Screen of Screens - beating the market with blended investment strategies
Of the 60-plus investing strategies tracked by Stockopedia, one of the consistently best performers in 2013 is one that really couldn’t be simpler. By picking those companies that qualify for the highest number of all our other long-only guru investing models, the Screen of Screens has produced a remarkable 36.4% return so far this year. So what’s the secret behind this potential profit factory - and can it be improved upon?
Ahead of the London Investor Show this Friday, when a lot of these strategies will be up for discussion, it’s worth recapping on a pretty surprising performance by the SoS this year, given that it can be a rather blunt instrument. After all, selecting shares just because they qualify for at least four other strategies (that may even appear to be at odds with each other) could seem tenuous. But look a little deeper and you’ll find an approach that takes an impartial view of investment style and manages to smooth out bumpy market conditions as a result.
Blending investment styles
It’s worth noting that our returns for the SoS are based on quarterly rebalancing but take no account of the trading costs that those portfolio changes might incur. Even so, there’s little doubt that it has a formidable record for finding hot stocks. And part of the reason is that it embraces the fact that one investor’s value stock can be another investor’s quality pick, income generator or momentum play. In fact, as Ed Croft will talk about at the show, the right blend of these approaches can be a highly effective way forward (if you’ve got the right tools).
Take Regenersis (LON:RGS), the AIM quoted company that repairs consumer electronics like mobile phones and laptops for big brand manufacturers. For most of this year it was only really flagging as a momentum stock, which wasn’t difficult for many small cap shares given the upward trajectory of the market. But in early October it qualified for a couple of growth screens in the form of James O'Shaughnessy's Cornerstone Growth (38.5%) and Jim Slater’s ZULU Principle (40.0%). Simultaneously, it made it on to the Charles Kirkpatrick Value screen (47.5%) and continues to be rated as a momentum stock as well. As you can see from the individual returns from those screens this year, each has delivered significant gains. So the appearance of Regenersis on the SoS is interesting because the rising price that might have made it interesting to momentum traders has now pushed it on to screens that focus on growth and value characteristics too. (I mentioned Regenersis in the recent Jim Slater article. I don’t hold any position in the company and, on that note, it’s essential to Do Your Own Research on stocks appearing on Stockopedia’s GuruModel screens !)
Another of the ways of reading the SoS is to see how qualifying companies change gear through different investing styles. Matchtech (LON:MTEC), the recruitment company, made it on to the SoS in mid-July just a couple of weeks before a trading update that sparked a re-rating in its price. But even as the company teeters on the edge of its 52-week high, the screens that it’s qualifying for are changing; in recent days it has appeared on Bill O’Neill’s CAN-SLIM growth and momentum model. For investors concerned about the scale of that recent price rise, the progression of Matchtech onto the CAN-SLIM screen may be some comfort - O’Neill himself gave little regard to P/E ratios, preferring instead to focus on the future.
Just add science
It’s fair to say that bullish sentiment this year has been kind to many investing strategies (a rising tide, etc) and that the SoS deserves a thorough test over the months and years ahead. Indeed, while the screen offers some really interesting insight into which stocks are ticking the boxes of various investment styles and models, it’s still fairly unwieldy (not least because shares can easily slip on and off quite quickly). So be warned - its performance may well take a turn for the worse at some point.
However, the underlying principle that great stocks can be found using a combination of investing styles really is the secret sauce behind the SoS. And with that in mind, the new StockRanks tool on Stockopedia really ratchets up the science behind the general concept. It does it by ranking the entire market for companies achieving the best composite scores based on their quality, value and momentum characteristics. In the case of Matchtech, for instance, a StockRank of 99 puts it in the top 1% of the market based on those QVM factors.
Until now, the SoS has offered some intriguing evidence that you can beat the market with stocks that score well across multiple investment models and styles. With StockRanks, investors can take a scientific approach to ranking the entire market in order to spot gaps and profit from opportunities that others fail to see.
Disclaimer
This is not financial advice. Our content is intended to be used and must be used for information and education purposes only. Please read our disclaimer and terms and conditions to understand our obligations.
8 comments
Was the screen of screens rebalanced on the 15th March?
Cadmunkey - Yes it was rebalanced equally into the qualifying stocks at the closing prices of 14th March.
Ed - thanks for linking that Mechanical Bull blog. Interesting to read... I'm intending to utilise something similar in the near future. My main issue currently is having enough capital to invest across a sensible number of holdings, but this should improve in the not too distant future.
I'm interested to see how such a strategy performs over the long term; in my eyes it does seem like a relatively prudent approach that can actually deliver very good returns.
Theres a few warnings Id like to make on following the screen of screens.
