What works in Zurich & Paris: Dividend yield (conclusion)

We have seen in our two previous articles that moderate, growing dividends covered by the progression of profits constituted the best possible yield strategy on the French and Swiss markets. However, the restrictive criteria used had the consequence of reducing the available securities so sharply that they made this approach unusable in a real portfolio, at least as a single strategy.

We have discussed the possibility of broadening the search horizon by selecting titles in the French and Swiss markets at the same time. Let's see what that gives:

The backtest gives us very good results, with a average annual return at 16,22%. Nevertheless, the number of securities, seven on average, remains insufficient to constitute a portfolio applicable in real life.

An alternative is to run the two portfolios (CH and F) in parallel. This gives us an average annual performance of nearly 18%, with 11 stocks on average. Here too, it remains a little meager for a real PF.

Keep it simple

By trying too hard to achieve excellence, we have come up against the impossible. Strategies based on average, growing dividends covered by profits, in France and Switzerland, are inapplicable on a daily basis as single approaches. However, they can be combined with other strategies, which is already a very good point.

Importantly, we have also demonstrated in our two previous articles that the following basic indicators perform better than the market:

• performance compared to peers within industries (Paris Stock Exchange)
• performance compared to the global market (Zurich)

In the highest quintile, for these approaches, there are nearly 80 stocks in France and nearly 50 stocks in Switzerland. This is more than enough to constitute a real portfolio.

We then limit the number of titles by keeping those which pay the best return, for example the first 20. In doing so, we obtain:

• for Zurich: 12 % average annual profitability since 2004
• for Paris: 7% average annual profitability since 2004

For Paris, we see a slight decrease compared to the figures obtained for the comparison with peers. This is due to a Yield Trap phenomenon, given that we targeted the 20 best yields, among the 5th quintile. This was not observed in Switzerland, as was already the case for the PER. To overcome this problem, we simply need to add one of the rules related to profits that we mentioned in our last posts (growth in profits for the last 12 months compared to the previous 12 months > 3%). We then obtain an average annual profitability of 8.75%. This is slightly more than what we had for the 5th quintile within industries.

The results are less good than with a more sophisticated approach, based on dividend growth, but they remain more than honorable compared to the performance of the respective markets. Above all, they can be reproduced in a real portfolio.

Conclusion

Through our various backtests carried out on dividends in Switzerland and France, we were able to highlight the following underlying trends:

• Moderate returns (compared to industry peers) are underpinned by growth 
• average and increasing dividends, supported by progressive profits, constitute the most effective approach, but it is necessary to resort to many criteria that make it inoperable as a single approach in a real portfolio, because of the small number of securities that compose it => it is therefore necessary to juxtapose this strategy with other approaches
• If we want to build a real PF using dividends as the only approach, it is necessary to stick to the most effective indicator on the market concerned.

In the next article, I will inaugurate a new series called "The War of the Wallets". This one echoes the different tests that I have already carried out in my work on the validity of some well-known (and less well-known) asset allocations. I will update some of the analyses already carried out and also review other approaches.

I will also take this opportunity to talk a little about the ETFs that represent the asset classes within these PFs. At the rate the ETF industry is progressing, there will soon be more of them than stocks. There is currently one ETF for every four listed companies. The number of the latter is growing by about 3% per year, while the number of ETFs is growing by nearly 15% per year. It is really becoming a jungle.

In the near future, I will alternate between backtests concerning factors specific to stocks (such as Yield here - in the "What works" series) and those relating to portfolios, made up of different asset classes (via their ETFs), in the "Portfolio war" series).

Here too, we risk having some surprises.

Navigation in the series

<< What works in Paris: Dividend yieldBacktest of the Price-to-Sales ratio in Switzerland and France >>