Canadian Money Forum banner

261 - 280 of 304 Posts

·
Registered
Joined
·
2,984 Posts
Any advice for when one should/could comfortably double from a 5 or 6-pack to a 10 or 12-pack?
I don't see why there has to be an amount to switch to having two stocks per sector rather than one. You are simply reducing your company specific risk when you do so, and that's a good thing.

ltr
 

·
Registered
Joined
·
17,032 Posts
Any advice for when one should/could comfortably double from a 5 or 6-pack to a 10 or 12-pack?
The criteria I would use is the amount of concentration in any single stock. This will depend entirely on your own investing situation. As a rule of thumb, you might not want more than 5% or 10% of your total investments to be in any single stock.

In my own situation, my 5-pack is part of several others investments (including S&P 500, gold, bonds). One stock in my pack is only 2% of my total.

In contrast, if my only investments were the 5-pack, meaning that a single stock was 20% of my total, then I would definitely diversify that further. I wouldn't want 20% of all my investments in a single stock.

Might as well give a 5-pack update: in the recent rally, the 5-pack was weaker than the broad index. The gap with XIU has narrowed significantly, and YTD the 5-pack is -8.7% and XIU -10.0%. So the performance of the two are starting to look very similar.
 

·
Registered
Joined
·
1,163 Posts
The time to switch is rather subjective and dependent on how that switch were to occur. Are you in accumulation or withdrawal? Are you going to sell to do so? Registered vs. non registered? etc. etc. If I was in a registered account during accumulation I would pick away with new money. If its non registered I may want to generate a capital loss by selling and adding a similar stock. I may also decide to trim back a winner or possibly an underperformer as part of my decision. The options are plenty. I would tend to agree with the others that the decision is a more about diversification and unless the portfolio is very small would not matter too much. I think the better approach would be to assess your current pack determine the additional stocks and go from there. I do have some core positions (Banks, Telcos, Pipelines & Utilities) across Canadian, US and International. I hold many positions most of them small gambles making up 1 or 2%. In hindsight I would enjoy the simplicity of a pack portfolio, perhaps 6 pack Canada, a dozen US, and another 6 International. However, I am not sure if I would be as engaged in my portfolio if it was set up that way. some would argue that it should be boring and to set it and forget it. I enjoy the process of profiling and selecting stocks so I consider it part retirement planning and part leisure.
 

·
Registered
Joined
·
208 Posts
Very interesting this 5-pack approach. Maybe I'd do a 10-pack just to have a backup for each of the 5 selected stocks, but still.

I've just backtested some 5-pack I did but instead of selecting the 5 biggest market cap for each of the selected sectors, I selected 5 stocks from different industries with high growth, high Sharpe ratio, high Sortino ratio, low drawdown and high worst year. This biggest risk is that these names have not necessarily proven themselves for more than 10 years and are not necessarily large caps and it's just a quick test based on historical returns, not potential future returns.

From TSX, this is what I ended up selecting. AQN, CJT, RPI-UN, CSU, BYD. Backtesting this selection on equal weight from Jan 2013 until Dec 2019, you get 38.62% CAGR, 12.06% stdev, best year 65.16%, worst year 13.81%, max drawdown -8.19%, Sharpe ratio 2.73, Sortino ratio 7.81.

Yup, Sharpe 2.73 and Sortino 7.81 on a 38.62% CAGR portfolio!

If you are wondering why I didn't include YTD with COVID crash, well it was just to have complete years, but COVID crash didn't really affect the portfolio except for a max drawdown of now -13.58%. Backtesting this selection on equal weight from Jan 2013 until Jun 2020, you get 39.34% CAGR, 14.23% stdev, best year 65.16%, worst year 13.81%, max drawdown -8.19%, Sharpe ratio 2.37, Sortino ratio 5.86. As expected, the Sortino ratio took a hit from the COVID crash. That's still solid returns.

All that being said, historical returns does not guarantee future returns. It was just to show one of the best combinations I've found so far on the TSX in a statistical point of view which is different than an investment analysis where you must do your due diligence to have a better idea on potential future returns.

Just to put things in perspective, that's AMZN as portfolio 1 in blue and my proposed 5-pack equal weight as portfolio 2 in red. See how stable it is.

