Author Topic: Mechanics of SWR?  (Read 2667 times)

Raoul Duke

  • 5 O'Clock Shadow
  • *
  • Posts: 3
Mechanics of SWR?
« on: May 10, 2017, 08:08:59 PM »
Hi folks,

Please forgive a very naive question but can anyone shed some light on the raw mechanics of how one withdraws income from a portfolio please?

Let's say the portfolio is $1MM and the planned SWR is 3%, i.e. require $30K income per year. Say the dividends from the portfolio are 2%, so that's $20K. From poking around the mmm blog and this forum, it sounds like the remaining $10K is obtained through selling off 1% stock (which has hopefully experienced some growth > 1%). However, although after selling 1% of stock the net asset value should still be higher than start point (assuming growth > 1%), the number of remaining shares will be reduced, which will therefore produce a slightly lower dividend in the subsequent year, which will lead to a slightly larger amount of shares being sold to meet the deficit, and ...... this looks like a compounding negative situation? Is this an accurate analysis or am I missing some key point?

Greatly appreciate some clarity here - thanks!

ysette9

  • Walrus Stache
  • *******
  • Posts: 8930
  • Age: 2020
  • Location: Bay Area at heart living in the PNW
Re: Mechanics of SWR?
« Reply #1 on: May 10, 2017, 08:16:18 PM »
The key point you are missing is that ok average, over the long term your 3% withdrawal rate (however you take it; it makes no difference if from dividends or selling stock) is less than the average growth rate of your portfolio. If your equities portfolio grows at a real rate of 7% and you are taking out 3% (or 4%) then you are still coming out ahead.

This ignores sequence-of-returns risk which is a possibility with a 4% withdrawal rate. At a 3% withdrawal rate you are so totally bullet-proof you don't need to worry about that

maizefolk

  • Walrus Stache
  • *******
  • Posts: 7400
Re: Mechanics of SWR?
« Reply #2 on: May 10, 2017, 08:24:13 PM »
What you're missing is that in the average year dividends per share would also go up (average increase of ~5% a year).

So in the average year you'd end up have slightly fewer shares, each worth slightly more money and each paying a slightly higher dividend per share than in the previous year.

Raoul Duke

  • 5 O'Clock Shadow
  • *
  • Posts: 3
Re: Mechanics of SWR?
« Reply #3 on: May 11, 2017, 07:16:52 PM »
Folks,

Thank you all for taking the time to reply, I appreciate it. What you've said makes broad sense and i guess i was aware of the "sequence of returns" issue, which could make things more difficult for an early retiree if several bad years occur in a row at the beginning of retirement. Not sure one can really plan for that however.

Regarding maizeman's comment that dividends, on average, increase by 5% per year, out of interest i looked at the dividend paid out from Vanguard's S&P 500 index fund VFIAX (data from 2002 until present) and this doesn't quite jive. The reason i pick this example is because this index fund is popular among mustachians. Looks like they've stayed relatively constant around 2% over the 15 year period. Maybe maizeman was referring to dividends from individual stocks?

Thanks again!

maizefolk

  • Walrus Stache
  • *******
  • Posts: 7400
Re: Mechanics of SWR?
« Reply #4 on: May 11, 2017, 07:44:20 PM »
Raoul, dividends are paid in absolute amounts, not percentages. You're looking at dividends as a percentage of share price rather than in absolute payout per share, so the increase in the value of the index is camouflaging the increase in the amount of dividends paid out.

Let's pretend you bought an ETF fund which tracks the S&P 500 in 2002 and the units cost $100 a share at that point.

In 2002, the S&P 500 had a dividend yield of 1.79%, so each ETF share would pay you $1.79 in dividends per year ($100*0.0179). 

15 years ago the S&P index was at 1097, today it is at 2394, a total increase of 118%, so today the ETF share you bought in 2002 is worth $218. Today the dividend yield of the S&P 500 is 1.94% so each share is paying you $4.23 in dividends per year ($218*0.0194).

That means the amount of money you receive per ETF share has increased 136% over fifteen years, which works out to dividends increasing at an annual rate of 5.9%

Edit: L.A.S. did a great job of explaining this more clearly than than I managed, thanks!
« Last Edit: May 11, 2017, 07:46:10 PM by maizeman »

Raoul Duke

  • 5 O'Clock Shadow
  • *
  • Posts: 3
Re: Mechanics of SWR?
« Reply #5 on: May 12, 2017, 06:11:18 PM »
Hey L.A.S and maizeman - these are great replies! Thanks so much. I really appreciate the clarity (like i said, I'm naive to this stuff...) and i must say it sounds pretty encouraging.

The learning continues.......

Cheers!