Since inception, the Internet has offered inves-tors a forum in which to compare notes and discuss ideas. But the growth of social networks such as Facebook has taken this concept to new levels with so-called “expert investment communities” — groups of experienced investors who share their analyses of the markets with anyone who is interested. Network providers employ different models, but all have the same basic purpose — to allow investors to bounce ideas off others before making buy and sell decisions.

On www.cakefinancial.com, for instance, Facebook users get updates when anyone in their chosen network buys or sells a stock. They can also track friends’ portfolio returns, trade stories and post messages.

What distinguishes many of these newer sites from older online forums is transparency — participants are often required to link their online community accounts with their actual investment accounts. The point with these new sites is to provide a platform that verifies actual performance and identifies outperformers.

For instance, those who sign up at sites such as www.covestor.com or www.vestopia.com are encouraged to write about their trades, explaining the strategy backing each move. Top performers across different strategies and risk profiles are then ranked on a variety of criteria, allowing other inves-tors to peek over their shoulders and hopefully improve their own odds. These followers track members’ trades both on the site and through email notifications, mimicking them as they choose.

Although these two communities are composed of both professional and armchair analysts alike, there are few barriers to entry. Some other sites, however, are a bit more discriminating.

When Columbia Business School professor Joel Greenblatt created www.ValueInvestorsClub.com a few years ago, he capped the membership at 250 people to keep things exclusive and avoid some of the nonsense found on many existing financial message boards. With a small attrition rate as members drop out, the forum has flourished.

To be accepted, applicants must submit a detailed, minimum 500-word analysis of their favourite stock pick, which is judged by Greenblatt. Members are required to post a minimum of two picks per year, and each stock idea is rated by members on a scale of one to 10.

While not all the ideas posted on ValueInvestorsClub.com will necessarily be winners, participants expect that most will pan out over time. Non-members, who can read ideas on the site on a delayed basis, can also piggyback on the highest-rated ideas from some smart, experienced investors. But should they bother? And what can they expect to learn about value investing in the interim?

In a recent paper entitled Fun-da-mental Value Investors: Characteristics and Performance, University of Chicago PhD candidate Wesley Gray and the University of Missouri’s Andrew Kern examined the investment decision process and aggregate performance of the professional value investors who participate in
ValueInvestorsClub.com. The conclusion? By focusing mostly on mismatches between stock price and intrinsic value, the pros get it right more often than not.

Using a sample of about 3,000 investment recommendations by forum members from January 2000 through June 2008, along with relevant stock return data, the researchers concluded that most of the site’s value seekers generally employed no more than three criteria when making decisions. None of the recommendations seemed to use statistical asset-pricing models from the academic literature.

Instead, picks focused on variables such as high book-to-market ratios, and assigned much more importance to factors such as discounted cash flow and earnings multiples. Most recommendations (87%) focused on long positions concentrated in small companies with a slight value tilt. The balance of suggestions opted for short positions in growth stocks.

Most member recommendations paid off for holding periods of up to one year, the study found, but declined beyond that. Interestingly, the study also found that the short positions generated positive returns over a longer time horizon. IE