Left to my own devices, the Stockpot column would feature endless variations of value and income investment strategies.
Stockpot’s one experiment with momentum trading was illuminating but not successful.
There was certainly plenty of momentum, but most of it was of the going backwards variety, at a rate of knots that would have given even the most hard-core adrenaline junkie the fits.
You can read the grisly details by working through the Stockpot articles and clicking on the ones that have the words “Bombed out” in the title.
Those articles detail how, over a period of less than six months, I ran a virtual portfolio and turned £10,000 of (thankfully) imaginary money into less than £5,000 by trading stocks that were “bombed out” but ostensibly on the comeback trail.
There was a period, however, when the portfolio performed sensationally, rising 46% in just 11 days, and that performance was interesting enough to make me wonder whether a few tweaks to the stock filter to eliminate some of the obvious “dumb buy” stocks would make a recovery stocks portfolio viable.
You can read the original ideas behind the “Bombed Out” portfolio by clicking on the link below.
Essentially, the idea was to take stocks that had fallen by more than third over the last year and then from that universe of stocks identify those that had been on a rising share price trend over the last month.
Now, most “virtual portfolio” investment workshop articles in the media ignore the fact that when one trades shares, one buys at the “ask” price and sells at the “bid” price.
(One has no idea why one has started talking like Princess Anne).
When I set up the “Bombed Out” portfolio, I decided to factor in bid/ask (or bid/offer) spreads and dealing costs into the experiment, largely because I knew the stock screen would be recommending lots of buys and sells each week, and I wanted to demonstrate to any novice investors out there that trading frequently is a quick way to lose money.
Making it work is a "big ask"
In a nutshell, what made the performance of the “Bombed Out” portfolio so disastrous was not the share price movements but the wide bid/ask spreads of so many of the companies that ended up in the portfolio.
A lot of the companies generated by the stock screen were tiddlers or minnows, and these stocks tend to have very wide bid/ask spreads because they are not heavily traded, and a market maker (the party setting the bid/ask quotes) likes therefore to have a wide spread to provide extra insurance against having a big pile of stock on the books that can’t be sold.
As an example of how wide these spreads could be, some of the stocks that ended up going into the “Bombed Out” portfolio needed to rise 20% or more just to break even.
Adding a criterion to the stock screen that limited the universe to companies with a market value of more than, say, £100mln, would probably have eliminated almost all of the companies with crazy spreads, but would also have eliminated some of the portfolio's big winners, such as MySQUAR Ltd (LON:MYSQ) and Bilby PLC (LON:BILB), both of which have a market capitalisation of around £27mln.
So, this time round – yes, we are reviving the recovery stocks virtual portfolio idea – I thought it would make more sense to exclude any stock with a bid/ask spread that is more than 5% of the underlying stock price.
In other words, no stock would be showing a loss of 5% (excluding assumed dealing costs of £15 a trade) the minute it joined the portfolio.
Changing the sell trigger
In the original “Bombed Out” portfolio, a mechanism was needed to indicate when would be a good time to sell, and I chose the 50-day moving average.
Specifically, when a stock moved below its 50-day moving average, it was jettisoned from the portfolio.
The 50-day indicator was chosen because it seems to be very popular with technical analysts who spend their time gazing at charts looking for semi-mystical signals with all the fervour of ancient astrologers.
I don’t want to place the blame for the “Bombed Out” portfolio’s failure on technical analysis and the arguable reliability of the moving average as a genuine buy/sell signal, because as we have already established, the main cause of the underperformance was the crazily wide spreads, but this time round I thought we would try our luck with the relative strength index (RSI) as a buy/sell indicator.
This is another popular indicator with technical analysts, and is generally calculated over a moving 14-(trading) day period on a scale of 1 to 100.
The indicator is said to be a good guide to whether a stock is ‘overbought’, which is to say people have piled in too enthusiastically into a stock that has been going up, or ‘oversold’, which is where people have been panicked into selling a stock that has already taken a dive.
People who have done a lot more research into this than me – some 7.5 billion of them at the last count – reckon that an RSI of more than 70 means a stock is ‘overbought’ while a stock with an RSI of less than 30 is ‘oversold’.
With the revamped recovery stocks virtual portfolio, I plan to sell any stocks in the portfolio that have an RSI of 70 or more, and also to ignore any potential new candidates that have an RSI of 70 or more.
Because the stock screen will, as in the original “Bombed Out” portfolio, be looking for stocks that are up at least 2% in the last week, 4% in the last fortnight and 10% in the last month, I doubt there is any need to stipulate any rules on ‘oversold’ stocks.