We don’t rely too much on the top-down perspective but we do industry research and we check the demographic trends to see if certain companies would benefit from those trends.

Q: How do you generate investment ideas?

A: We get our ideas from numerous sources, including news publications, quantitative screens, investment conferences, company visits, and quarterly and annual reports. We also talk to many different money managers. We know most of them on a personal basis and we visit them frequently, or talk to them on the phone. We attend many of the same meetings together and we network with other professionals all the time. I think that being tight-lipped on investing ideas is detrimental to our shareholders since it keeps you in a vacuum.

Q: What is your buy and sell discipline? How much turnover does the fund generate?

A: The average holding stays in the fund for about four to six years. We rarely sell a mutual fund unless there’s a significant change that we’re not happy with. Typically, we do all our work on the managers upfront and we know exactly what’s happening and why.

On the securities side, turnover is a bit higher. We typically sell stocks when the price targets are met. We periodically re-evaluate our stock holdings and if the intrinsic value hasn’t grown, a stock becomes a sell candidate. But some of our core positions, which are great businesses, may be in the fund for 10 years. Other holdings may just be trading at cheap valuations and we purchase them to exploit the valuation anomaly.

An example of such as stock is PetroKazakhstan, which we bought in 2005 when the stock was down about 50% over three months. Subsequently, they became a takeover target and we made a nice return in just a couple of months. That’s an example of a short holding period, which wasn't a function of timing the stock, but was rather an example of exploiting a valuation anomaly. Typically, if a stock we think is fundamentally sound moves down, we’ll purchase more.

We like to purchase stocks at a discount of 30% to our estimate of intrinsic value, and that serves as our margin of safety. We like to see the free-cash-flow yield of 10% or more. We have companies that can consistently earn return on invested capital of above 15%. The ideal company should have a current ratio of above one.

Q: What is your portfolio construction process? How many securities do you hold?

A: Right now we are invested in about 25 mutual funds, 33 stocks and a limited partnership. Usually, we have a little over 50 holdings, but the top 10 holdings represent approximately 40% of the portfolio, so it’s relatively concentrated. On the mutual fund side, we typically invest in very concentrated mutual funds, but with a global scope. So the combination is unique.

Another unusual aspect is that we have the flexibility to invest in hedge funds. We have invested in a limited partnership, Armor Capital Partners L.P. Armor is a conservative, absolute value investor which focuses on undervalued securities from around the world. It represents about 3% of our mutual fund.

The weightings in the portfolio are based on our level of conviction, so we may overweight certain businesses or funds that offer the best risk/reward potential. That's one of the ways to control risk.

Q: What are the other ways to control risk?

A: The most important risk control is the uniqueness and the independence of our investment ideas and the quality of the due diligence on each company or manager prior to purchase. Buying at reasonable valuations also helps buffer against downside risk. Investing in under-leveraged or wellcapitalized companies is also important, because in an inevitable downturn, they’ll be able to take advantage of industry flux.

Another factor is investing in concentrated mutual funds. Most people would argue that a 25-stock portfolio is more risky than a 500-stock portfolio but I would disagree. Investing in 25 stocks is safer because the manager knows those companies more intimately. We like our managers to invest in their best ideas. Typically, concentrated managers have had more success over the long term than diversified funds, which often end up mimicking their index. Our approach is to combine a couple of dozen concentrated mutual funds to reduce any stockspecific risk.

Nevertheless, our allocation is spread among many different regions, sectors, companies and market capitalizations. We have a flexible strategy, which allows us to go anywhere to exploit the largest valuation gaps we can find.