Over the years, I've interviewed enough portfolio managers to fill a couple of Boeing 747s, and it's gotten to the point where there's very little that any of them can say to surprise me. Then I met Ted Aronson. He runs Aronson + Partners, a firm that manages more than $2 billion for big investors like Ameritech, the Florida Retirement System, the MacArthur Foundation and New York University. You normally need a minimum of $25 million to invest with Aronson, but he and his partners, Kevin Johnson and Martha Ortiz, also manage a mutual fund, Quaker Small-Cap Value, which was launched two years ago and holds just $11 million in assets (minimum investment: $10,000).

Aronson, 47, has been managing money since 1974. He works out of a converted duplex apartment in an old building in downtown Philadelphia. Unlike most money managers' offices, it's not decorated with giant abstract paintings and stiff mahogany furniture. Instead, there are plenty of soft places to sit, and dozens of bulls and bears made out of glass, ceramic, wood, papier mache' or plush. Everywhere you turn -- even lining the walls of the office bathroom -- are antique stock and bond certificates to remind you that many of the most popular investments of the past turned out to be terrible failures. Aronson's biting humor even extends to the office postage meter, which stamps outgoing mail with the slogan, "Don't confuse brilliance with a bull market."

Ted Aronson can barely open his mouth without challenging conventional wisdom. In fact, he's so blunt and provocative that even if you have no interest in his mutual fund you can become a wiser investor simply by hearing what he has to say.

Q. You've said that investing in an actively managed fund (as opposed to a passively run index fund) is an act of faith. What do you mean?
A. Under normal circumstances, it takes between 20 and 800 years [of monitoring performance] to statistically prove that a money manager is skillful, not lucky.

Q. Hold it. Did you say 800 years?
A. That's right. To be 95 percent certain that a manager is not just lucky, it can easily take nearly a millennium-which is a lot more than most people have in mind when they say "long-term." Even to be only 75 percent sure he's skillful, you'd generally have to track a manager's performance for between 16 and 115 years. Now let's say you finally get to the point, after 20 years or something, where you can say to us, "OK, Ted, Kevin, Martha, you're it, you're great, you really do have skill." Then the whole thing changes. I retire, Kevin loses his mind, Martha becomes an alcoholic, we sell our firm to United Asset Management and take on $10 billion when we can only manage $2 billion.

Q. You're kidding, I hope.
A. I'm kidding about our firm, but retail investors need to know how the money management business really works. It's a stacked deck. The game is unfair.

Q. You make it sound as if picking a good fund manager is completely hopeless.
A. Not completely. Start by looking for a very low fee structure. Then ask whether there's a finite amount of assets that the manager can run in his style and whether he's committed to limiting the fund to that size. Also, if the firm has an edge, ask why it should endure. If their money-making machine is inside one person's head, your risk is that person's lifespan -- or the chance that he'll get recruited away to another company.

Q. Over the past ten years, the U.S. stock market has returned 19.2 percent annually. What do you foresee for future returns?
A. They will be lackluster, and you don't have to posit a crash. The numbers just don't add up. Even if you buy all the "new paradigm" crap about how technology investment will lead to unheard-of levels of productivity, stock prices are so high that you can't get anywhere near decent returns over the next 10 or 20 years without taking valuations to levels that would be double or triple the all-time highs. Let's say real economic growth gets up to 3.5 percent, and you can get real earnings growth of another 1.5 percent. Add to that 1 percent to 1.5 percent for dividends and you come up with a long-term-bond-like return-not much higher than 5 percent or 6 percent. And that presupposes that valuations stay at least as high as they are now. People are compounding recent rates of return, at which money doubles every couple of years, into the indefinite future-and it just doesn't compute.

