Exit strategy

True, experience tells us that in the investment world nothing potters for too long. If there appears to be equilibrium, something will disturb it. Even so, the income fund looks to be as comfortably balanced as I could wish for - there is room for just one new investment and no current holding is threatened with eviction.

That's a cue for the question: why do I evict holdings? First, it does not happen often. In all things, Bearbull is a firm believer in the notion of ‘less is more’ - and that certainly extends to running investment portfolios. Trading is altogether too much effort for an uncertain reward. So I do as little of it as I have to and invest for the long term. You can see that from the time I have held some of the fund's investments. Of the nine holdings, the average holding period - for what that's worth - is approaching four years. The oldest holding - GlaxoSmithKline - has been there for almost 13 years (though it was topped up in 2007 and again in 2009) without ever adding much value to the fund.

Granted, there is a sensible case for saying I should have worked harder to find a better home for the capital tied up in Glaxo. Yet that chunk of capital always produced an acceptable income yield. It started out at 4 per cent back in 2000 and is now running at 5.3 per cent of the holding's present market value.

Besides, Glaxo's share price never fell through its stop loss - though it ran close to it for quite a while - and that's usually the deciding factor. Let's state that as a rule: I sell a holding when - and only when - its price falls through its stop-loss level. In over 14 years of running the Bearbull Income Portfolio, there have been few exceptions to that rule. So let's explain how I use stop losses, with topical reference to shares in animal feeds processor Carr's Milling (CRM), currently the best-performing holding in the fund.

Following results for the year to end August, which, I confess, I haven't looked at yet (‘less is more’), shares in Carr's stormed through £10. That prompted the fourth upward shift in the Carr's stop-loss level since I bought the shares almost four years ago. Back then, I paid 440p per share and set my usual stop-loss boundaries - plus and minus 20 per cent. That meant that if the Carr's price fell 20 per cent to below 352p, I would sell. But if it rose above 528p, I would revise the stop-loss levels. So in April 2010, when the share price hit 530p, I re-set, with 424p as the selling trigger and 636p as the level where I would once more revise.

And so on. The aim - obviously enough - is to run with shares while their price is rising, but to limit losses if it starts to fall. When a share price rises fast - and instinct tells me it's starting to get ahead of events as investors get too excited - then I will compress the stop-loss boundaries. I've just done that with Carr's Milling. With its latest base of 1,010p, its boundaries are now plus and minus 15 per cent. So I'll sell if the price drops below 859p and re-set if it tops 1,162p. I could even compress the boundaries asymmetrically in order to capture more of the upside while limiting the downside (and occasionally have done that). That might mean limiting the upside to 10 per cent before I re-set while keeping the ‘sell' trigger at minus 15 per cent.

True, one effect of this method is that I will never get out at the top. But investors should be wary of trying to do this anyway as many aphorisms encouraging them to ‘leave something in the price for the next guy’ attest. The plain fact is that no one is bright enough to know when the top has been reached. Much better, therefore, to exit when there are still punters wanting to get on board.

The major negative of this system is that, especially in a bear market, it can give a sell signal before a security's price has been able to reflect the merits of an investment proposition. Still if I am really confident this has happened I can always override the rule. But that's rarely the case.