Financial Jungle recently wrote about a study put out by Tweedy Browne that purports to show the clear superiority of investing in high dividend yield stocks. One of the examples quoted in the Tweedy Browne study is the Dogs of the Dow.
Constructing the Dogs of the Dow is simple - after sorting the stocks in the Dow Jones Industrial Average (DJIA) by their dividend yields, pick the top ten stocks and invest equal amounts in each stock. Next year, construct a new list and repeat the process again. In Beating the Dow published in 1991, authors O’Higgins and J. Downs calculate that the Dogs strategy would have returned 17.9% annually, compared to 13.0% for the DJIA for the 1973 to 1998 time period.
Did the Dogs of the Dow actually outperform simply buying and holding the DJIA by a massive 5% per year? The answer, it turns out, depends on who does the measuring. In a paper titled, The “Dogs of the Dow” Myth, author Mark Hirschey of the University of Kansas, calculates the returns for the same time period and comes up with an annual return of 15.3%, 2.6% less than the O’Higgins numbers cited in the Tweedy Browne study, for the same time period!
The outperformance also depends on the time period under consideration. While there is still a 2.3% outperformance for the 1973-98 time period, Hirschey find that the outperformance for the 1961-98 time period is only 1.55%.
The excess returns discussed thus far do not account for the excess trading costs (in the past trading costs were significant) and turnover compared to simply buying-and-holding the thirty stocks in the DJIA because the Dogs portfolio needs to be rebalanced every year to sell the stocks that aren’t Dogs anymore, buy the new Dogs and rebalance the existing Dogs to equal weights. If held in a taxable account, the excess turnover in a Dow Dogs portfolio would create an additional drag in the form of taxes. Hirschey estimates these excess costs and concludes:
Thus, implementation of a Dow Dog investment strategy results in added annual brokerage costs of 0.34%, plus added income taxes on dividends of 0.74%, plus added capital gains taxes of 0.50% — or 1.58% in annual transactions costs. These transaction costs are roughly equivalent to the previously unexplained excess returns of 1.55% per year for the Dow Dog strategy over the 1961-98 period.
In a future post, we’ll take a look at the High Yield, Low Payout paper quoted in the Tweedy Browne study.
Bookmark: del.icio.us Digg StumbleUpon
19 responses so far ↓
1 Charles Martineau // Apr 22, 2008 at 10:59 pm
Something really important to mention that it is not mention above, is the “investor learning” effect where, if everyone who does the Dogs of the Dow strategy starting Jan 1st; some savvy investors will invest prior to this date in the top 10 high D/P and get an extra return from everyone purchasing the top 10 stocks on Jan 1st. This was first reported by Grant McQueen in ” Does the DOW-10 Investment Strategy — Beat the Dow Statistically and Economically?” 1997
2 Canadian Capitalist // Apr 22, 2008 at 11:20 pm
Charles: Excellent point. While I didn’t include it, the Myth paper discusses it noting that about $20Billion worth of investments in the Dow strategy. If a large number of investors chase a strategy, it may not work anymore.
3 Gregory // Apr 23, 2008 at 7:24 am
For good read exposing this myth, among others, is Ken Fisher’s The Only Three Questions That Count.
4 Joseph // Apr 23, 2008 at 7:49 am
It would be really, really nice if you started to focus on Canadian issues and securities.
5 Canadian Capitalist // Apr 23, 2008 at 9:23 am
Gregory: I haven’t read Ken Fisher’s book. I’ll put in a request at the library.
Joseph: As Canadian investors we can’t just ignore the US, which makes up close to 45% of the world stock markets, or for that matter, equities in the rest of the world. Out of curiosity, I took a look at the posts in April so far - out of 18 posts, 7 were of Canadian interest only and 8 were of general interest. Only three posts were about US markets that would still be of interest to Canadians. IMO, our focus is already firmly on Canadian issues.
6 ThickenMyWallet // Apr 23, 2008 at 3:17 pm
My issue with the Dogs of the Dow strategy is always that it is a high dividend yield stock because the share price has fallen. Thus, while yield is high, expectation of future performance is quite low. Its giving with one hand and taking with the other.
7 Four Pillars // Apr 23, 2008 at 6:56 pm
Joseph - the dogs of the dow is a universal trading strategy - the catchy name obviously applies to a US market but the principal of buying the highest dividend stocks can be used in any market.
“Toads of the TSX”….?
8 Financial Jungle // Apr 24, 2008 at 12:42 am
I have 3 points to make:
(1) As I alluded to in our previous discussion, it’s very important to examine expenses on both sides, otherwise you get a lopsided comparison.
Index investing isn’t a friction-free exercise.
Remember that Vanguard only came into existence in the mid 70’s. Prior to that, investors would have to incur transaction fees. I’m not sure what kind of MERS Vanguard was charging in the early days, but I’m pretty sure they were much higher than today seeing that the trend is going down.
