What Happens If You Just Pick the Stocks That Beat the Market the Most?

We had a pretty obvious question about Persistfolio: why not just find the stocks that beat the S&P 500 by the most and build a portfolio from those? If that gives us roughly the same stocks as Persistfolio, then all the additional statistical work isn't really adding much. So we tested it. We took the 20 stocks selected by Persistfolio and compared them with the 20 stocks that had the highest historical returns above the S&P 500, using 5, 10 and 15 years of history.

The overlap wasn't that high:

• 5 years: 10 of 20 stocks were the same

• 10 years: 8 of 20

• 15 years: 10 of 20

So Persistfolio clearly wasn't just giving us a list of stocks that went up the most. We then looked at the stocks where the two approaches disagreed to see what was different about them. The biggest difference we found was volatility. For stocks picked by Persistfolio but not by the simple-return method, average historical volatility was 30.6% over 5 years, compared with 55.2% for the simple-return picks. Over 10 years it was 34.0% versus 48.3%, and over 15 years it was 26.6% versus 45.2%. Maximum drawdowns showed the same general pattern.

What's interesting is that Persistfolio doesn't have a rule telling it to pick low-volatility stocks. We weren't screening for that. It seems to happen because Persistfolio isn't only interested in how much a stock beat the market. It looks for statistical evidence that the historical outperformance was persistent. Think about two stocks that both beat the S&P 500 over ten years. One did it fairly steadily. The other had a few huge years, some terrible ones, and ended up with the higher total return. If we're simply ranking by total return, the second stock wins. Persistfolio can see those two histories differently. A very large return isn't necessarily the same thing as a persistent one.

That sounds like a fairly small distinction, but it changed roughly half the stocks in our 20-stock portfolios. Whether that actually leads to better results is something we don't know yet. Lower historical volatility doesn't mean lower future volatility, and it certainly doesn't guarantee outperformance. So we're keeping both sets of portfolios and tracking them from here: the stocks selected by Persistfolio on one side, and the stocks we would have selected simply by taking the biggest historical outperformers on the other. We'll see which approach actually does better.

Learn more about Persistfolio: