Monday, February 8, 2021

Revisiting the Canadian TULF Dividend Growth Portfolio

One of the most read entries on this blog is my 2016 take on the Canadian TULF Dividend Growth Portfolio. When re-reading the post and a couple of critical comments that it generated, I thought it would be worthwhile to revisit the entry and take a different perspective. Since it's been almost five years since the entry, my personal viewpoints have changed, and I'm going to attempt to argue in favor of the TULF portfolio by presenting counter examples to invalidate my three criticisms of Tom Connolly's four stock portfolio. 

For any readers who aren't familiar with Tom Connolly, he published The Connolly Report investment newsletter for over 30 years, and is one of the best known Canadian dividend growth advocates. Mr. Connolly set out the idea of the four stock TULF portfolio, with "T" standing for telecommunication companies, "U" for utilities, "L" for low-yielding dividend growth stocks with growth potential, and "F" for financials. To ensure high quality companies, Mr. Connolly suggested investors limit themselves to the S&P/TSX Dividend Aristocrats Index, which consists of 87 members and can be traded under the CDZ ishares ETF. It is worth noting that Mr. Connolly had certain rules regarding eliminating any "high yield" stock with a payout over 6%, and then removing cyclical stocks (i.e. energy and mining) as they were perceived as too risky to include in a four stock portfolio.

My three criticisms regarding the TULF portfolio are include below in italics, with my counterarguments in normal font. 

1. Lack of sector and geographical diversification

Once you’ve filled up your portfolio with telecoms (4 candidates), utilities (5 candidates), and financials (12 candidates), you're left trying to identify low-yielders with above average dividend growth potential in the 28 remaining companies. Since Mr. Connolly fails to define what he considers a low-yielding company, it's fair to focus on the lower yielding half of the 28 remaining companies in search of candidates that would provide above-average dividend growth and sector diversification. 

Although a couple of the larger utility companies and financial firms that Tom recommends have exposure outside of Canada, limiting a portfolio to TULF companies would expose you heavily toward the small, resource-centered Canadian economy. Personally, I aim  to keep at least 30% of my investment capital dedicated to international equities in order to provide better geographical diversification.

Going through the 87 companies in S&P/TSX Dividend Aristocrats Index, I came up with some names to include in portfolios that would provide some geographic and sector diversification:

Sample geographically diversified portfolio: T: Telus (most international Canadian telecommunication player), U: Algonquin (or Fortis, or Emera would provide US exposure), F: Bank of Nova Scotia (Latin and South America exposure) or Brookfield Asset Management (global exposure, but a lower dividend yield), L: Magna (international exposure to vehicle manufacturers)

Sample portfolio to provide sector diversification: F: Onex (multisector holdings), T: Telus (health exposure in addition to telecommunications), U: Capital Power (producer and energy trader), L: Dollarama (general retailer)

Although achieving strong geographic and sector diversification is going to be a challenge with only 87 Canadian based companies to choose from, the sample portfolios above show promise in at least providing some geographic and sector diversification. 

2. Identifying low-yielders who can and will grow their dividends quickly is difficult and not necessarily relevant

Accurately identifying low-yielders who can continue to grow their dividends at a fast rate for an extended period of time is as difficult as timing the market. These special types of companies are even harder to find in the small and often domestically focused Canadian market. Furthermore, I have yet to see a valid case made from a mathematical standpoint of why including low-yielders in a portfolio is necessary. Assuming you get very lucky and select a company currently yielding 1.0% that can grow dividends at 20% for 10 years, your yield on cost would grow to 6.2% at the end of the period. In contrast, if you pick a company currently yielding 5%, that grows their dividend by a paltry 2% over 10 years, you end up with the same yield on cost of 6.2%.

Instead of trying to argue the math presented above, which is futile, I'll instead present the below names of companies (all part of the 87 S&P/TSX Dividend Aristocrats Index) who have relatively low yields, but have been able to grow their dividends annually over prolonged periods of time, which I checked through using the Canadian Dividend All-Star list at December 31, 2020.

Metro Inc.: Yield at Dec 31st = 1.6%, 26 year streak of raising dividends, 1-yr div growth = 12.5%, 5-yr return = 22%
Candian National Railway: Yield = 1.6%, 25 year streak of raising dividends, 1-yr div growth = 7.0%, 5-yr return = 73%
Alimentation Couche-Tard Inc.: Yield = 0.8%, 11 year streak of raising dividends, 1-yr div growth = 19.0%, 5-yr return = 38%
Brookfield Asset Management: Yield = 1.2%, 9 year streak of raising dividends, 1-yr div growth = 12.5%, 5-yr return = 88%
Open Text Corp.: Yield = 1.8%, 8 year streak of raising dividends, 1-yr div growth = 6.7%, 5-yr return = 81%

Keeping in mind that the 5-year returns presented above don't include dividends, the benefit of adding these low-yielding stocks to a portfolio becomes obvious: potential for large total returns. 

3. Energy vs Utility Companies?

While Mr. Connolly suggests removing cyclical stocks such as energy and mining companies, he goes onto say TransCanada (classified in the Energy sector) was the first stock he owned. I also find it odd he advises against including energy companies and yet dedicates a whole category for utilities. There are blurry lines and high correlations between energy and utility companies. For instance, Enbridge Inc is classified in the index as an Energy company which would make it ineligible for the TULF portfolio. However, living in Quebec, Enbridge is viewed as a utility company by its many natural gas customers who pay their monthly bill to Enbridge subsidiary Gazifere. By discounting an entire category of companies he deems to be cyclical, Mr. Connolly leaves the investor with an even smaller potential universe of companies from which to select. 

The best way I can think of  to support Mr. Connolly's argument to include utility companies, and remove energy companies, is to compare the performance of indices representing these two sectors (which I did using 

TSX Energy Capped Index ($TTEN) 5-year Performance: -37%
TSX Utility Capped Index ($TTUT) 5-year Performance: +47%

It's hard to argue against 84% outperformance over the past five years. Although I do think a longer time frame would make for an even more meaningful comparison. 

Before wrapping up, I can think of a couple more bonuses for the Canadian TULF portfolio:
- Four stock portfolio is easy to construct, follow, rebalance yearly and would have low transaction costs
- Strict criteria can help automate decisions
- Starting with only quality companies that are part of the S&P/TSX Dividend Aristocrats Index is likely to lead to less big losses of capital.

Does your portfolio include all of the TULF components?