Six years ago Leuthold Group stopped using Wall Street research for individual stock selection for its portfolios. How has it fared? Great.
Before the change, a Leuthold analyst would review Wall Street research regarding stocks. Then the firm's investment committee would vote on the analyst's recommendations, "which were based primarily on Street research and opinion," noted Steve Leuthold, chairman of the Minneapolis-based firm, in a report.
"The result from this more or less conventional approach was at best disappointing," he wrote. "For a number of years prior to 1995, we had tracked our individual stock selection within (its peer) group or sector, against the performance of the entire group." The firm's stocks typically underperformed the group average over different time horizons, from three months to one year. "We tinkered with the process, but nothing seemed to help," he noted.
"Finally, in 1995, I said `the hell with it,"' he added. "We quit paying for individual Wall Street stock research and disbanded the committee. Stock selection became a quantitative function, employing a multifactor ranking system."
The firm's goal had been to produce value added of 200 to 300 basis points above the portfolio's group-only performance. As it turned out, "the results have been far better than that," ranging from more than 500 to 1,300 basis points annually.
Mr. Leuthold acknowledged the firm "never did have the commission clout to be on a broker's A list, or `first call,' list. ... Our research input was exclusively from written reports."
Still, "the message is clear," he added. "The market tends to overvalue Wall Street's research favorites and often undervalues stocks in the same group." He views this tendency as "a market inefficiency created by excessive institutional enthusiasm focused on the stocks `the analysts like the best.' With a quantitative selection approach, the computer doesn't know what `everybody' likes the best."