LOS ANGELES. Think of Roy Weitz as the Anti-Cramer. The television Jim Cramer is intense, bombastic, and in-your-face. He presents investing in the time-tested manner of World Wrestling Entertainment.

Roy Weitz, founder of www.fundalarm.com, is the polar opposite--- modest, measured, and unassuming. His website, established in July 1996, is non-commercial and independent. Basically, Mr. Weitz has a priceless public hobby, a website you can visit without pop-up ads, fancy graphics, or breathless announcements of Inevitable Wealth or Impending Poverty.

The basic content of his site is based on a simple idea--- that we need to know when to sell a mutual fund as well as when to buy it. To provide guidance, Mr. Weitz puts funds in four categories: No-alarm funds, one-alarm funds, two-alarm funds, and three-alarm funds. A fund sets off alarms when it underperforms its target benchmark for one, three, or five year periods.

A patient man, he says you should actively consider selling a fund if it registers the full three alarms. Even then, he suggests further examination. If the fund is an index fund, for example, he points out that it will inevitably trail its index by a small amount because expenses will reduce the return below the benchmark. So an index fund can register three alarms and not be a sell candidate.

The further a fund is off its benchmark, he suggests, the more inclined you should be to sell it. To that end, he also provides a measure of risk so you know how much more, or less, than its benchmark this fund could lose in the future.

I asked if he would identify any limitations his measures had.

"One limitation is that the alarms are skewed to whatever (investing) style is in favor. If value investing is doing well, there will be a lot of growth funds with three alarms. If value is in favor and you have a 3-alarm value fund, that's something you should pay attention to. There are a surprising number of these," said Weitz.

"The alarm is a good reason to look further. It's the beginning of research, not the end."

I asked for examples of how his measures could best be used.

"Some of the best users see Fund Alarm as a first or last opinion--- an overview of a fund. If they come up with a positive impression and then discover it's a 3-alarm, they know it's time to do more research.

"Another important thing is peer group ranking. If you have a 3-alarm fund that's better than its peer group, you'll probably be better off ignoring the 3-alarms."

I asked if he saw any trends in different fund firms as a whole.

"Yes, the American Funds group is beginning to fall off the no-alarm list. That may be a turn in the market. Value may be going out of style. It may be an inflection point. Remember, they had a lot of 3-alarm funds back in 1999 when value was out of style.

"Also, Putnam, Morgan Stanley, Merrill and other brokerage oriented funds tend to have a lot of 3-alarms. Janus, on the other hand, seems to be pulling out (of its 3-alarm period). It's improving."

I asked what he thought of Fidelity Magellan, since it would now be ranked as a 5-alarm fund, trailing the S&P 500 Index for 1, 3, 5, 10, and 15 years, if there that many alarms.

"It's still enormous. That means its 'bets' are larger than a lot of whole fund families," he pointed out.   To put this in some perspective, consider this recent data: the largest single stock holding in Magellan is worth nearly $2.3 billion. That's more in a single stock than the total value of all but 418 of the 17,400 mutual funds in the Morningstar database.

His expectation? More and more Magellan assets will be siphoned off into other Fidelity funds.

I asked what he thought of the mutual fund industry in light of the recent scandals.

"It's very reasonable to hold these people to a high standard because they have good jobs. They're paid well.

It's interesting work. They have public esteem. It's prestigious work with some financial security."

On the web:

Fund Alarm