Areader wrote to me recently to ask how I decide when to sell a stock. It’s a good question, because investment columns do tend to dwell much more on the stock selection process, rather than the equally essential decision of how to exit gracefully.

The bare truth is, of course, that most exits are made when stocks have in some way or other disappointed. You don’t exit from winners, you hold them. Indeed, running winners and selling losers is one of the basic tenets of investment.

However, selling losers is probably the single most difficult proposition that investors face, because of a psychological tendency (from which I am far from immune) to avoid admitting mistakes.

Theory makes it sound so simple. All stocks in a portfolio should have a stop-loss through which a fall will enforce a sale. The troublesome word is “enforce”. Share price movements are not always neat, so a 10 or 15 per cent fall, or a drop through a trendline, does not necessarily mean anything is wrong with the stock itself.

If you sell without finding out why a share price is falling, you could dump a good share during a rare moment of weakness and lock in a loss. But, conversely, if you don’t sell until you do have a solid reason to do so, you could easily halve your investment – or worse.

All this is particularly true of the thinly-traded shares of the Alternative Investment Market (Aim). Over the past couple of years, there have been quite a handful of them that have caused me losses.

Aim-listed agricultural company Landkom (dealt with in this column last September) and eastern- European property outfit Carpathian were two that I held too long in the face of gradual and unexplained declines. So, when the bad news did strike, I was already more than 50 per cent down. The same is true of property refitting firm Your Space plc, although at least I had cut my stake by two-thirds at the end of 2008.

You usually have more information to go on with the larger companies, so stop-losses can be enforced more quickly.

With oil company Dana Petroleum, it was clear that the shares were gradually losing their premium rating over peers. So, after having held for a year, I sold out last October for a loss of five per cent. That decision was made easier because there were other sector companies, such as BP, offering better value and hefty dividends.

But, even with large companies, there are times when I ignore the stop-loss. Such has been the case with hedge fund group Man. I bought the shares last October for 360p, partly because of the dividend yield which then exceeded 8 per cent, but mainly in anticipation of a bounce-back in investment performance. My timing was poor. I had already missed the best of the recovery, which ran months ahead of any positive news, and the expected improvement did not arrive. A poor performance by the company’s flagship AHL fund and continuing withdrawals by investors have dragged the shares down to 220p, a 38 per cent fall from my buy price.

So the question is: would I buy Man now, were I not already aboard? The answer is, tentatively, yes. It is a notoriously volatile stock and periods of bad performance have in the past been followed by strong recoveries. The bigger question is whether the investment model is broken, and I don’t think it is. If I’m wrong, I will lose more. In the meantime, though, Man is holding its 12 per cent dividend yield – and that remains a comfort.

Nick Louth is an active private investor, writing about his own investments. He may have a financial interest in any of companies, securities and trading strategies mentioned.

Copyright The Financial Times Limited 2018. All rights reserved.

Comments have not been enabled for this article.