20349
 
  • Like
Reactions: Pluto

·
Registered
Joined
·
17,032 Posts
I've just backtested some 5-pack I did but instead of selecting the 5 biggest market cap for each of the selected sectors, I selected 5 stocks from different industries with high growth, high Sharpe ratio, high Sortino ratio, low drawdown and high worst year.
. . .
From TSX, this is what I ended up selecting. AQN, CJT, RPI-UN, CSU, BYD. Backtesting this selection on equal weight from Jan 2013 until Dec 2019,
To do a proper back test, you have to see the world as it was back in time. In other words you have to simulate what you would have seen and known back then and only use knowledge available at that time. Otherwise you are adding massive hindsight bias. Did you extract those metrics at their readings on January 2013?

You also have to use the index that existed back then, not today's index constituents. For example CJT was only added to the TSX Composite in 2019 so it would not have showed up in your screening back in 2013.

I also suspect that AQN would have failed your screening since it had recently crashed very severely with a horrible drawdown, about 80% drop from 2006-2009 along with wretched performance since early 2000s.
 

·
Registered
Joined
·
208 Posts
To do a proper back test, you have to see the world as it was back in time. In other words you have to simulate what you would have seen and known back then and only use knowledge available at that time. Otherwise you are adding massive hindsight bias. Did you extract those metrics at their readings on January 2013?

You also have to use the index that existed back then, not today's index constituents. For example CJT was only added to the TSX Composite in 2019 so it would not have showed up in your screening back in 2013.
I've thought about this and I could start that analysis as of 2015, therefore with at least 2 years of data on CJT to take the decision and it would not have changed much, except maybe buying CNR instead of AQN.
 

·
Registered
Joined
·
17,032 Posts
I've thought about this and I could start that analysis as of 2015, therefore with at least 2 years of data on CJT to take the decision and it would not have changed much, except maybe buying CNR instead of AQN.
But why would you even test CJT when it wasn't in the TSX index at the time?
 

·
Registered
Joined
·
208 Posts
I may have missed when did CJT got promoted from TSXV to TSX or something like that and where to find that information, but we have information on CJT since 2012 and from beginning of 2013 to end of 2014 it was promising. I do look at TSXV stocks and I do own TSXV stocks at the moment and some of them got promoted to TSX.

One could've start at year 2016 if you wish at that would be 4 years at 30% CAGR on a very steady climb.

20350
 

·
Registered
Joined
·
17,032 Posts
I understand, just pointing out that to really test how something would have done in the past, you need to make sure you are seeing the world as it was back then. That extends to the activity of "finding" the stock(s) to begin with.
 

·
Registered
Joined
·
3,872 Posts
I did own CargoJet several years ago ...there was very little to suggest it is a 6 pack candidate...unless you are trying to pick the best returns using hindsight.
 

·
Registered
Joined
·
208 Posts
Ok, I get that you guys don't like my choice of CJT since it seems too much biased. I get that this is all hypothetical and that it is an idealised case. It is easy to talk about the good decisions we could've taken when looking at the past. What's hard is to take the good decisions in the present for the future. This next one I'll present seems realistic. There's maybe someone out there in Canada which has done this combination from this kind of analysis. I'm not saying that one should invest based on this kind of analysis. Please do you due diligence and invest based on your own style and risk tolerance. Statistical analysis of the past does NOT guarantee future results.

Here's another scenario that I hope you will find more realistic and less biased. Say we just had the crash in 2008 and now we are about to start year 2012. I'm an investor looking for best performers in the last 3 years (2009 to 2011). CJT cannot be part of the choice then. I look also at RPI-UN and from 2009 to 2011, it doesn't seems appealing.

I end up making these choices instead from the data I have from 2009 to 2011. Remember that I'm an investor looking for stocks that performed well after the crash. I'm looking for 5 stocks in 5 different industries/sectors which have had high CAGR in the past years, high Sharpe ratio, high Sortino ratio. I also want the combination of these 5 to give even better statistics and to have a decent stdev when combined.

I end up selecting these 5 from backtesting from 2009 to 2011.

20355


Combined in equal weight from 2009 to 2011, this is the result of their backtesting.

20356


Now, I'm convinced and go all-in in that 5-pack portfolio starting in 2012. Here's what would be my results as of today.

20357


20358


That's 29% CAGR over more than 8 years. I'll take it. Also, who knows, maybe from my screenings I would've been interested by the rising CJT and RPI-UN to replace IFC and IIP-UN? After all, at the end of 2015, looking to assess if my 5-pack still has the best performers, I would've had enough data to decide to switch for CJT and RPI-UN, maybe?

