Stock-Picking Takes Concentration

eBay's stock split 2 for 1 last summer, and its price has nearly doubled in the last 12 months, climbing from $44.12 to $83.15. UnitedHealth's stock has also split in the past year, while its price has risen from $44.40 to $65.71 over the last 12 months.

The Janus fund is less expensive, though, charging 0.83% a year against the average of 1.67% among the actively managed funds tracked by Lipper.

But not every manager backs winners in a small portfolio.

Tjornehoj points to the American Heritage (AHERX Quote) fund, which dwindled to $1.1 million after investing 23% of its capital in penny stock ADM Tronics Unlimited and another 21% in Senetek (SNTK Quote), a biotech company.

Neither turned out to be worth the big bet. Senetek dropped from a high of $4 a share in March of 2000 and is now trading around 70 cents a share. ADM Tronics briefly topped 60 cents a share back in early 2000, but dropped to single-digit cents by the beginning of last year before rising to 37 cents recently.

Those sustained losses drove the fund's annual expenses up to a whopping 10%, and the fund has only recently begun the long, hard climb to breaking even.

"The joke was: 'Why not just buy the company?'" Tjornehoj says, adding that a $10,000 investment in the fund in 1994 is now worth about $1,200.

Sarat Sethi, a partner and portfolio manager at Douglas C. Lane, says he is skeptical about hiring a fund manager "to beat the pants off the market," because those claims don't hold up over long periods of time.

But he says that for the average investor, exchange-traded funds offer similar benefits as concentrated portfolio funds, and at a lower price. "Every industry is very narrow, and ETFs balance that out."

Glen Craman, a financial adviser with Pearson Financial Group in Lake Oswego, Ore., says bad experiences in the 1980s prompted him to avoid concentrated portfolios and sector funds, because market-timing never fit into his investment strategies. "At one time, our office was publishing 18 newsletters about these things, and while they made for great academic reading, the practical implications were ineffective. I haven't found a system [for market-timing] that really works."