Posted by: Aaron Pressman on September 24, 2007
With behavioral or psychological stock market analysis in vogue, many researchers have concluded that investors systematically under-react to news events like a jump in corporate profits or a dividend cut. Because the initial market reaction is small, there’s theoretically an opportunity for savvy investors to jump in. Some these anomalies have even become the underpinning for quant-like exchange-traded funds. But beware of Greeks bearing gifts, as Virgil once said. A lot of the effects vanish when looked at carefully.
One of the landmark papers in the field, by Louis Chan, Narasimhan Jegadeesh and Josef Lakonishok, in 1995 found that investors react too slowly when companies report surprisingly good earnings. They found, especially for stocks that were already performing well, an earnings surprise was a reliable indication of further outperformance over the following year.
There’s an exchange-traded fund that operates, in part, using this principal. The Zacks Small Cap Index fund (Symbol: PZJ) uses price momentum and earnings surprises along with a few other criteria to pick a basket of small cap stocks. It’s outperformed the Russell 2000 Index by a couple of percentage points over the past year, though it’s lagging the index in calendar 2007.
But are there other kinds of news that can be reliably used to pick stocks? Some have concluded that there is a similar opportunity in dividend announcements, particularly dividend reductions and ommissions. If investors failed to appreciate the significance of this negative signal immediately, it might offer the same kind of trading potential as the earnings surprise.
But a new paper by finance professors Yi Liu of Capital University, Samuel H. Szewczyk of Drexel University, and Zaher Zantout at the American University of Sharjah that will be published in the Journal of Finance says those earlier studies were just seeing another manifestation of the earnings effect.
The professors studied 2,337 dividend reductions or omissions between 1927 and 1999. At first, the data looked promising. They compared the performance of each company reporting a dividend misstep with a company in a similar industry and of similar size. At first they find a seemingly significant effect — the dividend cutters underperformed their peers by 6% to 15% over the next year, depending on which matching criteria are used. The effect was most pronounced for small companies and hardly visible for stocks of big companies.
But many companies that cut their dividend are also suffering from poor earnings. When the professors compare their bunch of dividend cutters to firms that had similar earnings performance but didn’t cut a dividend, the effect evaporates. “Our result indicates that firms that reduce or omit their dividend have the same post-event price drift as firms that have similar earnings performance but do not reduce or omit their dividend,” the concluded.