Q. But haven't stocks returned somewhere between 9 percent and 11 percent annually over the long run?
A. I love equities, I'm not a weirdo and I don't live in a bomb shelter. But in a very real sense, the compounding of stock returns over long periods is a fraud. It really is. No one has ever gotten those returns. Prior to the birth of index funds 25 years ago, the man on the street had no reliable way to earn the stock market's average return. I actually know someone who bought Microsoft at the IPO, although he sold most of it on the way up. There are people who have hit the mother lode like that-but not many. The rest of us mere mortals are lucky if we can hang on to the same mutual fund for 20 years. I recently went to my best friend's father's 80th birthday party and heard him musing about how much stock $10,000 would have bought when he was born. I said, "Equities have compounded by 11.5 percent since then. No wonder your son is smiling-because that $10,000 is now worth $60.5 million and you're getting old!" It's ridiculous. He's worth nowhere near that much.

Q. Prof. Jeremy Siegel of the Wharton business school says stocks have returned an average of 8.5 percent annually ever since 1802.
A. In 1802, the total value of all the stocks in Jeremy Siegel's market index was around $3.5 million. If you compound that at 8.5 percent annually for 196 years, you get $30.8 trillion. Yet the total value of the U.S. stock market today is only $11 trillion. So where did the other $20 trillion go?

Q. You tell me.
A. It went up in smoke. People bought good stocks that later went bankrupt, like the Philadelphia Guano Co. that's helping to paper the wall in our bathroom here. They bought bad stocks that never were included in any index. They took money out of stocks when they needed cash. They spent their dividends instead of reinvesting them. They sold everything during market panics, and they didn't get back into the market until after it had gone back up. They didn't buy and hold, they bought and sold-so commissions and bad stock picks and taxes (in the 20th century) ate away at their wealth. That's why this religious belief that stocks return 9 percent or 10 percent or 11 percent over the long run is just. . .guano.

Q. Prof. Siegel says much of the difference comes from the lack of dividend reinvestment.
A. He's absolutely correct. And that's why it makes no sense to assume anyone will earn that kind of return in the future. You have to subtract from the historical averages any portion of the return-fees, taxes, dividends-that isn't reinvested.

Q. So stocks are a bad idea?
A. Absolutely not. I think over the very long term -- ten years or more -- stocks will outperform bonds, and bonds will outperform cash, and all will be right in the universe. But the next five or ten years will be very different from the past five or ten; I could see flat returns for up to ten years. What I'm saying is that it's a bad idea to assume future stock returns will resemble those of the past.

Q. What's your investment philosophy?
A. There's a cartoon from The New Yorker that shows two lions on a hilltop, and one says, "I'm not picky -- whatever's wounded." Over time, I think the typical stock that's wounded -- that's out of favor in the market -- will outperform.

Q. Why don't more funds beat the market?
A. Because they can't.

Q. Why not?
A. Costs. Funds charge annual expenses of 1 percent or so; then it costs them another 1.5 percent to 2 percent to buy and sell their stocks each year. And there's something else: After three or four years of this water torture [lagging the S&P 500], active managers haven't gotten smarter. If anything, I think they've gotten stupider. They're scared. They're like cowering dogs. It's very hard to imagine them doing anything but just following the crowd -- because if they don't mimic the index, they'll get beaten by it again. A lot of people think active managers will outperform again when small stocks do better than the S&P 500, but there's nothing written in the Scripture that says it can't go on like this for our natural lifetimes.

Q. Doesn't history prove that small stocks outperform large stocks over time?
A. That's a crock. Small-caps don't outperform over time. When all is said and done, the returns [of small and large stocks] are very similar. Sure, the long-run numbers show small stocks returning roughly 1.2 percentage points more than large stocks. But those historical averages are a crock. The costs of buying and selling small stocks are a lot higher, and the averages don't reflect that. The extra trading costs easily eat up the entire extra return -- and then some! So over the long term, small stocks may even underperform. I'd like to think, on balance, they're a breakeven. But with a "whatever's wounded" investment philosophy, you should look at small stocks today precisely because they're so out of favor.

