The idea is that parallel lines extending into the future are drawn on the price chart of a share - the top one some way above its current price, the bottom one some way below. If subsequent price movements take the share price through the top line, then new parallel lines are drawn on either side of the higher price. If the share price falls through the lower line, the stop loss has been triggered and the holding is sold. Hence, the box metaphor - if the price falls through the bottom of the box, the shares are sold; if it goes through the top, a new box replaces it.
The big question then becomes: how far beyond the share price should the top and the bottom of the box be set? Initially, I fix the upper and lower limits comfortably distant from the price I have paid for a new holding - I want the share price to have plenty of room to wiggle about within its box. The upper and lower limits are also often asymmetric. Pretty consistently, I apply a plus 20 per cent gain to the upper limit, but I will leave more room on the downside - at least 25 per cent initially, more if I think the share price is likely to be volatile. That way, I limit the likelihood that I will find myself being told to sell a newish holding before whatever investment merits I think I have spotted have had the chance to do their work.
Thereafter, the process is to continue shifting the parallel lines upwards as the share price rises. In that way I can run my winners. However, I also want to maximise my profits. So, for shares that have risen far - especially if they have also risen fast - progressively I tighten the upper and lower levels.
Take shares in milk processor Dairy Crest, which was a big winner in the Bearbull Income Portfolio some years ago and for which I have kept detailed records. Having bought the shares at 145p each, I set the upper and lower limits of 174p and 108p respectively. But the share price gathered momentum and I re-set my hold-or-sell limits as the price passed through 181p, 226p, and 279p. After that, I limited the wiggle room that I gave the share price. But it still scampered upwards - so I re-set at 331p, 372p, 417p, 475p and, finally, at 532p. As it turned out, the share price peaked at 566p in May 2002. Not that I sold at that price. I was being dictated to by limits set when the price was 532p and these told me to sell if the price fell through 478p. In the event, Dairy Crest's share price dropped sharply in the second half of that year. I sold at 462p and the price was 350p by the end of the year.
Remember that while Dairy Crest's share price was surging upwards, I had no idea where it would go to. I simply wanted to be along for the ride, or to alight if the ride switched to reverse. But that's the great merit of using the box method. Because we have no way of anticipating future price movements, we need a simple plan that can secure most of the profits, while limiting losses if the price starts dropping too far.
Granted, the system's merits may be better attuned to a bull market than a market like the current one. Even in a bear market, however, the system clearly has one big merit - if stop-loss triggers are obeyed, it turns invested portfolios increasingly into cash as the market falls, thus limiting losses. That is why, for example, the Bearbull portfolios have ridden out this bear market quite well so far.
Yet, obviously, it comes with a cost - if the London equity market were to recover strongly, then someone using the box method conscientiously would miss the bounce. But that's true only if this method were being used in isolation. However, it should be just one gadget in an investor's toolbox, which should also hold the means to seek out attractive situations even as other holdings are being dumped.
Besides, you can make exceptions. For example, when I returned from 10 days away in mid-October to find most holdings in the Bearbull funds blaring out sell signals, I had a simple choice - realise losses that were far greater than the stop-loss system had allowed for, or re-set the upper and lower limits on all holdings to take account of the new reality. I chose the latter course.
And just last week, I ignored a sell signal from my holding in Chesnara in the Bearbull Income Portfolio. True, the market is fretting that Chesnara's investments will include hidden exposure to the toxic waste of the credit crunch. Maybe. Yet it seemed odd to sell the shares soon after Chesnara's management had launched a share buy-back programme. You could hardly imagine a stronger signal that Chesnara's dividend, which produces a 14 per cent yield with the shares at 105p, is going to be maintained. Check it out.
