Our screen of screens is beating its benchmark

The theory was first conceived in 1952 by the economist Harry Markowitz, who won a Nobel Prize for his work in the field. What is less well known is how its author invested his own cash. Instead of following his method, Markowitz chose to minimise his future regret at large possible stock market losses by splitting his savings equally between bonds and stocks.

As Brian Christian and Tom Griffiths note in Algorithms to Live By, the story could be presented as the Nobel Prize winner’s irrational adoption of a mental shortcut. Instead, Christian and Griffiths write that “it’s precisely because of the complexity of real life that a simple heuristic might in fact be the rational solution”.

Every stock screen is a heuristic. Whittling down a stock market’s myriad choices is a simple strategy to help investors identify companies that meet their criteria.

The problem is that this isn’t always a smart route to a diversified equity portfolio. A screen for ‘value’ stocks, for example, might identify a range of firms of different sizes and in different sectors. But those same results might all be equally highly exposed to a recession, which is why they were all cheap to begin with.

By using multiple screens at the same time, we can try to spread risk and factor concentration. That is one way to think about our Screen for All Seasons, now entering its seventh year, which casts six variously shaped nets over the premium and junior UK markets in a bid to upgrade our heuristic.

In aggregate, last year’s 60 stock selections delivered a 1.5 per cent negative total return. While nothing special, it marked the fourth year in succession in which the screen has beaten its benchmark, which is a split of the FTSE All-Share and Aim All-Share. Another poor outing for the junior index dragged down the negative total return of the combined indices to 8.7 per cent.

This takes the screen of screen’s total return to 40.5 per cent in the six years it’s been live. Though well ahead of its benchmark’s cumulative 8 per cent return, I’m minded to ignore that delta given Aim’s consistently dilutive effect on the latter’s performance. A better proxy for UK equity performance, the standalone FTSE All-Share, is up a more respectable 24.7 per cent in the period.

While the screens in these pages are intended as a source for further research rather than off-the-shelf portfolios, any real-world attempt to replicate the Screen for All Seasons’ methodology is likely to have stuck more closely to market returns. Indeed, by incorporating a chunky 1.5 per cent annual dealing charge to reflect the portfolio’s numerous and sometimes illiquid holdings, its six-year return drops to 28.3 per cent.

Each of the six screens uses a ranking system to create a selection of 10 stocks. When there are not 10 stocks that meet a screen’s full criteria, a weakened version of the test is employed, and details of any tests failed are given in the accompanying table of stock picks. Full details of the criteria of each screen can be found below along with the ranking methodology. All screens are conducted on all the constituents of the FTSE 350, FTSE All Small and FTSE Aim All-Share, although several screens have tests to remove companies below a certain size.

The screens

The stocks

An abridged version of this year’s 60 stocks is included in the table below. Lots more fundamental data and forecasts can be found in the downloadable excel version of the table here: