Copart was found on one of our numerical screens for attractive margins and positive trends in fundamentals. The thing about Copart that stood out was its high revenue growth. At the time, it was growing around 20%, had high margins and good returns on capital.

Another thing that made Copart different is that it had an Internet portal at the time when the Internet was just gaining traction. By using the Internet to open its auctions to a wider audience, the values that were received at the car auctions increased, which helped the company capture market share.

In the research process, I visited three or four salvage yards to meet with the operators and the buyers and to understand how the auctions worked. I met with the management, both here in our offices, and in their San Francisco area location. We then built our own earnings model and determined how much capital was required to run the business.

In 2003 the stock was extremely weak because company management had gone out and expanded too quickly and accumulated a lot of real estate to do so. Their business slowed and Wall Street hated it. It was important for us at that time to understand management's philosophy. They could look out three to five years and see huge opportunities for Copart because their service was so much better. Management knew they were going to get more cars to auction but they simply didn't have the land to accommodate the inventory of cars, so they were investing in real estate at a time when it wasn't convenient for Wall Street. The stock was punished, but it was a good time to be buying the stock, which we did.

Q:  Why did you believe that Copart would continue to grow?

A: Copart's margins were better because it was using the Internet at a time when all its competitors were still holding oldfashioned auctions at their salvage yards with an auctioneer and bidders walking from car to car. By using the Internet, Copart was attracting more buyers and the prices at auction were higher because there were more competitive bids.

The company eventually stopped holding physical auctions completely and went to an all-Internet auction, which had a tremendous positive impact on margins because all the costs of holding physical auctions went away. Copart had better operating margins, faster growth and better returns on capital than the rest of its market. And the chief executive, Willis Johnson, owned about 20% of the stock. All those things gave us confidence that this was, and still is, a long-term growth company.

Another example is Pediatrix (PDX), a physician practice management company that serves neonatologists and emergency care pediatricians. First, the number of sick babies that are delivered and need emergency care every year is very consistent, so it's a very stable business. Pediatrix works with approximately 30% of the emergency care pediatricians and neonatologists in the United States, which is a market position that is very difficult to replicate. The company has been the largest physician practice management company ever since it controlled 10% or 15% of the market, and it has continued to grow. Pediatrix has a long history of making physician practices more profitable, which is in part a result of better contracting with insurance companies. Their size is a great competitive advantage. So, these are two examples of why we focus on the business model and the sustainability of the margins.

Q:  Generally, how do you build your positions?

A: Our minimum positions tend to be 1% weights and our top holdings tend to be 5% weights. The stocks at the top of the portfolio with the largest weights are the ones we have known the longest and the ones we believe to be the most sustainable, consistent growers.

The smaller weights are positions where we think we found a consistent grower but it's going to take more time getting to know the company better and watching the management team execute their business plan.

Q:  Could you describe your buy/sell discipline?

A: Our baseline valuation for a new position is a trailing P/E multiple that is equal to, or less than, our estimate of the long-term forward earnings growth rate. If a company is growing 20%, then ideally, we'd like to pay 20 times earnings.

When we get confidence in the sustainability and quality of a particular model, we'll increase our weighting in the stock. We are constantly comparing the quality of different businesses. If we think a company's prospects are becoming brighter or if a company has made some significant gains while another company seems a little weaker, then we may swap those positions.

Sometimes we'll have a 1% position in a quality company but think the valuation is too high, and our models tell us that we can't make our targeted 15% return going forward. We will likely sell that position and wait to buy the stock again when it is more attractively valued.

In all, there are really only two times we sell a stock. The first is when our original investment thesis is broken and the other is when we think the valuation is too high. If we don't see the potential for out-sized returns, then we sell the stock. We benchmark the portfolio against the Russell 2000 Index.

Q:  What is the total number of holdings that you generally have?

A: We run a fairly concentrated portfolio. We currently have 69 holdings and that number could be 10 lower or 10 higher at any given time as we buy or sell positions. Right now, our 20 largest positions make up about 60% of the portfolio.

Q:  What do you do to mitigate risk?

A: We take a simple approach to managing risk. We stay disciplined and focus on buying long-term growth businesses and paying a decent multiple.