To help me think about this, here is a REALLY crude example:
In scenario 1, I buy $10,000 worth of stock XYZ, yielding 5%, and after three months I get my first dividend of $125. Let's assume the share price does not move at all. I reinvest (with no transaction costs) that $125 into more shares of XYZ, so after three months, I have $10,125 worth of XYZ.
Scenario 2 is the same, except that instead of reinvesting the $125 dividend back into more shares of XYZ, I take that money and initiate a new position into stock ABC, which also yields 5%.
In either case, I am in the exact same position financially: after 3 months, my $10,000 worth of stock has turned into $10,125 worth of stock. In scenario 1, however, it looks like my $10,000 has "grown" to $10,125, for a three-month return of 1.25%. In scenario B, it looks like my initial $10,000 has just sat there flat, and I happen to have a new small investment in another company.
The $125 is either part of the return on the initial investment or it isn't, regardless of whether it gets reinvested in XYZ, buys new shares of XYZ, or sits in your account as cash. It seems a mistake to say that is counts as "return" only if automatically reinvested back in the company that issued it.
So with all that said, I think I have to see it the way Kerim and cannew present it. Shares bought with dividends (whether reinvested or put into other stocks) have to be treated the same as "new money."
That said, tracking and accounting all that does get tricky. Cannew's system sounds good, but with only 35 stocks, you're looking at tracking 140 transactions a year just for dividends! I guess there are worse problems to have, though!
EDITED TO ADD: I guess one real question is how do you track the return on stock XYZ when you take the dividends as cash and/or invest them elsewhere.
In scenario 1, I buy $10,000 worth of stock XYZ, yielding 5%, and after three months I get my first dividend of $125. Let's assume the share price does not move at all. I reinvest (with no transaction costs) that $125 into more shares of XYZ, so after three months, I have $10,125 worth of XYZ.
Scenario 2 is the same, except that instead of reinvesting the $125 dividend back into more shares of XYZ, I take that money and initiate a new position into stock ABC, which also yields 5%.
In either case, I am in the exact same position financially: after 3 months, my $10,000 worth of stock has turned into $10,125 worth of stock. In scenario 1, however, it looks like my $10,000 has "grown" to $10,125, for a three-month return of 1.25%. In scenario B, it looks like my initial $10,000 has just sat there flat, and I happen to have a new small investment in another company.
The $125 is either part of the return on the initial investment or it isn't, regardless of whether it gets reinvested in XYZ, buys new shares of XYZ, or sits in your account as cash. It seems a mistake to say that is counts as "return" only if automatically reinvested back in the company that issued it.
So with all that said, I think I have to see it the way Kerim and cannew present it. Shares bought with dividends (whether reinvested or put into other stocks) have to be treated the same as "new money."
That said, tracking and accounting all that does get tricky. Cannew's system sounds good, but with only 35 stocks, you're looking at tracking 140 transactions a year just for dividends! I guess there are worse problems to have, though!
EDITED TO ADD: I guess one real question is how do you track the return on stock XYZ when you take the dividends as cash and/or invest them elsewhere.