Buying shares: where do you start?

Alternatively, it's OK to rely on serendipity - a company's name hoves into view for whatever reason, then give the company and its share rating the once-over. If it looks interesting, delve deeper. Relying on serendipity might seem casual, but the random nature of so much in the stock market means it's not obviously worse than any other approach.

However, here's one exercise that I use maybe twice a year. It's very simple and it helps equity investors do what they should be doing anyway - buying shares when they are still deeply unpopular and selling them when they are too popular. For London's listed equities, I ask a database to generate a list of all share price movements from one, three and five years ago up to today. Then I display the list based on one of the time periods and in ascending or descending order (it does not matter which period or which order).

So the opening list is a much longer version of the table shown here, which extracts the most interesting shares. As I scan the full list, I am looking for two features - first, the acceleration of a trend; second, a change in trend.

Acceleration of a trend is shown quite neatly by the performance of shares in cleaning and maintenance services supplier Mitie (see table). At 281p, the price is up 22 per cent over five years, 24 per cent over two years and 36 per cent over a year. The implication is that if - a big if - the share price is faithfully reflecting progress in the company, things are getting better and better, so the shares may still be good value. But, given the direction of the London market, currently it's easier to find the trend pointing to a decline that will get steeper before it bottoms out. Maybe none is starker - or more ominous - than shares in Homeserve, whose business model may be as broken as the household pipes and drains its insurance policies are supposed to cover. Certainly, that's what the share price data implies.

However, it's more interesting when a trend changes direction because it helps equity investors spot unpopular shares in companies that are starting to get it right. From the data that I trawled through, the table majors on this feature. Interestingly, not many FTSE 350 constituents show this pattern, the best two are Qinetiq and Dixons Retail.

Qinetiq, which provides varied training, testing and buying services for the defence industry, lost its way as it reduced its reliance on the UK's Ministry of Defence. But it seems to be getting itself back on track and the change in direction of the share price looks well established. Full-year results next week could put the finishing touches to the share price recovery. Meanwhile, in the case of electrical goods retailer Dixons, the revival is in its infancy and, therefore, not established, but it could have far to go.

However, the four to which I would give more attention are Xchanging, which provides white-collar services, such as IT and buying, to other companies; fund manager Liontrust; training and education software and services provider Tribal; and PuriCore, whose products kill germs for healthcare and food-industry customers. The reasons are because in all four cases the breakaway from the share price trend looks arresting; also because I know two of them vaguely (Tribal and Liontrust), which is a head start that should not be ignored; and because the other two - Xchanging and PuriCore - seem to be getting their act together. At Xchanging, cost savings combine with a cash-rich balance sheet yet the shares trade on a 'smoothed' PE ratio of only 10 times. Meanwhile, at PuriCore sales in the first quarter of 2012 are racing ahead and the company - which, admittedly, has been thoroughly lossmaking for each of the past five years - is generating cash.

Use the table by all means. Some readers may prefer the look of, say, 888, easyJet or McKay Securities. But, if you do, remember this: the list is about as far removed from a recommendation as it's possible to get. It's a preliminary stage - just the start of the start really.