Dogs of the TSE
Re: Dogs of the TSE
Why does he pass on Husky? That seems like a very arbitrary rule (i.e., if it makes him uncomfortable, it seems it would fall under rule #1).
-
- Veteran Contributor
- Posts: 5923
- Joined: 27 Feb 2005 07:14
- Location: Canada
Re: Dogs of the TSE
Well, from the article he says:Spudd wrote:Why does he pass on Husky? That seems like a very arbitrary rule (i.e., if it makes him uncomfortable, it seems it would fall under rule #1).
"I always remove Husky Energy (HSE) from the list since its dividend payment history is erratic. The company pays out more when they make more and pays out less when they make less. This is really not the steady blue chip business that we are looking to invest in.
I've held HSE for years in my own BTSX, but each to their own set of mods to the system.
ltr
Re: Dogs of the TSE
I would pass on HSE for other reasons. I believe it is primarily a vehicle for Li Ki Shing to fund the rest of his $32B ventures and HSE's offshore China ventures are only as secure as Li's relationship with the current regime. Risky in my opinion.
Still, that is hardly a valid reason to punt HSE from consideration in the Dogs index.
Still, that is hardly a valid reason to punt HSE from consideration in the Dogs index.
finiki, the Canadian financial wiki The go-to place to bolster your financial freedom
- scomac
- Veteran Contributor
- Posts: 7788
- Joined: 19 Feb 2005 09:47
- Location: The Gateway to Wine Country
Re: Dogs of the TSE
The more rules, or personal assessments that come into play, the less this becomes a simple yield based strategy. Instead of the Canadian version of the Dogs of the Dow, it could be more aptly called the Tom, Dick or Harry version of beating the TSX depending upon who's running the show. It appears to be getting further and further away from what was originally envisioned (and proven) many years ago. More factors aren't necessarily better.
"On what principle is it, that when we see nothing but improvement behind us, we are to expect nothing but deterioration before us?"
Thomas Babington Macaulay in 1830
Thomas Babington Macaulay in 1830
Re: Dogs of the TSE
IIRC, until the issue of former income trusts arose, David Stanley followed the rules rigorously. It's tempting to start up a Classic Dogs screen at the end of the year and compare it with Grant's.
Re: Dogs of the TSE
Sold TA for a loss on 10/01/2014 at $13.90. Replaced it immediately with CPX at $21.81. Same sector, same geography, but a much healthier company on most metrics. TA is down 24% from where I sold it. CPX is up 19%.like_to_retire wrote:I have to admit his rule #4 gives me pause, and I may revisit this as a new rule for my own BTSX that I have followed for several years. I'll have to give it more thought. I remember that FWF member ig17 was pontificating on this very issue some time ago in this thread. I wonder what his resolution was.
IIRC, I sold TA because I was very uncomfortable with it. Way too many problems and no light at the end of the tunnel. Grants' rule #1 makes perfect sense to me.
Guilty as charged. I use Dogs screen for ideas when I have new money to invest, along with other value screens. Annual BTSX turnover is one of the main reasons why I deviate from the canonical strategy. The turnover has been quite high in the last few years. I'm in a high tax bracket so I prefer to buy and hold.DenisD wrote:The first rule leaves me "uncomfortable". IMHO, it's no longer a stock screen, but stock picking.
I've held HSE since the inception of my BTSX-like portfolio, but sold it this Tuesday at $28.67.like_to_retire wrote:I've held HSE for years in my own BTSX, but each to their own set of mods to the system.
The reason is simple. I bought HSE in 2012 and then added more in 2013. My ACB on it was $27.96. That was with oil trading around $100. I collected two years worth of dividends and sold it at a tiny gain after oil price got cut in half. This is the kind of thing that makes you go WTF?!? Either HSE was incredibly cheap when oil traded at $100, or it's incredibly expensive now. I have no idea what it is so I'm out.
Re: Dogs of the TSE
How many years can a company be a Dog until you call it dead?
Specifically referring to TA here. It seems to have been on a downslide for the last 10 years with no light in sight.
Specifically referring to TA here. It seems to have been on a downslide for the last 10 years with no light in sight.
Re: Dogs of the TSE
Isn't that thinking counterproductive to the entire concept? I thought the point of the strategy is to buy beat-up stocks with the surrogate market being the dividend yield. And so if the dividend yield is high when the company is doing well then that's a perfect reason to include it because the shares themselves will go up.like_to_retire wrote:Well, from the article he says:Spudd wrote:Why does he pass on Husky? That seems like a very arbitrary rule (i.e., if it makes him uncomfortable, it seems it would fall under rule #1).
"I always remove Husky Energy (HSE) from the list since its dividend payment history is erratic. The company pays out more when they make more and pays out less when they make less. This is really not the steady blue chip business that we are looking to invest in.
I've held HSE for years in my own BTSX, but each to their own set of mods to the system.
ltr
Re: Dogs of the TSE
For the Classic Dogs, it would be dead when it's removed from the index you're using.nisser wrote:How many years can a company be a Dog until you call it dead?
-
- Veteran Contributor
- Posts: 5923
- Joined: 27 Feb 2005 07:14
- Location: Canada
Re: Dogs of the TSE
The BTSX is considered a value play. Screening for the highest dividend yield will result both in stocks whose dividend is simply high, along with stocks that are beaten down and so the yield has moved them into the list (the latter being the value play). It's a crap shoot every year on the anniversary date as to what proportion of those two categories is in the majority.nisser wrote:Isn't that thinking counterproductive to the entire concept? I thought the point of the strategy is to buy beat-up stocks with the surrogate market being the dividend yield. And so if the dividend yield is high when the company is doing well then that's a perfect reason to include it because the shares themselves will go up.
It was unwanted sector concentration and US/CDN currency valuation that led me to make my own modifications to the standard BTSX, so I sure don't begrudge anyone coming up with their own set of rules as Ross Grant has done. I'd rather see a set of hard rules applied rather than a blanket rule that a stock is removed if they're 'uncomfortable' with it, because to me that starts to smack of stock picking.
ltr
Re: Dogs of the TSE
Maybe Ross should call his portfolio the Comfortable Dogs.
Re: Dogs of the TSE
I like Norm's version of the Dogs screen that he runs in Money Sense, the Safer Canadian Dogs.
I am a big fan of strict quant rules. Unfortunately, they are not foolproof. COS passed Norm's quant screen on Oct 16th. Ironically, the screen appeared in the very same article where Norm warned about the likely dividend cut:
http://www.moneysense.ca/invest/stocks/ ... ds-at-risk
I wonder if COS would have passed Grant's qualitative comfort rule. Probably not.
Norm uses quant tests to eliminate potentially unsafe dogs. Grant's rule #1 is qualitative in nature, but the goal seems to be the same: eliminate unsafe companies.Safer Canadian Dogs
My safer variant of the Dogs of the TSX tracks the 10 stocks in the index with the highest dividend yields provided they also pass a series of safety tests, such as having positive earnings. The idea is to weed out companies that might cut their dividends in the near term. Just be warned, it’s a task that’s easier said than done.
I am a big fan of strict quant rules. Unfortunately, they are not foolproof. COS passed Norm's quant screen on Oct 16th. Ironically, the screen appeared in the very same article where Norm warned about the likely dividend cut:
http://www.moneysense.ca/invest/stocks/ ... ds-at-risk
I wonder if COS would have passed Grant's qualitative comfort rule. Probably not.
Re: Dogs of the TSE
I always try to sell a stock "before" a dividend cut, not after. I don't always get it right, but sometimes I surprise even myself.
Re: Dogs of the TSE
Thanks for the shout out ig17!ig17 wrote:I like Norm's version of the Dogs screen that he runs in Money Sense, the Safer Canadian Dogs.
Norm uses quant tests to eliminate potentially unsafe dogs. Grant's rule #1 is qualitative in nature, but the goal seems to be the same: eliminate unsafe companies.Safer Canadian Dogs
My safer variant of the Dogs of the TSX tracks the 10 stocks in the index with the highest dividend yields provided they also pass a series of safety tests, such as having positive earnings. The idea is to weed out companies that might cut their dividends in the near term. Just be warned, it’s a task that’s easier said than done.
I am a big fan of strict quant rules. Unfortunately, they are not foolproof. COS passed Norm's quant screen on Oct 16th. Ironically, the screen appeared in the very same article where Norm warned about the likely dividend cut:
http://www.moneysense.ca/invest/stocks/ ... ds-at-risk
I do have a general policy of just letting quant methods run. But I also like to point out potential dangers - thus the cautionary note on COS.
For those that want it, I also run the standard dogs of the S&P/TSX 60 screen each week in the SNW over here.
Ideally, it would be nice to avoid the Laidlaws and Yellow Pages of the world when investing in the dogs. But it isn't clear how to weed them our without giving up a great deal of upside from the dogs that recover.
