Lots of misinformation in this thread, which can confuse the newbies. So let's state the mathematical facts:
1. A dividend is mathematically equivalent to selling an equivalent amount of stock.
2. If you're expecting X% return, it doesn't matter if you realize it through dividends, or by selling assets. You will never run out of either.
As a result of the above two mathematical facts, focusing your portfolio on yield alone, will only serve to reduce your diversification, causing your portfolio to be more risky, for equivalent return. If you do this in a taxable account, taxes will result in your portfolio being more risky, for a lower return.
OK, this is interesting. Please explain! I mean I understand that a dollar is a dollar, no matter where it comes from (is that what you mean by #1?), but how do you explain #2, if I'm selling shares, and the price is going down, then I'll need more shares to get the same dollar amount, so they could run out, whereas dividends never will....because I'll still have the underlying share.
Ok, a simplified version. Suppose you have one stock that pays a 4% dividend and is worth $100. You want to use a 4% WR. At first glance all seems fine and dandy - you can take the 4% dividend and keep every share, right? Well, not so fast. Let's say it drops 20% - the dividend would go up 20% (good... right?) - except that during economic downturns companies often cut their yield. Problem - now you need more than a 4% yield to have the same amount of real dollars.
Even more important, there is inflation to consider. If the company's share price stays the same and it's dividend stays the same you will have less purchasing power every year. Even if you have a company with a 5% yield you are still relying on growth to maintain a 4% WR - after just 13 years the yield alone can no longer sustain the WR if inflation goes along at 2%. If inflation is higher (3%) it's a mere 9 years.
This isn't to say that this strategy can't work. If the company's stock price grows at the rate of inflation and the yield stays at or above 4% you can sustain a 4% WR. But it's no "magic bullet." You will always need either the price to increase OR the yield to exceed WR + inflation.
Now let's look at using an ETF that tracks the SP500. It has a yield of 1.9% right now, meaning that to maintain a 4% WR you must sell some of your shares. This scares people, and leads them to the erroneous conclusion
"if I am selling shares I will eventually have no shares left!"
This is not necessarily the case, and here's why: The stock market goes up.
Ok, want a slightly more detailed explanation? Consider this example.
Suppose you have shares of the SP500 ETF with a current yield of 1.9% and the share price of this ETF is currently $100. You own 10,000 shares with a value of $1MM. Using a 4% WR you would have $40k/year to live off of. Let's assume real-adjusted growth of 5%
6%/year***, which around the historical average. During the first several years you will need to sell some shares to cover the difference between the dividend and your 4% WR. For example, in year 1 your dividend will supply $19k, and you will need to sell 210 shares @ $100 to generate the remaining $21k.
Oh no! people say...
Now I have 9790 shares, eventually i'll be doomed! However, there's two important things going on here. 1) each share is now worth more (the market goes up!), and 2) over time the need to sell any shares will go away. As the share price of the ETF goes up, the dividend (historically) goes up with it. Each time the share price goes up you need to sell less shares to meet your 4% WR target. Remember, your WR is fixed to when you started taking it out, indexed to inflation.
In year 2 you will need to sell just 196
192 (14 fewer shares) to make up the difference. This is because each share is worth more ($105), and you need to sell less of them because you will get more from the dividend ($19.17k).
By year 5 you will need to sell 141 shares (share price $117, dividend provides $22.2k)
By year 10 you will need to sell just 69 shares. (share price $142, dividend provides $28.3k)
After
15 17 years the dividend (which is a function of the share price) will actually provide more income than you need, even though the yield is still hovering around 1.9%. Remember, the yield is a function of the share price. After just three years the number of shares will actually start to increase as the increase. The total number of shares you will have to sell over the first
15 17 years will be 1,554 (leaving you with 8,445
8,623 shares now valued at $220 each, giving you a portfolio of $2.2MM
From this point forward the total number of shares owned would increase every year.
Here's how it would look in graphical form.
Year, Share price, % Withdraw from dividend,
1 $100 47.5%
2 $105 47.9%
3 $110.3 50.3%
4 $115.8 52.8%
5 $121.5 55.4%
...
10 $155.1 70.8%
...
15 $198.0 90.4%
By year
15 18 the yield surpasses the real-adjusted $40k necessary for the example above (e.g. you can now get $40k/year from the yield alone.. thanks to a steadily rising share price). At that point total # of shares start to increase as the leftover dividend buys more shares. If we look at the 30 year mark, that 1.9% dividend will provide 2.5x expenses.
*(note -
this is obviously a very simplistic example, where the share price increase every year by 6% in real adjusted returns, and the WR is kept exactly constant at 4%/$40,000. It is meant ONLY to illustrate how having a portfolio of only an SP500 ETF can result in more shares over time that are worth more money. In reality it would not be so smooth a ride, as the market will jump up and down. The dividend also fluctuates, and has been above 3% for the majority of the last 100 years... but not since the mid 1990s)
** if you want to look at actual historical scenarios just plug numbers into FIREcalc nad cFIREsim - given that the SP500 yield has never gone below 1% and typically stays above 2%, anytime a portfolio exceeds 2-4x its starting value by extension the yield exceeds the input WR, and the # of shares owned increased. It's not surprising considering that roughly 3/4 of 30y simulations using a 4% WR wind up with more money than they started with.
***EDIT: I went back and changed the numbers in the example to reflect 5% growth before a dividend. This is more in line with the historical, real-adjusted returns of the SP500.