Stock Screen Review 2017

What's the point?

Reviewing the weekly screens that I monitor in this column has some fanciful aspects. First off, for most investors the main purpose of screening is to generate ideas that can be researched further before making an investment decision. However, to monitor the usefulness of the screens, I treat each one as an equally-weighted portfolio of shares based on all the ideas generated.

What’s more, while this review looks at the screens over several fixed time periods running to 1 January 2018, the actual screen results themselves are arguably only really relevant around the time each screen is conducted. To give an extreme example, if a share makes it into a value-orientated screen at 100p and roars up to 150p over the following week, it probably would not appear in the subsequent screen. Indeed, one big week could have accounted for all the year's gains (possibly more).

And then there is the existential question of what the historic performance of the screens means anyway. The screening column started a few years into the current, long-running bull market and has yet to face any really serious market headwinds, although the Brexit vote did knock a number of screens for six. The prevalent fair-weather conditions, mean I remain cautious about reading much into the near-blanket market outperformance of these screens since inception.

What’s more, the fact most screen are conducted just once a year means the tables presented here would probably not pass much muster with a serious statistician. And as my primary focus is on each screen’s ability to generate good ideas, as long as a screen stays true to its core objective, I am not afraid to tinker with criteria to make improvements or to make a screen more productive when it struggles to come up with a decent number of results. Therefore, not all of the screens are entirely consistent over time, although I’d argue each one is broadly consistent and I try never to tinker just for the fun of it. I think the virtues of being flexible are illustrated by the excellent performance of my Jim Slater/Zulu Principle inspired screen since I amended it to take account of a letter received from the great man himself before he passed away in late 2015. The changes he encouraged making to the screen, and which I duly put into practice, have been followed by a truly stellar run.

One big plus that this review can boast, in my opinion, is that all the screens have been conducted in 'real time', with the ideas behind them based on common-sense attempts to mimic a range of popular and proven investment styles. This contrasts to strategy performance data based on back-testing, which is often found to tell investors far more about what has been successful in the past rather than what will be successful in the future. 

It may not be scientific, but…

All in all, the spirit in which to take this review is probably as a rough and ready assessment of whether the screens have been throwing up some potentially useful stock ideas. And while it is interesting to note what strategies have been performing best and worst, the runs they are currently experiencing, whether good or bad, are more likely to be broken than not.

The performance given in the tables (both for the screens and indices) is based on Thomson Datastream total return data, which means it includes dividends paid and reinvested. The screens are compared with the indices from which they select stock ideas. One anomaly to this is the Genuine Growth screen, which selects stocks from the FTSE Aim 100 as well as the FTSE All-Share; the sole index it is currently compared against. I’ll look at using an amended benchmark when I next update this screen.

While my weekly column provides performance numbers for screens adjusted for notional cost associated with a real-world enactment of each screening strategy (these vary from 1 to 2.5 per cent depending on the liquidity of the stocks screened) it has not been possible to factor costs into this review. That means the numbers are an overstatement of any likely real-world returns.

As well as the raw performance numbers relating to the screens, I’ve provided a compound annual rate of return to make comparisons with the indices easier. I’ve also provided numbers for compound annual outperformance or underperformance against the index. All the tables are ordered from best to worst compound annual out- or underperformance.

Top of the pops

The out-and-out king of the strategies I regularly follow is my Great Expectations screen. This screen is all rock and roll. It looks for strong earnings upgrades from brokers, coupled with exceptional share price movement. This screen pays no heed to the question of valuation. Intuitively, this feel like a screen suited to a bull market and particularly a bull market powered by purchases of 'safe' assets at any price by central banks, which has had a knock-on effect on valuations of riskier assets while increasing the attractions of 'growth'. Since I first ran my Great Expectations screen six years ago it has produced a knockout 27 per cent compound annual return. What’s more, it’s achieved these returns targeting the largest companies the London market has to offer, selecting shares exclusively from the FTSE 350. Perhaps sometimes elephants do gallop after all. This is the first screen I normally run each year and it will feature in next week's column.

Despite the nature of the current bull market, by no means all of the top-performing strategies from this review are value blind. The John Neff screen, which is the only screen to outperform Great Expectations over the past year, and the High-Yield Small-Caps screen both have a strong focus on valuation. So too does the Cash Magic screen, which ranks second for relative performance since inception. I find it particularly pleasing to see the Cash Magic screen in second spot as it was devised in response to a reader request for a screen incorporating the cash-return-on-capital-invested (CROCI) measure of business quality (suggestions are always happily received).

The other reason the performance is pleasing is because the ideas for the screen are based on the excellent notions for blending measures of value and quality advanced by hedge fund star Joel Greenblatt. Mr Greenblatt’s own Magic Formula screen does not fair so well in the table. That said, the numbers in the table are based on the 30-stock version of his Magic Formula screen. The 10-stock version, which I also monitor, has actually produced far stronger results, returning 228 per cent since inception almost seven years ago compared with 139 per cent from the 30 stocks. That said, with increased portfolio concentration has come increased volatility.

Down at the bottom of the list for relative returns since inception is the Late Bloomers screen, which has only managed to keep pace with the FTSE All-Share. What’s more, as already mentioned, the numbers in this review ignore costs, so anyone who followed the strategy wholesale in the real world would have underperformed. The Late Bloomers screen is based on looking for companies valued at a low multiple of sales with the potential for, and signs of, margin recovery. I plan to try to soup this one up a bit at the time of its next outing, although keeping very much with the screen’s current spirit.

Performance since inception

5-year performance

3-year performance

1-year performance

Source: Thomson Datastream