I've been running a series of back tests on variations of the method when I happen to be near a Bloomberg terminal and will likely modify my "safer" variant based on some of my findings. But the work is preliminary at the moment and data mining is a real issue.
If anyone has a variant of the method they'd like to see back tested, let me know.
Re: Dogs of the TSE
I'd be interested to see at what point is it worth just abandoning a dog that's been a dog for many years. How many consecutive years has TA been a dog now?NormR wrote:
If anyone has a variant of the method they'd like to see back tested, let me know.
-
- Veteran Contributor
- Posts: 5923
- Joined: 27 Feb 2005 07:14
- Location: Canada
Re: Dogs of the TSE
Its price began a slide starting back in August 2008 when it reached a peak of $37.38 and still hasn't recovered.nisser wrote: How many consecutive years has TA been a dog now?
I've had it in my BTSX since Nov 2011 and blindly followed my rules since then.
It's results in my BTSX have been.
Nov 2011 - 2012 = -28.96%
Nov 2012 - 2013 = -9.77%
Nov 2013 - 2014 = -23.52%
Nov 2014 - today = +2.52%
My BTSX has done extremely well, but would have done a lot better without this dog. I wonder if I should consider a "comfortable" rule.
ltr
Re: Dogs of the TSE
TA entered BTSX in 2010, when David Stanley was forced to switch the index from the DJ 40 Canadian Titans to the S&P/TSX 60. TA wasn't eligible before 2010 because it wasn't part of the DJ 40.
Canadian MoneySaver, June 2010:
Canadian MoneySaver, June 2010:
David Stanley wrote: There was some upheaval with the index this year. As I mentioned in an earlier column (September 2009), the DJ 40 Canadian Titan Index was jettisoned half way through the year in favour of the larger DJ 60 Canadian Titan Index. I said then and still maintain that a true Canadian blue-chip index would contain no more than 40 stocks and preferably less. If the granddaddy of all blue-chip indices, the Dow Jones U.S. 30 Industrial Index, can cover that huge market with 30 stocks, I fail to see why Canada should require more than that.
...
Our new index is the S&P/TSX 60.
Re: Dogs of the TSE
I'd like to see a back test that limits the investable universe to the top 30 companies, as suggested by Stanley.NormR wrote:If anyone has a variant of the method they'd like to see back tested, let me know.
In fact, two back tests. One would rank the companies by the market cap. Another - by the revenue.
Re: Dogs of the TSE
Wondering how far back you can backtest Canadian screens with Bloomberg? What are your views on the optimum length of a backtest?NormR wrote:Bloomberg
About three years ago, I made a change to my large-cap screens to eliminate the worst losers. A stop loss algorithm O'Shaughnessy has mentioned is: sell after a 50% loss if the percent return of the stock is in the bottom decile of the universe. I decided to not buy a stock if it's three or six month percent return is in the bottom decile of the universe.
A year ago, P123 started providing data on Canadian companies. So I decided to backtest my change for my Canadian and US large-cap screens. I ran rolling backtests with 10, 15 and 20 stock screens. Sometimes the changed version was a little better. And sometimes it was a little worse. But it was always less volatile. I decided to keep the change. More comfortable.
- Peculiar_Investor
- Administrator
- Posts: 13267
- Joined: 01 Mar 2005 14:52
- Location: Calgary
- Contact:
Re: Dogs of the TSE
Is a stock screen expected to be correct 100% of the time? If so, why not pick the single best choice, rather than ten (or some other number). My understanding of the rationale for selecting the ten highest-yielding stocks in the index was for diversification purposes as there was no expectation that all ten would show positive, index-beating, returns. Having ten choices in play reduces the impact of making one poor performing selection.
As others have noted, once you start building exceptions into the process, it seems to defeat the whole purpose of a screen, which is to eliminate/reduce the role of human judgement in the selection process.
As others have noted, once you start building exceptions into the process, it seems to defeat the whole purpose of a screen, which is to eliminate/reduce the role of human judgement in the selection process.
Absolutely agree. Seems to me that when 'exceptions' start getting included in the screen it usually means that whatever made the screen work in the first place is no longer working and it might be a signal to stop using the screen, rather than to continue to "refine" it.scomac wrote:The more rules, or personal assessments that come into play, the less this becomes a simple yield based strategy. Instead of the Canadian version of the Dogs of the Dow, it could be more aptly called the Tom, Dick or Harry version of beating the TSX depending upon who's running the show. It appears to be getting further and further away from what was originally envisioned (and proven) many years ago. More factors aren't necessarily better.
finiki, the Canadian financial wiki New editors wanted and welcomed, please help collaborate and improve the wiki.