I know that most of the time that's not what happens in real life and it's still an idealised case, but these idealised cases have to happen to some people, don't you think?
 

·
Registered
Joined
·
17,032 Posts
I don't hate CJT by the way, in fact I hold it in my growth portfolio. I know that I confuse everyone by talking about my various portfolios. My 5-pack (BCE etc) is my big portfolio, most of my Canadian equity. But I have a much smaller growth/momentum portfolio too. It holds crazy things like WDO.

The reality is that there are a million ways to stock pick, and many of those million ways are pretty good approaches. As long as you have a reasonably sane stock picking approach, with sufficient diversification (minimum # of stocks and sectors) you'll be on good footing.

I would look at some practical things too, such as: how often will the portfolio be updated and refreshed? How much time and energy is involved in doing all this? Do you think you can keep doing that maintenance for 20 or 40 years? What's the plan when a stock turns into a disaster?
 

·
Registered
Joined
·
208 Posts
I don't hate CJT by the way, in fact I hold it in my growth portfolio.
I was not saying that you didn't like CJT, I was saying that I understand that you guys find that the first scenario that I presented was biased on CJT because back then it didn't have sufficient history for that kind of analysis.

The reality is that there are a million ways to stock pick, and many of those million ways are pretty good approaches. As long as you have a reasonably sane stock picking approach, with sufficient diversification (minimum # of stocks and sectors) you'll be on good footing.

I would look at some practical things too, such as: how often will the portfolio be updated and refreshed? How much time and energy is involved in doing all this? Do you think you can keep doing that maintenance for 20 or 40 years? What's the plan when a stock turns into a disaster?
Yes, that's why I don't like when people are opposing technical analysis to fundamental analysis. I'm just using both and also other strategies. In my opinion, there's never too much information and there's always something new to learn. At some point, the investor will find the tactics that seemed the best when looking only at a few key things in the fundamentals, in the statistical analysis of its history, in some technical indicators, in some trends and news.

I like pretty much that 5-pack approach. So far I'm holding more stocks in my main portfolio because I'm making experiments and playing around, but I feel like the 5-pack is less maintenance and still a very solid strategy. As I said, maybe I'd go for a 10-pack to reduce a bit the risk. Though, on the other side, I was even thinking of going to a 3-pack with only CJT, AQN and RPI-UN, but that would mean more risk and more watching. Still, it's pretty interesting to see that this 5-pack idea would've outperformed even an ETF like HQU (2x daily bull NASDAQ). I'm still not sure of my final strategy.

Personally, since I don't hold that much money at the moment, I'll rebalance every year and refresh every year if I find new stock ideas which seems better options than the current ones I own. On my personal portfolio, I'll be buying twice a year to reduce my commissions. On my matrimonial portfolio, I'll be buying monthly since it's bigger investments, but there will also be withdrawals.

For stocks turning to disaster, it'll depend of the stock's history and current context. I have a risk threshold for each.
 

·
Registered
Joined
·
17,032 Posts
As I said, maybe I'd go for a 10-pack to reduce a bit the risk. Though, on the other side, I was even thinking of going to a 3-pack with only CJT, AQN and RPI-UN, but that would mean more risk and more watching. Still, it's pretty interesting to see that this 5-pack idea would've outperformed even an ETF like HQU (2x daily bull NASDAQ). I'm still not sure of my final strategy.
Some of stocks you're listing are small cap and mid caps. These are inherently more volatile and I think you'll need more than 5 to get sufficient stability.

RPI.UN for example is less than $1 billion market cap and very thinly traded. I would expect some massive volatility to eventually hit that stock, so be careful. For example the CEO himself owns 19% of the shares! Another company executive owns 5%. Potentially very dangerous.

Because things like RPI.UN have the tendency to absolutely crater (when large holders exit), you need more stocks in the portfolio to diversify against such disasters.
 

·
Registered
Joined
·
208 Posts
Some of stocks you're listing are small cap and mid caps. These are inherently more volatile and I think you'll need more than 5 to get sufficient stability.

RPI.UN for example is less than $1 billion market cap and very thinly traded. I would expect some massive volatility to eventually hit that stock, so be careful. For example the CEO himself owns 19% of the shares! Another company executive owns 5%. Potentially very dangerous.