(2) Moreover, indexes also have turnovers. Since 1961, of the 30 Dow components, only *4* remain on the list today.
http://djindexes.com/mdsidx/downloads/DJIA_Hist_Comp.pdf
(3) For registered accounts, the dividend & capital gain tax (1.25% drag) wouldn’t matter for both strategies.
If we account for the indexing MER and turnovers, the Dogs of the Dow is still a superior strategy.
9 Financial Jungle // Apr 24, 2008 at 2:33 am
>>”some savvy investors will invest prior to this date in the top 10 high D/P and get an extra return from everyone purchasing the top 10 stocks on Jan 1st. ”
In theory, yes, but that’s not what’s happening in reality. Footnote 10 of the Myth paper (revised in 1999) said:
>>”At the turn of the year 2000, Chevron ($59.7 billion) and Goodyear ($5.1 billion) will be dropped from the 1999 Dow Dogs and be replaced by 2000 newcomers International Paper and SBC Communications. With roughly $2 billion each in targeted buying, International Paper and SBC Communications may enjoy some price strength tied to Dow Dog buying. Given its relatively modest market capitalization, price strength may be especially notable in the case of International Paper. With roughly $2 billion each in targeted selling, Chevron and Goodyear, especially, may be vulnerable to some price weakness tied to Dow Dog liquidation.”
As it turned out, Chevron and Goodyear outperformed International Paper in January 2000. Couldn’t dig up SBC Communications; they merged with AT&T.
>>”the Myth paper discusses it noting that about $20Billion worth of investments in the Dow strategy.”
Most market participants don’t have the conviction to following through with the Dogs strategy on a consistent basis. $20 billion was only an estimate, but I’m sure much of it would’ve migrated to tech stocks by end of 1999, one of the worst year for the Dogs (Dow and TSX) on record.
http://www.dogsofthedow.com/dogs99p.htm
http://www.ndir.com/SI/articles/MS0205.shtml
10 Canadian Capitalist // Apr 24, 2008 at 10:35 am
FJ: The big story here is the incredible shrinking out performance. According to the paper cited here, for the same period described in the Dogs of the Dow strategy, the gross out performance is 2.3%, not 4.9% as shown in the Tweedy Browne study. The Dow Dogs advantage before accounting for excess costs over five years is all over the map: 3.26%, -0.28%, 8.79%, 2.75%, 0.36%, -0.84% and -3.4%. I wouldn’t call it “still a superior strategy”.
“I’m sure much of it would’ve migrated to tech stocks by end of 1999, one of the worst year for the Dogs (Dow and TSX) on record.”
I’m not so sure. Plenty of people were chasing the Foolish Four strategy in the late nineties.
11 Financial Jungle // Apr 24, 2008 at 12:55 pm
>>”According to the paper cited here, for the same period described in the Dogs of the Dow strategy, the gross out performance is 2.3%, not 4.9% as shown in the Tweedy Browne study.”
I’m not sure why the discrepancy in outperformance. I like to see Micheal O’Higgins respond to this Myth paper if he’s still around.
>>”The Dow Dogs advantage before accounting for excess costs over five years is all over the map: 3.26%, -0.28%, 8.79%, 2.75%, 0.36%, -0.84% and -3.4%. I wouldn’t call it “still a superior strategy”.”
This is the type of performance I was expecting from any dividend strategy, and exactly what I was hoping for. I want a strategy that will rock people’s boats once in a while. This is why dividend investing is simple to follow but difficult to execute because investors always surrender to the latest trends eventually.
If investors were chasing Dogs with $20 billion in 1998, how did Dogs outperform Dow in 2000 and 2001? Why didn’t we see a spike in International Paper’s stock when it was introduced to the Dogs list in 2000? My guess is that most of the Dogs followers surrendered to tech stocks during 1999.
We always come back to this: the theory that investors are always rational, and always exploiting market inefficiencies aren’t exemplified in the real world.
12 Canadian Capitalist // Apr 24, 2008 at 1:43 pm
Here’s one more paper that comes to the same conclusions, namely, while there is evidence for gross out performance, after trading costs and taxes, the premiums vanish.
Link
“We always come back to this: the theory that investors are always rational, and always exploiting market inefficiencies aren’t exemplified in the real world.”
I think we are in agreement that that’s not true (as the previous post talked about the validity of indexing even if markets are not efficient). But my view on the subject is that neither are investors always irrational and market inefficiencies that do exist are easily exploitable.
13 FinancialJungle // Apr 24, 2008 at 3:39 pm
CC: Just want to make a general comment that we’re probably in the 6th or 7th inning of our discussion, and I’d rather not have to keep repeating the 5th inning.
