Follow These Canny Share Buyers

New research suggests it could be worth following company directors.

At last academics have some encouraging words for stock pickers! They may have won Nobel prizes proving it's pointless trying to beat the market, they may have nothing nice to say about chartists, but when it comes to following director share buying, at least two professors have shown it's potentially worthwhile.

The University of Exeter Professors Alan Gregory and Ian Tonks have published research showing company directors tend to buy low and sell high -- the Holy Grail of investing.

Studying every director trade between 1986 and the end of 2003 -- and discarding certain sectors, for reasons which aren't immediately obvious, as well as smaller trades -- they found insiders tend to buy their company's shares when the share price is more lowly rated by the usual fundamental metrics, such as P/E or book value. They also found directors tend to sell after price rises takes the share into what the professors term 'Glamour' territory, when these value credentials have diminished.

It proved a good strategy. The strongest outperformance was found with smaller companies, where directors' buys beat baskets of similar firms' share prices by as much as 20% over two years. The abnormal return was not as great for large cap value stocks, but it was still 6% over the period.

As for director selling of their company stocks when the value is evaporating, the professors found any outperformance was "not generally significant".

Private investors rejoice!

Following directors' deals has long been a popular strategy among private investors. In the early 20th Century, when regulation of the markets was incomparably lax compared to today, it would probably have made you rich.

Even today when we're regularly told the market is efficient and that such opportunities were long ago traded away, newspapers, magazines and numerous online data providers all catalogue the latest director trades for share pickers to pour over.

The theory is that company directors know the value in their business better than anyone -- and this research appears to show it's true.

A quick way to a quick buck?

Still, it's one thing to mine data for an apparent discrepancy you can profit from, but another to do so in practice.

For one thing, I don't know why the professors excluded investment trusts, property companies, insurance companies and banks from their research. It may have something to do with different fiduciary duties, or it may be they didn't fit the data…

There's also the practical issue of actually buying at the price the directors do. In my experience, director buying almost always moves the price upwards (suggesting the market is well aware it's a good signal) and in bullish conditions it may never fall back again.

Unless you know in advance when the director will buy -- which you can't -- you can't expect to follow the logic of the academic study to see the same outperformance. By the time you've drunk your second coffee on a Saturday morning and turned to the relevant section of your newspaper, the price could be far from what the director paid.

To explore this, I've conducted my own entirely un-scientific study of the seven director deals I followed up for The Fool this year.

In each case I originally researched the share as a direct result of director buying, at intervals ranging from a few hours to a week or so after the trades took place. This isn't what the two professors tracked, but it's how most private investors treat such information.

I wanted to see three things -- how well the directors did (presuming they held until today), how a reader buying on my article would have done (not that I thought all the shares were cheap), and what the FTSE All-Share returned over the same period to today's 2,691 level:

I make the average director average gain 56%, but the reader one 29%. The FTSE All-Share average gain over the same period was 20%

Conclusion: No shortcut to riches

The good news, superficially, is that a reader buying all the shares I wrote about would have beaten the market, with an average 29% gain versus an average 20% gain for the All-Share.

But there's bad news, too.

Firstly, most of the outperformance is down to two shares: Petra Diamonds and Entertainment One. I almost bought the former but thought the latter too expensive. In the end I bought neither -- a big mistake!

Secondly, the directors did much better than readers. Only Charles Taylor Consulting and Business Post Group/UK Mail offered readers better prices than the directors' paid, and neither company has shot out the lights since.

Of course, seven share write-ups hardly covers all directors deals done since April, and even the longest holding period is much shorter than the two years cited in the study.

But personally, I'm satisfied that while significant director share buying will always be a good reason to look at a share, particularly if it's a small cap, blindly buying after reading about a director's purchase with no regard for price or fundamentals is unlikely to see you beat the market. And as my caution with Entertainment One showed, looking at the fundamentals is no guarantee of success, either!

Not easy, this stock picking game.

More from Owain Bennallack:

• Treat Yourself To This Investment
• Google Gatecrashes Rightmove's House Party
• This Moat Has Been Breached