In my previous posts, I tried to create dividend stock portfolios that can outperform the market by a big margin. In those cases, I created screening methods that show such promise, but the demand was to rebalance the portfolio every four weeks and replace the stocks that no longer comply with the screening requirement with other stocks that comply with the requirement. Since most investors do not have the opportunity to rebalance the portfolio every four weeks, I created a new screening method that requires rebalancing the portfolio just once a year. Furthermore, in order to decrease the maximum expected drawdown to a lower level than that of the benchmark I had to be satisfied with a bit lower return, but still much better than the benchmark.
I have searched for profitable companies that are included in the Russell 3000 index that pay solid dividends with a low payout ratio and high dividend growth over the past five years. Those stocks also would have to show low debt.
The screen's method that I use to build my portfolio requires all stocks to comply with all following demands:
- Dividend yield is greater than 2%.
- The payout ratio is less than 75%.
- The annual rate of dividend growth over the past five years is greater than 5%.
- The forward P/E is less than 15.
- The PEG ratio is less than 1.50.
- Total debt-to-equity ratio is less than 0.40.
- The eight stocks with the lowest payout ratio among all the stocks that complied with the first six demands.
I used the Portfolio123's powerful screener to perform the search and to run back-tests. Nonetheless, the screening method should only serve as a basis for further research. All the data for this article were taken from Portfolio123.
After running this screen on June 26, 2013, before the market open, I discovered the following eight stocks: Valero Energy Corp. (VLO), Assurant Inc. (AIZ), AFLAC Inc (AFL), PartnerRe Ltd. (PRE), Schweitzer-Mauduit Intl Inc (SWM), AXIS Capital Holdings Ltd (AXS), Exxon Mobil Corp. (XOM) and Chubb Corp. (CB).
The table below presents the eight companies, their last price, their market cap and their industry.
The table below presents the dividend yield, the payout ratio, the annual rate of dividend growth over the past five years, the forward P/E, the PEG ratio, and the total debt-to-equity ratio for the eight companies.
In order to find out how such a screening formula would have performed during the last year, last 5 years and last 14 years, I ran the back-tests, which are available by the Portfolio123's screener.
The back-test takes into account running the screen once a year and replacing the stocks that no longer comply with the screening requirement with other stocks that comply with the requirement. The theoretical return is calculated in comparison to the benchmarks (S&P 500, Russell 3000), considering 0.25% slippage for each trade and 1.5% annual carry cost (broker cost). The back-tests results are shown in the charts and the tables below.
One year back-test
Just a matter of curiosity, the table below presents the eight companies originated by the screen formula one year before, on June 25, 2012.
Five years back-test
The table below presents the eight companies originated by the screen formula on June 25, 2008.
Fourteen years back-test
The table below presents the eight companies originated by the screen formula on September 11, 1999.
The long-term dividend screen has given much better returns during the last year, the last five years and the last 14 years than the S&P 500 and the Russell 3000 benchmarks. The Sharpe ratio, which measures the ratio of reward to risk, was also much better in all the three tests. Furthermore, the maximum drawdown, which normally is much bigger in a small portfolio than in the benchmarks, was much smaller in all the three tests. One year return of the screen was at 39.93% while the return of the S&P 500 index during the same period was at 20.90% and the return of the Russell 3000 was at 21.92%. The difference between the long-term dividend screen to the benchmarks was much more noticeable in the 14 years back-test. The 14-year average annual return of the screen was at 15.85% while the average annual return of the S&P 500 index during the same period was only 1.79% and the return of the Russell 3000 was at 2.45%. The maximum drawdown of the screen was only 39.97% while that of the S&P 500 was at 56.39% and the maximum drawdown of the Russell 3000 was at 57.07%. Although this screening system has given superior results, I recommend readers use this list of stocks as a basis for further research.