition that the divisor needs to be at least roughly the portfolio value at the time of the withdrawal. If the portfolio has a value of $1M at the start of the year, falls to $0.5M by mid year, and you sell $40k of equities for expenses at the end of the year your withdrawal rate is not 4%! There's probably some simple calculus to figure out what the effective withdrawal rate is when withdrawals occur throughout the year.

The previous year value works fine if you take your expenses out of the portfolio at the start of the year, like most studies and many here.

I agree, but I'm not about to create the spreadsheet calculations to try and estimate the WR based on period withdrawals throughout the year.

As I said above I know there will be other perfectly valid ways of calculating a WR.

 

[pb4uski]

Assuming the excess dividends (in cash) went back into your retirement account then this is exactly how I calculate my WR each year. It works for me but I'm sure others will have equally valid ways to do the calculation that will work for them.

+1 though I think technically it is just a retirement date calculation and then the $35k of withdrawals is increased each year for inflation. That said, many people look at WR based on current year withdrawal/beginning of year portfolio balance.

 

[audreyh1]

Since I withdraw a fixed percent of my portfolio each year, I calculate the amount based on the value 12/31 each year, and withdraw it on the first trading day of the new year.

 

[LongPrime]

When our IRA VA/FI annuities mature and kickin @ ages 69, 70 and 72 and 69, the amount will be in excess of our need, I plan to take any excess and add to trading account or start 529/annuity for any grandchild that we may have- Paying it forward. We will not have a set WR since the most of our annuities are tIRA and will be throwing off 5% and 6.5%, which is well above minimum MRD. Hopefully the Roth's can be held off for growth and inheritance. And the trading-nonqualified accounts be used as cushion to needs and wants.

 

[W2R]

The correct way to calculate your withdrawal rate - like the correct definition of retirement, the accurate way of determining the true value of gold, and the right way to measure the real rate of inflation - is how I do it. If you do it some other way you're just wrong...

Normally I compute my WR as a ratio of the spending money withdrawn (everything I spent minus tiny pension) divided by my portfolio value (not including house) on 12/31 of the prior year. I believe my method is pretty similar to David1961's method or Haha's method or Audreyh1's method.

This gives me a slightly lower WR than the way YOU do it! I recall that on your recent thread you mentioned that in your method, the denominator is the portfolio value on your first day of retirement, NOT at the beginning of the year.

Where most of us differ, if anywhere, is in the denominator only AFAIK.

 

[2B]

There's a lot of talk about withdrawal rates (WR) and to my knowledge, there is no agreed-upon definition.

I agree that there are many different opinions and many ways to calculate a WR. I also suspect that nobody, no matter how convinced that they have a fool proof method, will actually stay with that approach during their entire retirement. Too many things change.

I am personally planning on a variable WR with a basic level that covers my "comfortable but with few splurges." SS and a bevy of small pensions cover this level but the SS won't kick in for a few years. I have a do-it-myself annuity created out of CDs set to bridge this gap but this also represents a very small part of my portfolio. The remaining portfolio can be tapped at up to 5% annually. A paid for house represents the final emergency fund to the possible nursing home stay if the portfolio is bled to far.

The one advantage, if you care to agree, of being a OMY-er is that you can build up a signicant portfolio above what most people would consider "enough" by most financial calculators. I've been a OMY-er for about 5 years now. This really is my last year.

 

[Dcharles]

I withdraw at 5 percent and cross my fingers and forget about it. To many unknowns I. The future to make an accurate prediction and I am taking an optimistic approach. I may need to change if things s
Go south, but the last thing I want to worry about each month is my withdraw rate. That just defeats the purpose of the good life in ER

 

[Ronstar]

I practiced WR for a couple of years before I retired. Just took expenses divided by beginning period portfolio value. Since i just retired 4/1/2014, I calc'd a preliminary WR by taking projected yearly expenses divided by 4/1/2014 portfolio value. I monitor monthly by keeping a running total of actual expenses against projected. I may set aside a certain amount of portfolio value outside of the WR calculation- to be used for home improvements.

 

[NW-Bound]

I pretty much agree with that, but with the additional condition that the divisor needs to be at least roughly the portfolio value at the time of the withdrawal. If the portfolio has a value of $1M at the start of the year, falls to $0.5M by mid year, and you sell $40k of equities for expenses at the end of the year your withdrawal rate is not 4%! There's probably some simple calculus to figure out what the effective withdrawal rate is when withdrawals occur throughout the year.

The previous year value works fine if you take your expenses out of the portfolio at the start of the year, like most studies and many here.

While my current AA is roughly 25% in cash (short-term instruments), I do not take out all of my annual expenses on Jan 1st. Heck, I do not even know exactly what I will be spending. So, I draw money each month as I need it. Cash is fungible, so it does not matter where I stash it and what account I spend from.

The OP method is reasonable to me, but I like to add that as I draw money throughout the year, more than once per month actually as I tend to incur unplanned expenses, I use the current portfolio value as the denominator. When the market goes up or down 10% in a year, it does not concern me that much, as my numerator fluctuates way more than that, due to home improvements, travels, gifts on a whim, etc...

However, if the market drops 50%, darn, I am not going to point to my beginning portfolio value and say that my WR [-]is[/-] was 4%. Darn it, it is going to be 8% soon if the market stays that low. I would be thinking of switching from canned cat food to dried cat food bought in bulk!

On the other hand, if the market goes up 50% (one can dream, yes?), I am not going to up my spending by 50% to match. That would be suicidal.

By the way, for expense smoothing, I look at the past 12 months and the 12 months before that to spot any troublesome trend.