Beating the TSX

Beating the TSX (BTSX) is an investment strategy developed by David Stanley and publicized in Canadian MoneySaver. It is a Canadianized version of the Dogs of the Dow strategy developed by Michael O'Higgins in his book Beating the Dow, and when applied to Canadian stocks is sometimes referred to as the "Dogs of the TSX". It is based on choosing the ten highest-yielding stocks in the index. However, index changes have required some modifications to Stanley's approach when used in Canada.

"Dogs of the Dow"
The original "Dogs of the Dow" simply selected the ten highest-yielding stocks in the thirty (30) stock Dow Jones Industrial Average and invested equal amounts in the ten (10) highest dividend-yielding stocks. The portfolio was reconstituted once a year.

Although this approach has been criticized in US usage, the purpose of this article is to discuss the Canadian background, not developments in US usage beyond the original findings in the original edition of Beating the Dow.

TSE 35
David Stanley first publicised this method in Canadian Moneysaver magazine in 1996, applying it to the closest match in Canada to the Dow Jones Industrial Average (DJIA). At that time, the "TSE 35", which was based on 35 stocks in the Toronto Stock Exchange, seemed the best match to the DJIA, and Stanley's backtesting suggested the approach was workable.

Dow Jones Titans 40
The TSE 35 became subsumed in what is now the S&P/TSX 60. In order to keep the selection base relatively small, Stanley changed his methodology to the newly-introduced "Dow Jones Canadian Titans" index, which had 40 stocks. The BTSX method continued using the Titans index until that index was discontinued.

Income trusts
During the period the Dow Jones Titans were being used, companies started to convert to high-yield equities called income trusts. The tax loophole favouring this conversion was later removed, and many former trusts converted back to normal companies, still offering higher than normal yields. Stanley decided to exclude income trusts and former income trusts from the BTSX selection process.

S&P/TSX 60
The discontinuation of the Dow Jones Canadian Titans index required a second switch, this time to the 60-stock S&P/TSX 60. However, the larger index, when sorted by yield, also contains several companies that were formerly income trusts. Stanley originally decided to continue excluding such companies (most of which were smaller components of the S&P/TSX60) from his stock picks. This requirement has now been rescinded.

Current methodology
The current methodology, as first established by David Stanley and modified by Ross Grant with Stanley's permission, takes the following form:
 * 1) Sort the 60 stocks in the S&P/TSX 60 by dividend yield.
 * 2) Invest equal dollar amounts in the 10 highest yielding stocks in the remaining list.
 * 3) If a member of the list cuts its dividend, sell it and replace with the next member on the reconstituted list on the date of the cut.
 * 4) Reconstitute once a year. (The current year end date is Dec. 31.)

(If fewer than 10 stocks are desired, the list is resorted by price from lowest to highest and stocks are selected beginning with the lowest priced.)

Performance
Over 25 years, the 10-stock "BTSX" portfolio has outperformed the comparison index 18 times, under performed six times and tied once. The average annual total return versus the average annual return of the Total Return index (using the appropriate comparison index for each period) is shown below.
 * {| class="wikitable" border="1"

! ! Portfolio ! Index
 * +Beating The TSX returns vs. the index (%), 1987-2017
 * Average Total Return (%)
 * 12.4
 * 9.6
 * % Increase
 * 29.1
 * }
 * 29.1
 * }
 * }

The above calculation shows only the average return including dividends for the BTSX, and does not include the value of compounding reinvested dividends. When the dividends are included, BTSX has over 30 years produced over double the amount of the underlying indexes from $1000 invested ($26273 vs $12850 for the Total Return index).

Criticism of performance report
It has been pointed out by forum member nisser, that BTSX is very dependent on start date. A 3-year rolling-return analysis of BTSX from 2003 to 2018, offset by one month, reveals that the median backtest barely beat the market. Starting on the wrong month would have underperformed the TSX market by up to -34%. A lucky start month would have delivered an excess return of 48%.

Possible reasons for outperformance
The BTSX strategy can be viewed as a blend of value investing and contrarian investing, using dividend yield as the value metric. Restricting the selection base to a large-capitalization index results in an attempt to capture the large cap value advantage in the multifactor model.

Arguments for the strategy

 * Easy to understand (clear set of rules)
 * No research needed
 * Has been shown to work
 * Relatively low cost to implement

Arguments against the strategy

 * Lack of diversification (too few stocks)
 * Tracking error relative to broad (cap-weighted) indices
 * Sector concentration (e.g. financials)
 * In a taxable account, tax drag due to yearly reconstitution
 * Based on synthetic returns (backtesting) ignoring the market impact of trading big positions.