If one basket reaches 40% of the portfolio, most probably some of the stocks will hit their price targets. Then we sell some of the stocks to reinvest the money in the lower-weighted basket. The same applies for the capitalization ranges. In a way, we are actively re-balancing the portfolio in the areas that have underperfomed the market, which works very well for us.

Q:  Can you give us some examples of each of these types of companies?

A: Growth industry leaders include companies like Gilead Sciences, which is a dominant HIV biotech company, or Chicago Mercantile Exchange Holdings. The consistent growth basket includes companies like DirectTV, which has 13-14 million subscribers, each writing a check of $50 per month. It is a very predictable business - people do not usually terminate their broadcast TV, they only occasionally switch from provider to provider. DirectTV is getting more subscribers than all other satellite and cable providers together. Another very predictable business is FedEx. They handle more than 2 million packages per day; it’s a very high volume business encompassing the globe. If some customers leave them for DHL or UPS, it wouldn't have a great impact.

In the emerging growth basket, there is a broad range of companies – from Imax, which does the big screen theaters to Sportingbet PLC, which is a London-based bookmaking and online gaming service.

Q:  How do you handle the stock-specific risk?

A: We model a GAAP and a pro forma income statement, which is reflective of how Wall Street arrives at their numbers. We have our own pro forma, which reflects our analysis and shows the earning power of the company. Often our analysis of the earnings quality, together with the analysis of the balance sheet and the cash flow, will raise certain flags in advance. For example, if are worried that a company is selling a lot of its products into the distribution channel but recognizes it as revenue to make a quarter, we know that accounts receivable will rise relative to sales.

We try to be proactive and sell stocks ahead of earnings disappointments. When we are right, we may sell early by a quarter or two. The growth stocks come with high expectations and when they miss the earnings expectations, they may go down by 20% or 30%. Our average position size is about 3%, and a 30-percent drop in one stock will decrease total return by 90 basis points, which we want to avoid. We spend a lot of time trying to eliminate the blow-up side of the equation and to figure out the upside at the same time, because the winners are those that generate benchmark and peer outperformance.

We are usually pretty confident on a stock selection, but the timing issue is a lot more difficult. We still think that Comcast is a great company to own, even though we've lost 3% in that investment over the two and half years. Our fundamental thesis continues to track what we expect. Eventually, we’ll be right on keeping it, but it has been a long time waiting.

Q:  Do you set price targets on the stocks you own?

A:We use price targets in two ways. First, companies with more finite opportunities may be a great investment, but may not be long-term winners. For example, a local community bank that is growing its earnings at 10% or 15% may be an attractive investment, but it is not going to become Citigroup. You may be able to capture a move with that bank from $10 to $22, but it is not going to $500. So at $22, we are happy to be a seller.

With other stocks, like Gilead Sciences, the dominant HIV franchise, we know that from time to time the stock is going to be expensive. But if we are confident on the fundamentals, we compensate for the high valuation by trimming back at price target and maintaining exposure to the position. If it trades down, we’ll take the opportunity to buy more stock. It is an example of how volatility can be a friend if you are confident on the fundamental work.

Cisco, which we don't own, returned 85,000% in the decade of the 90’s. The stock had an average P/E of 80, which I would consider pretty high. But if you sold Cisco the first time it got expensive, you would leave great return on the table. If we find the biggest winner of our time, we don’t want to sell it when it get expensive. We want to trim it and buy more when it comes down.

The reality of running a mutual fund is that you cannot make 85,000% on Cisco because if you capture that much return, by the end of the decade Cisco will represent 99% of your portfolio. People buy mutual funds because they want measured diversification and our position size is typically 2% to 4%.