Q. When might they come back in favor?
A. Compared with large stocks, small stocks haven't been this cheap since 1973. That doesn't mean they can't get even cheaper -- but in 1998, the S&P 500 stocks outperformed the Russell 2000 [small stocks] by roughly 30 percentage points. Are the S&P's fundamental business prospects really that much better? Not by a mile -- not by a mile. These things really do go in cycles, and at some point small will be beautiful again. I wish I knew exactly when.

Q. What's a quant?
A. Someone (like us) who relies only on information that can be reduced to numbers. We think we can find an edge over a lot of other smart investors by crunching a lot of numerical information.

Q. Do you think your fund, Quaker Small-Cap Value, can beat the market?
A. Yes. What quants do is to identify what has worked and what is working. We can really drill down to the sources and causes of return. It works.

Q. Among your fund's largest holdings are companies like Allegiance, Keane and Mariner Post-Acute Network. What can you tell us about their businesses?
A. Nothing whatsoever. I don't know what their products are. In fact, I don't even know the names of their CEOs.

Q. Excuse me?
A. A lot of money managers make a big deal out of kicking the tires and meeting company management, but we've got a few problems with that. Company executives don't always tell the truth. More often, they themselves don't know what the future holds. Finally, when you talk to them, they're good salesmen. They can make you believe their PR even when they don't. We see no point in gathering information from unreliable sources.

Q. So you know nothing about these stocks?
A. I didn't say that. I said I don't know anything about their products. Our numerical analysis tells us oodles of things about the value of a stock relative to its peers, as well as what top management is doing with its own money (through insider trades) and the company's money (through share buybacks, etc.). We look at more than a dozen factors that measure value, management and Wall Street's sentiment. We want only the cheapest stocks by these measures -- whatever's wounded.

Q. What do you charge to run the fund?
A. Our normal management fee is 0.9 percent. But we can only earn that rate in a year when we beat the Russell 2000 [small-stock index] by three percentage points. If we return more, we earn more [up to 1.5 percent]; if we return less, we earn less [down to 0.3 percent]. It's a direct way to align our interests with the shareholders'. Obviously, if more assets hurt our performance, that will lower our fee. So we won't let the fund get too big.

Q. But according to Lipper, Inc. only 106 out of 5,190 U.S. stock funds charge performance-incentive fees. If they're so good, why are they so rare?
A. How many of those funds outperformed their benchmarks?

Q. Not very many, I'm afraid.
A. It's obvious. If those fund managers got paid better only when their performance was good, that would decrease their income.

Q. Do you invest in your fund?
A. All of my family's retirement money is in the fund. But because the fund trades a lot, it's not suitable for taxable investments. So all of our taxable money is elsewhere.

Q. Where?
A. In Vanguard index funds. I've owned Vanguard Index 500 for 23 years.

Q. You're an active fund manager who has nearly all his money in index funds?
A. Once you throw in taxes, it just skewers the argument for active management. Personally, I think indexing wins hands-down. After tax, active management just can't win.

Q. Your fund did very well against the Russell 2000 index in 1998. Why?
A. I wish I could say it was because we are geniuses. The truth is almost embarrassing. We got inflows of cash [from new investors] in the extreme down-markets of April, July and August [1998]. When the market is down, any cash sitting in the fund from the night before makes you look like a star. If no cash had come into the fund, we would have lost almost 5 percent for the year [vs. an 8 percent loss for the Russell 2000 index]. Thanks to the cash alone, our performance for the year was flat-five percentage points better than it would have been if no money had come in.

Q. You're saying that you performed so well not because of what you did but because of what your customers did?
A. Customers can distort funds' performance records in major, major ways. When you look at a fund's returns you're not just looking at what the manager has done in isolation. You're also looking at a mirror -- maybe even a circus mirror. Depending on when it comes in and out, the public's money can make a fund look dramatically better (or worse) than it really is. That can really gum up the works, and it tilts the whole goddamn game.