As far as investment strategies go, there is nothing especially clever about stock screening. At their most basic, stock screens are simply about grouping data into lists. You could, for example, screen for companies whose chief executive mentioned the term ‘AI’ in the past month, or that operate in the European energy sector, or have a mnemonic ticker. The options are limitless.
The more common use of stock screens, which involves filtering the market to highlight companies with various financial and valuation attributes, is also tautological. In defining a set of parameters, the data-reliant investor merely shapes their options around criteria they decide are effective.
While such criteria can be indicative of a well-run business, or a cheap valuation, they are blunt tools for assessing a company’s future performance. Lots of considerations in a full due diligence process – such as a company’s competitive edge, the quality and discipline of its management, and broader business trends within a sector – are glossed over.
For these reasons, many fund managers treat screens as a starting point for ideas and further qualitative analysis, rather than ends in themselves.
But this doesn’t mean screening and active management are far apart, or should be overlooked. Both share a goal to beat the market, rooted in the notion that investors can do better than passively hug an index. Both also share a belief that insight – provided by data, proprietary research, some framework, or a combination of all three – is a better starting point than the efficient market hypothesis, and its logical asset of choice, the tracker fund.
Such a philosophy isn’t likely to ever unearth a Rosetta Stone of stockpicking. But it can still aim to highlight good investments at the expense of the obviously bad in a way that a FTSE All-Share index fund – with its guaranteed array of future winners and losers – never can. If quantitative investing has any edge on active management in this endeavour, it is in a dispassionate, rule-based approach that is unpolluted by the emotional or behavioural quirks of a human stockpicker.
The screens in these pages, for example, stay invested come rain or shine, and typically rebalance once a year to reapply their rules to an ever-changing market data set. Not only does this reduce the risk of sticking with companies thst take several years to crash and burn, but it avoids clinging onto hard-to-define investing principles such as ‘running your winners’, which can introduce their own biases.
Since it first appeared in this magazine, our High-Yield Low-Risk screen has been one of the most effective at turning this blunt, dispassionate methodology into a source of alpha. Its most recent selections have risen 10.7 per cent on a total return basis since they were selected at the end of March 2022. That was more than double the return from the FTSE All-Share, and the eighth time in 12 outings that the screen has managed to beat its benchmark.
| Name | TIDM | Total return (29 Mar 2022 - 24 Feb 2023) |
| Devro | DVO | 65.3 |
| Burberry | BRBY | 39.2 |
| BAE Systems | BA. | 33.1 |
| Games Workshop | GAW | 26.0 |
| Kingfisher | KGF | 3.2 |
| Rio Tinto | RIO | 1.4 |
| Mondi | MNDI | -7.0 |
| Ferrexpo | FXPO | -16.8 |
| Synthomer | SYNT | -48.4 |
| FTSE All-Share | FTALLSH | 5.3 |
| High-Yield Low-Risk | - | 10.7 |
| source: Refinitiv Eikon Datastream | ||
The long-term track record is more impressive. Since March 2011, the screen has returned 383 per cent, based on a once-a-year reweighting of picks and assuming dividends are reinvested. Although our screens are designed as a source of ideas rather than model portfolios, that headline return drops to 303 per cent if we factor in a 1.5 per cent charge to account for real-world dealing costs. Over the same period, the FTSE All-Share’s equivalent return has been 108 per cent.
Reassuringly, the strategy is also ahead of the market on a 10, five, three and one-year period, which is what investors would likely hope for in a methodology that claims to find stocks that are less risky than the average.
Whether its selections really are less risky is debatable, as I touch on below. However, several of the screen’s tests are interested in long-term consistency – measured over five or 10 years – which I view as a proxy for corporate stability across an economic cycle. Its tests also focus on dividends as a cornerstone of profitability, which is what a lot of UK investors like.
This year, I have kept a tweak to the key dividend yield test, first introduced in 2021, which asks for a forecast yield at least as high as the index median. In the past, this test mandated a historic dividend yield of at least 3.5 per cent, but the past year’s market shift – as well as the cash-balancing priorities of higher yielders – means my slight preference is for analyst estimates of future payouts rather than trailing yields. The full criteria are as follows:
■ A forecast dividend yield for the next 12 months above the median average for dividend-paying shares (the 'high yield' test).