Normal people… believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet. – Scott Adams
Normal people… believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet. – Scott Adams
- scomac
- Veteran Contributor
- Posts: 7788
- Joined: 19 Feb 2005 09:47
- Location: The Gateway to Wine Country
Re: Dogs of the TSE
Stock picking doesn't work 100% of the time even if your name is Warren Buffett -- just ask him. It's not supposed to be 100%. It's supposed to work on-balance. IOW, in order to be successful at stock picking, you need to have more than half of your picks work out and you have to know when to get out on the ones that don't. Ben Graham said that if your stock selection operations worked out 60% of the time you would do well. Granted that was written 60 some years ago, but I don't see how it doesn't apply today. Taken in the context of the BTSX, you've got 10 picks to lesson the impact of being exactly wrong and you've got an annual revisit that allows you to sell high and buy low on-balance.Peculiar_Investor wrote:Is a stock screen expected to be correct 100% of the time? If so, why not pick the single best choice, rather than ten (or some other number). My understanding of the rationale for selecting the ten highest-yielding stocks in the index was for diversification purposes as there was no expectation that all ten would show positive, index-beating, returns. Having ten choices in play reduces the impact of making one poor performing selection.
Stock picking isn't a game of perfect. A screen is just a very basic form of stock picking and a method that should mitigate the amount of human judgement that comes to bare.
"On what principle is it, that when we see nothing but improvement behind us, we are to expect nothing but deterioration before us?"
Thomas Babington Macaulay in 1830
Thomas Babington Macaulay in 1830
-
- Veteran Contributor
- Posts: 5923
- Joined: 27 Feb 2005 07:14
- Location: Canada
Re: Dogs of the TSE
But if you build hard exception rules into the process, and follow them religiously year after year, then there's no human judgement (or random stock picking) involved.Peculiar_Investor wrote:As others have noted, once you start building exceptions into the process, it seems to defeat the whole purpose of a screen, which is to eliminate/reduce the role of human judgement in the selection process.
But Stanley built the former trusts rule into his process and it continued to perform well and beat the TSX. Ross Grant has modified it somewhat and his version has also performed well. His return over the last 14 years is similar to Stanley's 28 year average.Peculiar_Investor wrote:Absolutely agree. Seems to me that when 'exceptions' start getting included in the screen it usually means that whatever made the screen work in the first place is no longer working and it might be a signal to stop using the screen, rather than to continue to "refine" it.
ltr
- Peculiar_Investor
- Administrator
- Posts: 13267
- Joined: 01 Mar 2005 14:52
- Location: Calgary
- Contact:
Re: Dogs of the TSE
An exception for income trusts I can definitely understand, they were 'odd-beasts' created specifically for tax, not business, reasons. So in reality the screen just had another parameter and less of an exception. But the other exceptions being discussed here such as HSE and/or TA are of the human judgement kind that make me question the process.
I'm not questioning the results, just the process.
I'm not questioning the results, just the process.
finiki, the Canadian financial wiki New editors wanted and welcomed, please help collaborate and improve the wiki.
Normal people… believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet. – Scott Adams
Normal people… believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet. – Scott Adams
Re: Dogs of the TSE
Nothing wrong with exceptions as long as they're rules and not judgment calls. And once you've changed the rules, don't call it the Dogs any more. Call it something like the Safer Dogs.
The number of companies in your screen is inversely proportional to your confidence in the stock picking ability of your screen. Similar to a human stock picker. Usually, returns and volatility are inversely proportional to the number of companies too.
I changed my screens from 10 to 20 companies when costs went down.
The number of companies in your screen is inversely proportional to your confidence in the stock picking ability of your screen. Similar to a human stock picker. Usually, returns and volatility are inversely proportional to the number of companies too.
I changed my screens from 10 to 20 companies when costs went down.
Re: Dogs of the TSE
Humm, sorry, I'm not sure how to test that one on Bloomberg. However, I would note that long-term reversal tends to be a positive for stocks that have fallen a great deal over many years.nisser wrote:I'd be interested to see at what point is it worth just abandoning a dog that's been a dog for many years. How many consecutive years has TA been a dog now?NormR wrote:If anyone has a variant of the method they'd like to see back tested, let me know.