TMF: As demonstrated through your past and current investments, you don't have any problem investing in high-growth companies. But a lot of other investors are reluctant to project such high future growth rates. How do you become comfortable with estimates for double-digit growth over several years?

Gay: For a company like Amazon (NASDAQ:AMZN), we don't believe it will grow at extremely high rates forever. I think the way that we get comfortable with owning names like Amazon or eBay (NASDAQ:EBAY), which are perceived to be growth names, is really to look at them on a more normalized basis over a long period of time, understanding that the growth rates are likely to degrade to more normalized levels.

We don't have to believe that Amazon will continue to grow at 25% for 10 years to buy it or to own it at $39 or $40 a share. All we have to believe is that its business model is sustainable; the returns on investment capital are going to be at least at this level or higher -- we think probably higher; that the company will continue to execute on its strategy; and really getting to know the managements and understanding who are the capital allocators and how their decisions are made leads us to that level of comfort.

We get comfortable [with growth estimates] by doing a lot of work, by really meeting with the companies, and understanding what their long-term strategy is to create value. That gives us the courage to invest in companies like Tyco (NYSE:TYC) or Qwest (NYSE:Q) or Amazon in the summer of '02 when everyone's throwing them out the window.

We spend a lot of time saying, "Okay, what's this business worth today? And if they continue to execute over the next one year, three years, five years, 10 years, what's it going to look like over those various time periods?" And [we] assign probabilities to those scenarios.

Bill [Miller] wrote a piece earlier this year entitled "Mulligan." What he did is he basically went through the analysis of eBay. Many people, on very traditional metrics, would view a company like eBay as being very expensive. What he compared it to was Microsoft (NASDAQ:MSFT).

TMF: Right. I remember seeing that.

Gay: Basically, the story there was Microsoft looked really expensive back in 1990. Many people believed it was overpriced. In fact, Bill wrote what he says now is probably one of the dumbest things he ever wrote, saying Microsoft was really expensive. In retrospect, it was really cheap. Many technology stocks in 1990 were really cheap, but no value managers owned them.

One of the things that we talk about is that there really are two types of value traps. One is a value trap where you buy a company that's a low P/E, but it deserves to be a low P/E because it's destroying value. The other type of value trap is when you avoid buying something that on an accounting metric standpoint appears to be expensive when in reality it's quite cheap, because fundamentally it's changing.

So whether it was an Amazon at $5 or $6 a share when there was a lot of concern about their business model and the acceptance of it. Or Dell (NASDAQ:DELL), which we bought between March and June of 1996 at about 6 or 7 times earnings, and people didn't believe that there was going to be an acceptance of their direct-delivery method of computers. But when you actually looked at the returns on invested capital, they were about 40%. And they went to triple-digit levels within three years. So it was quite cheap in retrospect. Or an AOL (NYSE:TWX), where people were focusing so much on one side of the business and not focusing on the potential of another side of the business. [Note: Bill Nygren made similar comments about AOL in his interview with the Fool.]

Actually, the same story could be told of Amazon as well. People focused on who can make a lot of money just selling books. Well, it's not just a book business. It's a fulfillment business. If it's just a book business, yeah, then it probably is not worth a lot more than where it is. But if you identified the value that we see as inherent in Amazon and that it is a fulfillment business -- starting out in the books, music, and video, but expanding out into other areas -- and it has, drawing on our Santa Fe connection, a lock-in or path dependence where once a user uses a system, whether it's an Amazon or AOL, they have a loyalty effect that causes them to go back to that because there is information there that is valuable to a consumer. They already have all the credit card information. I have a lot of nieces and nephews. They've got all my kids in the system, so at Christmastime, it makes it a whole lot easier to do Christmas shopping when you just go in and click who you want a gift sent to. Once you already have that, you have a relationship with a business, which is valuable, and it's difficult to really quantify.

TMF: What kind of returns are you looking for in the companies you invest in?

Gay: The minimum requirement for a discount before a new name enters a portfolio is typically 30% to 50%. So we have to believe with a fairly high degree of confidence that over the next three to five years we're going to at least earn that amount of return. We can be wrong and have been wrong. Hopefully, we can be right two-thirds of the time and still deliver good returns. But basically, that's our typical threshold before a new name would enter the portfolio. And then when the new stock enters, it's much more of a function of does it replace an existing name because it adds some level of better risk-adjusted returns than what we have in that name. Does it go into the portfolio simply because we've got cash that we want to put to work, and we think it's a good, new name that will add some level of valuation diversification that we don't have? So we want to keep the portfolio fairly focused in the names we have the highest confidence in.

We tend to be very patient, so many people might look at what we're doing and say, "Well, how crazy is it to invest in a security where there's a lot of headline risk and there's a lot of concern in the marketplace about the sustainability of their business or there's a lot of concern about their balance sheet." Oftentimes, it takes quite a while before the market gains confidence that the company truly has turned around.

If we buy something and it goes down, if the investment case is still in place, we typically are adding to it. One of the expressions Bill's fond of saying is, "Lowest average cost wins." [Note: In studies, Columbia University's Professor Bruce Greenwald found this strategy accounts for a lot of the performance of great investors.] If we buy something and it goes down and we still believe it's a good value, it's a much better value at a cheaper price. This may sound a little perverse, but the last thing we want a company to do when we initially start buying it is to go up right away. We need to build a position, and we tend to be liquidity providers. Valuation is always the key. We want to buy stocks that are trading at the lowest possible value that we think that they're likely to be at relative to what we think they're worth. If we begin to buy it and it goes up, we're reducing our future rate of return. We would tend to be somewhat cautious about adding to it.

TMF: You've explained when and why you buy. When do you sell?

Gay: Typically, we only sell a security when one of three things happens. When it reaches fair value -- when we don't believe we have the potential to continue to outperform the benchmark by owning that security over our investment time horizon. The second reason is when a better investment comes along: When one new name that we've come across in our research is better than something we currently own, we'll swap it. And then the third reason is when the investment case no longer applies. That occurs when we've made a mistake, when we misjudged the situation or when something externally changes. For instance, some new regulation or legislation that comes into place that would affect the economics of the business. For a company like UnitedHealth Group (NYSE:UNH), which is one of our bigger names in the managed-care space, if some new regulation that was passed that affected reimbursement rates for that business, it could fundamentally affect what we think the company is worth. So we pay a lot of attention to what's going on in the government front, because that can affect values as well.

Read Matt Logan's complete interview with Mary Chris Gay:

Are you on the hunt for value stocks? Join Philip Durell in his search for deeply undervalued companies in Motley Fool Inside Value. Each month his newsletter covers two companies that represent real value for investors. Start yourfree 30-day trialnow.

Fool contributor Matt Logan does not own shares in any of companies mentioned . The Motley Fool has a disclosure policy.