■ A one-year beta of 0.75 or less (the 'low risk' test).
■ Ten years of unbroken dividend payments.
■ Ten years of positive underlying earnings.
■ Underlying EPS higher than five years ago.
■ Underlying dividend higher than five years ago.
■ A return on equity of 12.5 per cent or more.
■ A current ratio (current assets divided by current liabilities) of one or more.
■ Market capitalisation of more than £100mn.
■ Dividend payments covered 1.5 times or more by earnings.
This year, no stock passed every test, although as in previous years, those that scored nine out of 10 have been allowed to pass. In five of the seven picks, it was the ‘beta’ test that obstructed a perfect score, which feels like the market’s way of saying that investors can’t have it all. In other words, the price of ticking those long-term, dividend-friendly boxes right now is higher volatility, which is one reason why I mentioned the screen’s low-risk credentials might be overstated.
The other reason is the reappearance on the list of Ferrexpo (FXPO). With a beta score of 2.8, shares in the Ukrainian iron ore miner are almost three times as volatile as the broader market. That isn’t much of a surprise when you consider the huge operational challenges and logistical issues created by Russia’s war in Ukraine, an ongoing probe into a subsidiary's potential underpayment of royalty payments, and the unrelated-but-nonetheless-unhelpful recent arrest of former chief executive and major shareholder Kostyantyn Zhevago at the request of Ukrainian authorities.
But the inclusion of the miner – one of three 2022 picks that posted a double-digit negative total return – still highlights the blunt nature of the screening process. Until we assign a quantitative risk-weighting system to the screen (and factor in all the judgement calls this might throw up) it’s hard to know how to exclude a stock as high-risk as Ferrexpo without changing the screen’s methodology. So while I am tempted to add a new rule stating stocks with a beta above two cannot be included, I am also hesitant to introduce tweaks that change the results on a case-by-case basis. We’ll know whether this was the right call in a year’s time.
Details of the tests failed are given in the table of 2023 stocks below. A downloadable version of the table, with more details on stock fundamentals, can be found here:
| TEST FAILED | Name | TIDM | Mkt Cap | Net Cash / Debt(-)* | Price | Fwd PE (+12mths) | Fwd DY (+12mths) | FCF yld (+12mths) | 12mth Chg Net Debt | Op Cash/ Ebitda | EBIT Margin | ROCE | Fwd EPS grth FY+1 | Fwd EPS grth FY+2 | 3-mth Mom | 3-mth Fwd EPS change% |
| RoE | DCC | DCC | £4,549mn | -£1,309mn | 4,607p | 10 | 4.3% | 11.6% | 113% | 75% | 2.3% | 9.3% | 6% | 3% | 3.8% | -1.1% |
| RoE | Ocean Wilsons Holdings | OCN | £338mn | -£153mn | 955p | 8 | 6.1% | 12.9% | 64% | 89% | 25.0% | 9.4% | - | - | 7.9% | 17.4% |
| Beta | Robert Walters | RWA | £367mn | £17mn | 490p | 9 | 5.0% | -2.2% | 72% | 49% | 5.1% | 20.0% | 24% | -5% | -14.0% | -12.8% |
| Beta | Kingfisher | KGF | £5,219mn | -£1,840mn | 269p | 11 | 4.3% | 3.8% | 105% | 62% | 6.5% | 11.6% | -17% | -14% | 7.9% | -3.3% |
| Beta | Ferrexpo | FXPO | £869mn | £141mn | 148p | 5 | 5.7% | 4.5% | 8% | 89% | 41.9% | 74.9% | -63% | -51% | -0.2% | 94.3% |
| Beta | Hilton Food Group | HFG | £619mn | -£470mn | 692p | 13 | 4.1% | 9.3% | 26% | 66% | 1.8% | 9.4% | -31% | 23% | 30.6% | 9.7% |
| Beta | Norcros | NXR | £187mn | -£84mn | 210p | 6 | 4.4% | - | 271% | 39% | 8.0% | 16.8% | -6% | -8% | 10.5% | -4.0% |
| source: FactSet. * FX converted to £ | ||||||||||||||||