Once again, these studies are comparing Dogs’ net return versus DOW’s gross return. I don’t see how that’s a fair comparison. We should either compare gross vs gross, or net vs net.
BTW, all these talks about taxes aren’t applicable to most retail investors when they haven’t maxed out their 401k or RRSPs.
>>”But my view on the subject is that neither are investors always irrational and market inefficiencies that do exist are easily exploitable.”
Didn’t you just finish reviewing Predictably Irrational? Again, this falls into what I was saying before. It’s not easily exploitable because value investors must go against the crowd. Market participants are only irrational maybe 25% of the time, so this strategy requires both conviction and patience.
Take Keynote System for example. It was sitting on my watch list for 4 months, until a nasty downgrade by an analyst, sending the stock to a stunning 25% decline. Going against the exodus, I put down half a position. The other half never got filled because the stock is almost back to where it was before the downgrade.
http://financialjungle.com/?s=keynote&x=0&y=0
After waiting around for ~1.5 year, I finally bought a couple of REITs when the sector absorbed a 25% haircut. Back in 1999, I could not have done this because I didn’t have the stomach it. (I’m not trying to blow my horn. That’s not my style, and I’m learning something new almost everyday.)
Beside, the merits of dividend investing don’t depend on market beating performances. Dividend investing promotes good investing habits, and dividends are far less volatile than capital gains.
http://financialjungle.com/2008/04/24/dividend-increases/dividend-increases-steady-income-without-market-volatility/
14 Canadian Capitalist // Apr 24, 2008 at 4:30 pm
“Once again, these studies are comparing Dogs’ net return versus DOW’s gross return. I don’t see how that’s a fair comparison. We should either compare gross vs gross, or net vs net.”
You say that I keep repeating myself. That’s because you never seem to agree that there were significant excess trading costs in the past that should be accounted for. In the paper I referred above, the authors estimate the Dow-10 trading costs as 0.59% per year and the Dow-30 costs as 0.02% assuming a 1% cost in the time period under study. We are talking about 150 basis points difference here and fully 1/3rd could be explained due to transaction expenses. But I do promise not to raise this point again
If you are talking about stock picking as a skill (I think that’s what you are taking about with Keynote Systems, which doesn’t pay a dividend), I don’t fall in the camp that it doesn’t exist. But then, it’s not a generic strategy anymore. It’s for every investor to track their performance and decide if they have good stock-picking skills and whether their results are skill or luck. My only claim is that beating the market is a not an easy gig and to beware strategies that tell you it is.
15 Financial Jungle // Apr 24, 2008 at 6:05 pm
I agree with your closing statement from the previous discussion that investors should be “paying attention to costs & taxes, not chasing performance, not trading too much, investing or reinvesting regularly.”
Personally, I don’t practice the Dogs of the Dow strategy; therefore my trading expenses and taxes are minimal due to low turnovers and the preferential tax treatments from dividend tax credits. There’s no reason to ever sell dividend stocks like Fortis, TransCanada or Royal Bank. I’m going to pass these stocks to my children anyway, so my investment time-horizon is forever.
For me, I appreciate the steady + rising paychecks steaming from a diversified dividend portfolio; so, even if I end up matching the benchmark by age 65, so what? I’m still better off, because my retirement income isn’t compromised by market hysterias.
But seriously, shouldn’t we be picking on daytraders instead of each other?
16 Phil S // Apr 24, 2008 at 6:56 pm
Regarding Charles Martineau’s first comment… What Charles is describing is very much akin to market timing triggered value investing. I would actually go one step further on his basic premise as I generally like to go shopping for stocks just before the end of the year, when everybody else is doing their tax loss selling, just to see if anybody is dumping companies I like.
Regarding Joseph’s comment. Unfortunately, if you already own the big 5 banks and you don’t want to buy any more resource stocks, you’re pretty much pooched in Canada. The plain reality is that there isn’t very much to buy on teh TSX. Now for me, I actually do like small and micro cap companies, but in reality it’s more like gambling than investing, so I don’t like to recommend any of them to anybody I like… But my point is that the NYSE and NASDAQ are chock full of wonderful businesses in all kinds of different sectors of the economy and most brokerages will allow Canadians to trade down there.
17 Rates Cuts, New Look, and Other Links | Million Dollar Journey // Apr 25, 2008 at 5:31 am
[...] Capitalist shows us that investing in the "Dogs of the Dow" may not be as lucrative as we might [...]
18 Personal Money Tips // Apr 25, 2008 at 12:42 pm
The shortcoming of the dog approach has always been transaction costs. But any pre-tax return over 10% is still very respectable. Smaller investor should keep a close eye on their trading fees however.
19 Weekly Dividend Investing Roundup - April 26, 2008 » The Dividend Guy Blog // Apr 26, 2008 at 8:35 am
[...] the Dogs of the Dow [...]
Leave a Comment