Because things like RPI.UN have the tendency to absolutely crater (when large holders exit), you need more stocks in the portfolio to diversify against such disasters.
Sure, smaller caps usually have higher volatility. Not sure where you see the volatility in this selection of picks, though. It has Sharpe & Sortino ratios way higher than even a low volatility ETF like XMW. I'm being cautious when it's getting close to micro cap, but small caps that have been there for more than a decade with a average volume*price over 500k$ are fine to me.

I remember you are also holding CJT like I do and it's only 5x the volume, 4x the cap of RPI-UN and I bet you are pretty happy holding CJT.
 

·
Registered
Joined
·
2,984 Posts
Not sure where you see the volatility in this selection of picks, though. It has Sharpe & Sortino ratios way higher than even a low volatility ETF like XMW. I'm being cautious when it's getting close to micro cap, but small caps that have been there for more than a decade with a average volume*price over 500k$ are fine to me.
I think James is just saying to be cautious about small cap companies as their volatility and lifespan might not be suited for a Pack with only 5 stocks. I would caution the same. In fact, I continue to caution James about a 5 pack with large caps because each stock represents 20% - too much.

Generally the definition of small cap comes from its market value of its outstanding shares and not its Sharpe & Sortino ratios or even average volume. There are some discrepancies in definition, but my own definition of a small cap would be around 1 billion.

I personally own a stock with a market cap of 1.5B (which some might define as the bottom of mid-cap), but that's an anomaly for me, because I like this stock in a sector that is quite thin on good companies, but then I have a 24 pack, which is 8 sectors of 3 stocks each, so that one stock is about 4%, rather than 20% for a 5 pack.

If you're going to assemble a 5 pack, use large caps to reduce the inherent high risk of that strategy.

ltr
 

·
Registered
Joined
·
17,032 Posts
I think James is just saying to be cautious about small cap companies as their volatility and lifespan might not be suited for a Pack with only 5 stocks. I would caution the same. In fact, I continue to caution James about a 5 pack with large caps because each stock represents 20% - too much.

Generally the definition of small cap comes from its market value of its outstanding shares and not its Sharpe & Sortino ratios or even average volume.
Yes exactly. Small caps are inherently volatile, even if they haven't shown volatility for the last few years.

And I hear like_to_retire's caution that my 5 large caps are too highly concentrated. It's a good warning. I just decided to compromise on this, because fewer positions are easier to manage. Additionally, I have many other investments besides my 5-pack. This pack is 10% of my overall investments so one of my 5-pack stocks at 20% weight is, in reality, 0.10*0.20 = 2% of my overall investments. Meaning that ENB is 2% of my overall, RY is 2% of my overall, etc.
 

·
Registered
Joined
·
208 Posts
I think James is just saying to be cautious about small cap companies as their volatility and lifespan might not be suited for a Pack with only 5 stocks. I would caution the same. In fact, I continue to caution James about a 5 pack with large caps because each stock represents 20% - too much.

Generally the definition of small cap comes from its market value of its outstanding shares and not its Sharpe & Sortino ratios or even average volume. There are some discrepancies in definition, but my own definition of a small cap would be around 1 billion.

I personally own a stock with a market cap of 1.5B (which some might define as the bottom of mid-cap), but that's an anomaly for me, because I like this stock in a sector that is quite thin on good companies, but then I have a 24 pack, which is 8 sectors of 3 stocks each, so that one stock is about 4%, rather than 20% for a 5 pack.

If you're going to assemble a 5 pack, use large caps to reduce the inherent high risk of that strategy.

ltr
Yes exactly. Small caps are inherently volatile, even if they haven't shown volatility for the last few years.

And I hear like_to_retire's caution that my 5 large caps are too highly concentrated. It's a good warning. I just decided to compromise on this, because fewer positions are easier to manage. Additionally, I have many other investments besides my 5-pack. This pack is 10% of my overall investments so one of my 5-pack stocks at 20% weight is, in reality, 0.10*0.20 = 2% of my overall investments. Meaning that ENB is 2% of my overall, RY is 2% of my overall, etc.
Yes I know small cap definition comes from market cap, I used Sharpe & Sortino ratios just to point out the great risk-return of the 5-pack portfolio I was analysing.

I also agree that more stocks would be better to reduce risk. Though, when looking at Buffet's BRK picks, about 35-50% of its holding is in AAPL only. More than 75% of its holding is in the 7 first stocks.
 
261 - 280 of 304 Posts
Top