November 26, 2001
Large-cap growth? Small-cap value? Where on the style grid can we put a fund? A Manager? We shouldn't put them anywhere because the traditional style grid used in the mutual fund industry to classify funds and managers is useless, according to a Colorado-based investment management firm.
"This grid is not working. I hope it's obsolete in five years," Craig Callahan, CIO of Meridian Investment Management, said to a roomful of financial advisers earlier this month at an asset management conference in San Antonio sponsored by Royal Alliance. Meridian manages the Icon funds, a series of 14 mutual funds with almost $700 million in assets under management.
One of the main problems that Callahan and Meridian have with the box-style grid, which has between six to nine boxes, is that the composition of each box in the grid is not consistent. Industries that are out of favor move down and left towards small-cap value, and industries in favor move up and right towards large-cap growth, Callahan said. As a result, using those boxes as a strategy assessment is counterproductive, he said.
"One way of using this grid is to hire six specialty managers," said Callahan. "That's a system that's not working. One problem is you cannot predict style specific performance. Any notion that you can do a search and hire a good mid-cap value manager is false."
For example, if pension plan trustees are conducting a mid-cap value search for a manager after 1998, they may find a manager with an impressive one-, three- and five-year track record. However, it might be possible that the manager was very good at analyzing banks and that sector comprised most of his holdings.
"What does he hold now?" said Callahan. "Banks have migrated out of mid-cap value. If he holds banks, he's accused of style drift. I think gold mining and energy are different skill sets."
That might be true for sector funds, but most funds encompass much more than single skill sets, said Russel Kinnel, a mutual fund analyst with Morningstar. It's true that industries move in and out of the style boxes, but they also move in and out of a manager's portfolio, Kinnel said.
"I do think that any simple configuration isn't going to capture any manager's style completely, but our categories do capture a fair amount of managers' styles," said Kinnel. "The style box can't possibly capture every style out there."
Callahan maintains that managers have been forced to alter their styles to adhere to the grid. Having managers that specialize in these classes causes void areas, he said. Managers are hesitant to buy stocks that are on the border of the box because if they move slightly, they'll be accused of style drift. Also, adherence to the grids causes artificial selling. Managers are often forced to sell a stock that is performing well, which not only changes the composition of the portfolio when the manager might not have done so, but also makes investors realize capital gains.
In Morningstar's case, the company bases its categories on three years' worth of data, and so one stock is not going to put a fund over the edge, said Kinnel. Both Morningstar and Lipper take into account the manager's stated strategy when classifying funds, and so a fund has to have a substantial position outside that strategy for the classification to change.
"The style grid is good for investors to use as building blocks for their portfolios, said Kinnel. "It certainly doesn't show the whole picture, but simplifies things in a useful way for investors."
Grid and Bear It
Callahan, however, offers a grid that he believes is more helpful to investors. That grid does not break the market up by size, but rather situates a manager depending on how important value and growth principles are in their investing strategies.
"We believe value and growth are two different scales," he said. "They are not the opposite ends of a continuum. Hot and cold, those are opposites. Value and growth are not opposites of each other."
A manager who says growth is very important but value is not, would place on the top right of the graph. A manager for whom value is everything would be placed on the bottom left. That scale allows managers to be placed in appropriate areas, and allows a classification for managers who lie towards the middle, said Callahan. Then investors can be surveyed and asked questions related to their investment objectives and preferences.
"Both managers and investors can be scored on the grid," said Callahan. "Then imagine matching the managers with a score that correlates where the investors are. I'm being sarcastic here. Imagine matching investments to the investor. The style grid today does the opposite. It forces everybody to fit the darn style grid."