Pair Says Forget Bulls and Bears; Look for the Dogs

Two years into the brutal Japanese bear market, the U.S.-based Nikko Japan Tilt stock mutual fund had shrunk from nearly $20 million in assets to about $2 million. By late last year, the fund’s managers were understandably receptive to the idea of a new approach.

Enter the “Japan Return Reversal Strategy”: Pioneered by two American business professors, the strategy held that the best way to play Japanese stocks--in bull or bear markets--was essentially to buy the issues that had performed worst in prior years.

If this sounds like simple market “contrarianism”--buying the stocks people hate most at any given moment--it is basically that. But economics professors James Poterba of MIT and Lawrence Summers of Harvard, who developed the strategy, also set out to discover whether a specific time period of lousy performance would yield the highest future returns.

There was indeed a magic number, the academics discovered: eight years. By reshuffling a portfolio of Japanese stocks quarterly so that you owned only the 25% of issues that had fared worst in the prior eight years, you could have far outpaced the Japanese market’s Topix index of 1,200 stocks between May, 1986, and July, 1991.

In that period--covering 3 1/2 years of Japan’s bull market and 1 1/2 years of its bear market--the “buy the dogs” strategy would have returned 387%. In contrast, the Topix index rose just 48.5%.

No wonder the Nikko stock fund’s beleaguered managers embraced the concept. The fund, since renamed Capstone Nikko Japan fund, shifted to the Poterba/Summers formula on March 17 of this year, with refinements: Stocks that might have fit the eight-year-loser profile were eliminated if the companies were severely distressed. By applying that kind of screen, Capstone Nikko filtered a final list of 50 to 70 high-quality yet ignored stocks for its portfolio.

Once in place, the strategy initially worked as well in real life as in the hypothetical model.

Between March 17 and June 30, as the Japanese market continued to sink, the Capstone Nikko fund lost 4.7% versus an 8.4% Topix loss. While the fund lost ground, losing less than the market index is a major victory in money management.

But when the Japanese market began to rocket in August, the stocks in the Capstone Nikko fund lagged. Result: The Topix index now is up 5.5% since March 17, while Capstone Nikko is off 1.5%.

The August results have dismayed Capstone officials, who have been trying to drum up new investment in their fund among big and small investors. If Japanese stocks are poised for a sustained rebound, Capstone Nikko can’t afford its “dog” strategy trailing the market for long--especially given that the fund is by far the smallest of the five U.S.-based mutual funds investing exclusively in Japanese stocks.

Still, judged over the ‘86-’91 period, the strategy’s performance remains intriguing.

Why should it work so well? Ken French, a University of Chicago finance professor, says buying downtrodden stocks makes sense because “the notion is that people overreact--so the stocks that have plummeted have probably fallen too much, and are going to come back up.” It’s a simple idea, but eminently logical, French says.

As for why eight years of bad performance should signal the turning point for the typical Japanese stock--versus some other time period--there is only speculation.

David Rochman, a vice president at Nikko Capital Management, says one theory is that Japan’s stock cycles correlate with companies’ capital-spending cycles.

And though the fund’s August results disappointed, Rochman says the nature of Capstone Nikko’s strategy is to bet on a catch-up of the laggard stocks over time--not a run with the market’s hares in an explosive rally.

Investors interested in this fund (it carries a 4.75% sales charge) can call (800) 262-6631 for more information. But this certainly isn’t a bet for your lunch money. Remember: Past performance--especially hypothetical performance--is no guarantee of future results.

Buying Japan’s ‘Dogs’

A stock strategy that involved buying out-of-favor Japanese shares would have paid off handsomely between May, 1986, and July, 1991, according to the strategy’s developers. But since the Capstone Nikko mutual fund adopted the strategy on March 17, results have been mixed: The strategy worked well for a few months, but failed in August.

Investment return: Japanese ‘Dog’ Period market* strategy May, ‘86-April, ’87 +66.0% +44.6% May, ‘87-April, ’88 +5.6% +93.9% May, ‘88-April, ’89 +13.4% +56.0% May, ‘89-April, ’90 -11.4% +23.9% May, ‘90-July, ’91 -15.7% -10.2% May, ‘86-July, ’91 +48.5% +387.0% March 17-June 30, ’92 -8.4% -4.7% March 17-Sept. 3, ’92 +5.5% -1.5%

* based on performance of Topix index, which measures 1,200 stocks on Tokyo stock exchange

Source: Nikko Capital Management