08-01-2017, 03:07 AM
No but seriously I can't see how this calculation would be unbiased. Only looking at survivors is the biggest problem I have about the arguments and articles for dividend growth investing. Again, I'm NOT attacking the conclusion that dividend growth may well be a criterion leading to market beating or at least matching and less volatile returns 
. But those comparisons only seem to suggest that your DG61-portfolio was better than the total stock market, which should be clear when selecting them using data available after the studied period.
Could you run a similar backtest with maybe the aristocrats of 2007 or someone's handpicked portfolio of DG stocks from that time, actually posted somewhere before the crisis? I came across this list of aristocrats from 1989:
http://www.valuewalk.com/2015/05/list-of...9-to-2014/
What I'm suggesting is something like watering down the superstars from this article:
https://seekingalpha.com/article/2148343...89-to-2014
Here there is also a clear survivorship bias, as nobody should think they would have picked exactly KO, PG, LOW, JNJ, MMM, EMR and DOV from the 26 choices in 1989 and seen those sky high returns. But the real question I think is just how great were these returns: From the list above also CL, PH and GPC actually managed to raise their dividend to this day. Add 3 more companies which I (born after 1989
) recognize and are still operating, namely K, TMK and IFF. Here are the total returns of these 13 companies between 1/1/1989 and 6/30/2017:
KO 2929%
PG 3158%
LOW 16266%
JNJ 4690%
MMM 2924%
EMR 1644%
DOV 2179%
CL 4709%
PH 3296%
GPC 1424%
K 850%
TMK 2969%
IFF 1585%
That's an average return of 3740% i.e. $13,000 put equally in each stock would now be worth approx $499,000, which equals CAGR of 13.655%. Compare this to $13,000 invested in S&P500 or the Wilshire 5000 which would have turned to approx $210,000 or $213,000 respectively during the same period, CAGR approx 10.3%.
So the "best" 13 of 26 had significantly superior returns than the market. How to account for the obvious survivor bias? The return of this half of the sample was in fact more than double the return of the market!!! Meaning that if you had put $1000 in each of the 26 stocks, you would have gotten a better return than putting $1000 in only in these 13 and burning another $13,000 in a campfire, which in turn would actually had gotten a better return than putting $26,000 in either of these indexes.
There's of course the another 13 stocks which would likely returned real money as well. A pitfall is the huge run of LOW, which should have been left untouched with no balancing and imagine if someone would have chosen only all the other 25 names? I'd really like to see an actual total return of all the 26 aristocrats of 1989, some have been aquired etc. I got my numbers using that Seeking Alpha article and the total return calculators on dqydj.com, and there might be tiny rounding errors.
Sorry for the long post and again for lifting this old thread, I'm just really thinking over and over about just going the index route and want to eliminate all possible pitfalls in the arguments favoring stock picking dividend growth stocks. At least for the last 28 years the approach seems legit, but that isn't a long enough time frame to draw too certain conclusions. What I'm mostly worried now is that since the late 80s the interest rates have been declining, how will this strategy fare if they steadily increase to 6-10% in the next 15-20 years?
. But those comparisons only seem to suggest that your DG61-portfolio was better than the total stock market, which should be clear when selecting them using data available after the studied period.Could you run a similar backtest with maybe the aristocrats of 2007 or someone's handpicked portfolio of DG stocks from that time, actually posted somewhere before the crisis? I came across this list of aristocrats from 1989:
http://www.valuewalk.com/2015/05/list-of...9-to-2014/
What I'm suggesting is something like watering down the superstars from this article:
https://seekingalpha.com/article/2148343...89-to-2014
Here there is also a clear survivorship bias, as nobody should think they would have picked exactly KO, PG, LOW, JNJ, MMM, EMR and DOV from the 26 choices in 1989 and seen those sky high returns. But the real question I think is just how great were these returns: From the list above also CL, PH and GPC actually managed to raise their dividend to this day. Add 3 more companies which I (born after 1989
) recognize and are still operating, namely K, TMK and IFF. Here are the total returns of these 13 companies between 1/1/1989 and 6/30/2017:KO 2929%
PG 3158%
LOW 16266%
JNJ 4690%
MMM 2924%
EMR 1644%
DOV 2179%
CL 4709%
PH 3296%
GPC 1424%
K 850%
TMK 2969%
IFF 1585%
That's an average return of 3740% i.e. $13,000 put equally in each stock would now be worth approx $499,000, which equals CAGR of 13.655%. Compare this to $13,000 invested in S&P500 or the Wilshire 5000 which would have turned to approx $210,000 or $213,000 respectively during the same period, CAGR approx 10.3%.
So the "best" 13 of 26 had significantly superior returns than the market. How to account for the obvious survivor bias? The return of this half of the sample was in fact more than double the return of the market!!! Meaning that if you had put $1000 in each of the 26 stocks, you would have gotten a better return than putting $1000 in only in these 13 and burning another $13,000 in a campfire, which in turn would actually had gotten a better return than putting $26,000 in either of these indexes.
There's of course the another 13 stocks which would likely returned real money as well. A pitfall is the huge run of LOW, which should have been left untouched with no balancing and imagine if someone would have chosen only all the other 25 names? I'd really like to see an actual total return of all the 26 aristocrats of 1989, some have been aquired etc. I got my numbers using that Seeking Alpha article and the total return calculators on dqydj.com, and there might be tiny rounding errors.
Sorry for the long post and again for lifting this old thread, I'm just really thinking over and over about just going the index route and want to eliminate all possible pitfalls in the arguments favoring stock picking dividend growth stocks. At least for the last 28 years the approach seems legit, but that isn't a long enough time frame to draw too certain conclusions. What I'm mostly worried now is that since the late 80s the interest rates have been declining, how will this strategy fare if they steadily increase to 6-10% in the next 15-20 years?