1. In the simulation theres no position sizing taking into account the stocks volatility, a lot of the stocks are small cap and very volatile so in a bull market you would expect the rewards to be greater , in fact you would need the reward to be greater. A better benchmark would be the FTSE allshare
2. Theres also no costs and stamp duty in the simulation
3. I presume the start date and consequent rebalancing anniversaries are also arbitrary , different dates could produce very different results
That said Im fascinated by it. Im currently tracking a watch list of QVM stocks, Ive taken anything with 99 or above as that also seems to be very similar in principle to the screen of screens. Do you have any plans to track something like that as well
Hi Tim,
Yes we've taken the same approach to these trackers as AAII have in the US, which is to make them tax and transaction cost free. Clearly in a real portfolio the returns would be less. Currently the average spread across Screen of Screens stocks (excluding the 2 sub £ 10m microcaps) is 60 basis points which is not that wide - especially versus a lot of the value screens. If one rebalanced twice per year at that spread it would only add up to 1.2% spread costs. There are 7 stocks with greater than £1bn market cap and 1 beneath £100m. Personally I believe the £100m to £1bn range is the best stock picking range for private investors willing to stomach some volatility in return for greater returns. The other stock market anomaly we haven't put in the QVM ranking system is 'Size' - the fact that small caps tend to outperform large caps - institutions in general can't focus on it but smaller investors are perfectly positioned to take advantage. The Screen of Screens does have a small cap bias, but again it's not as extreme as many strategies - such as the Piotroski screens which tend to pick up a lot more microcaps.
We are working on extending the rebalancing module, to incorporate more risk constraints - to enforce maximum exposures to certain sectors, put in place bid/ask spread limits and so on. This will I hope be live by year end or early 2015. At that stage we'll be looking to throw the tracking system open to all subscribers - in order to model your own strategies and I hope one day to print off deal sheets to those taking a systematic approach.
We won't be tracking the StockRanks in the same way as the screens - i.e. with just the top 25 stocks tracked. They are designed more as decile ranking systems - we are interested in how the deciles perform over time against each other and have an initiative to start tracking them as deciles. We'd expect much more volatility of returns across each of the precise 'centiles' than the deciles. Of subscribers who have been tracking returns - this blog by Mechanical Bull illustrated the top ranked QVM stocks performance last year. http://bullmechanic.blogspot.co.uk/2013/12/mechanical-bull-portfolio.html
We will be launching 4 model portfolios at some stage soon which will be partially constructed using the StockRanks. They will span across several of the core 'anomalies' of stock market returns - most specifically we have in mind a Quality+Value portfolio, a Deep Value portfolio, a GARP portfolio and an Earnings/Price Momentum portfolio. These core approaches will I hope act as illustrations of how to construct personal portfolios and also could be combined into a blended style approach.
Hope that helps,
Ed
Thanhs Ed - look forward to reading your research. I have isas and sipps set up for both me and my wife so we have a far amount of tax free funds to trade with. Everything is done via Halifax share dealing. Every month or so they run a promotion where the commission fee is reduced to 3.95 for two hours. I do nearly all my trading in that window. It both saves me money and forces me to stick with my approach of strategic ignorance.
I am very mindful of over trading but I think this kind of strategy can help with this. It forces me to hold stocks foran average of at least a year, a rule of thumb I have borrowed from Greenblatt. However the strategy provides a safety valve for stocks where the numbers rapidly become less favourable. Conversely it can allows you run winners for longer.
The other thing to say is thay the SoS seems to screen out a lot of the microcaps with large spreads. I haven't yet encountered issues here.
Anyway thanks for providing this fantastic product. I couldn't live without it!
I've been following the screen of screens pretty closely since the beginning of the year. Around six months ago I also noticed 4-5 gold miners were coming up in this screen. These did not seem very attractive investments, so this got me thinking about a way of filtering these out.
It was also around this time that I stumbled over the composite rank (now called QVGM rank) in Stokopedia. It seemed to me that both the SoS and the composite rank were applying the same basic principle in different ways, i.e. identifying stocks with good all round strengths. I reckoned that combining the two could be a good a way of selecting the very best all rounders. My approach was simply to add the two together. For example, DTG currently has a QVGM rank of 100 and appears in 9 Screens giving a total score of 109. Effectively, the approach uses the SoS to select from the highest scoring candidates in the composite rank. Importantly, this had the effect of filtering out those dubious miners.
The idea then is to buy within the top 6-8 stocks every month with the aim of maintaining a portfolio of around 20. Each month, the lowest scoring 1 or 2 stocks in the portfolio are sold and used to purchase more top scoring stocks.
It is a very simple mechanical approach, but it seems to be pretty effective it filtering out duds. The worst performing stock in the current portfolio is down 4% (which I only bought 2 weeks ago so it is still early days. Everything I have held for any length of time has increased in value. In five months the portfolio is up 24% compared with 0% for the FTSE 100 and 7% for the FTSE 250 and 20% for the SoS. The other thing I like is about incorporating the QVGM rank is that it provides a degree of stability that the SoS lacks and provides a basis for deciding what to hold and what to sell.
Obviously it is early days (I am only just starting to sell stocks in the portfolio) but I am very encouraged by the results. The next thing to look into is whether I should replace the QVGM with StockRank.
Hi nswier - we've been doing a lot of testing of QVGM vs QVM (StockRank) and our tests are proving that QVM is a more robust methodology - we'll show some results on Friday at the show and publish them here soon. It's partly due to the higher value weighting but also due to the fact that Growth isn't that predictive of future returns (a finding that O'Shaugnessy also corroborated). Value & Momentum blended with a quality filter seems to be the more optimal approach - which tallies with the academic thinking too.
Reading your approach - it's important to be careful about transaction costs. It's v. v. hard to get the balance right and all depends on tax, portfolio size and other personal factors. But certainly minimising the number of transactions each year can save a lot of money - especially in smaller cap companies.