Q:  What is the investment philosophy of the fund? A: Our philosophy is both behavioral and quantitative, which is an unusual mix. Instead of competing with the army of analysts trying to predict the changes in a company’s financial health, we measure the specific factors that prompt investors to make one investment instead of another. We use various quantitative measures to try to understand the factors that are rewarded or punished in the current market environment, and then predict the reaction to changes in a company’s financial health as we watch them occur in real time. We look at a universe of about 3,400 stocks that we consider to be liquid enough and we measure them on a daily basis. The excess return, over time, comes from adjusting to the psychology that determines the investment environment. The quantitative behavioral approach performed well both in the growth environment of the late ’90s, for example, as well as in the value-driven bear market in the beginning of this decade. The mutual fund was launched at the end of 2002 but I've been managing portfolios with this strategy since the end of 1998. Q:  Why do you believe that this philosophy is a better way to manage money? A: First, I don’t believe that it's the only approach that works or that it’s always going to be superior to other, more traditional philosophies. But because it is not highly correlated with traditional strategies, the Top Flight fund can fit well into a portfolio of traditional strategies and improve the risk/ return profile through strategy diversification. Second, I strongly believe in having a competitive edge. It’s very expensive to have a competitive edge, especially in the large cap space, if you have to discover more and better company information than the other participants chasing the same excess return. So we’ve found a place and an approach that allows us to maintain our edge over time. There is also the aspect of the anatomy of a trend; investment managers usually say that they want to identify trends and follow them. But for this to be profitable, the trend has to be observable and durable over the period that you are exploiting the trend. In my experience, trends in investor preference are both observable and durable – much more so than trends in sector performance or individual stock performance. Q:  Why do you think certain styles are durable, especially the styles that you focus on? A: We focus on different styles at different times. One of the reasons for styles being durable over a certain time frame is the tendency of retail investors to go with the best performing managers. As mutual funds in a specific style begin to perform well, investors reward those funds with their investment dollars, so the stocks that fit the specific profile of the manager get money inflows. You can see that in terms of pure price momentum coming from investors wanting to buy specific kinds of stocks. There isn’t a tendency, especially among retail investors, to buy what nobody else is buying. Using a shorter investment horizon (much shorter than that of most mutual funds) enables us to exploit these trends. Over a four to eight week period, these trends are very durable and exploitable. Q:  How does that philosophy translate into a strategy? A: Some people have termed this strategy short-term style rotation, while others use style momentum. I don’t like 'style momentum' because people can misunderstand exactly what’s being stated. The idea is that a specific factor or kind of stock tends to perform well in an environment over a four to eight week timeframe. We focus on identifying the emerging trend and exploiting it by investing in the right kinds of stocks. To be more specific, we have about 60 factors or reasons why people would buy one stock instead of another. Those factors include technical measures, such as momentum and volume; fundamental measures, like valuation and growth metrics and combinations of the two; expectation components, such as analyst estimate revisions and earnings-toprices; and basic company data, such as sector and industry membership and stated repurchases. Then we rank our entire universe of liquid stocks; we take the top and bottom deciles to construct virtual portfolios that represent the degree of reward or disdain that investors show to specific factors. We then select five to seven factors that show leadership or emerging strength and we construct a multifactor linear model. The model is used to rank our universe, and we use that ranking for all of our investment decisions. Q:  Could you describe your research process in more detail? A: Our research process is very different from that of traditional managers as our investment decisions are not that much based on information about the company and its business model. Instead, we’re interested in the profile of specific stocks and in the investors’ attitude towards stocks that match that profile. Given the short holding period, this is a high turnover fund. At any given time, we'd be buying or selling probably 10 to 20 stocks a day. Q:  So you use shorting in the fund? A: Yes. Currently we’re 53% long and 47% short, but we let the market move us in terms of that ratio; it’s not a top-down decision. For example, as the market dips, we will enter more of the long positions that we're interested in taking, and exit more short positions we’re interested in exiting. So the portfolio will naturally move into a net long position during market declines and into a fully hedged position during market rallies. There are times, especially on the long side, when we see a fast pitch opportunity based on the macro environment and we may move the portfolio to being as much as 100% long. Q:  In your experience, what factors have been consistently helpful when selecting stocks or metrics? A: I’m not sure that there are factors that are relevant at all times. For example, right now stated repurchases is a very relevant factor but it tends to be more relevant at the first calendar quarter, peaking in January. Although it is still part of our model, we’re seeing its impact declining. Earnings surprises tend to be a more relevant factor during the beginning of the earning season and then trail off towards the end, when revisions become a stronger factor. It’s very interesting that the model has at times pushed us to construct a portfolio that would seem counterintuitive. The best example is 2003, when at one point the model was putting us into stocks with low prices, high leverage, high valuations, negative earnings, and negative earning surprises. That list of characteristics is not something that many investors would want for the long term. Yet we positioned the portfolio in that way and were able to participate in one of the most spectacular short covering rallies in recent history. Capital was leaving the short positions and pushing those share prices up, the model was able to point that out, and we capitalized on the trend. I can’t think of any traditional strategy that could have been more successful in capitalizing on the psychology of that particular market environment. Q:  Investors usually engage either in fundamental or in value shorting. How do you approach short positions? A: Our approach to shorting goes back to our philosophy: they’re behavioral shorts. We’re shorting stocks that investors don’t want to own right now or that they would trim. So we want to understand the factors that investors are punishing and where they're taking away their investment dollars. For the short positions, our holding period tends to be shorter than for the long ones, just because of the way stocks move. Successful long positions tend to drive up over long periods of time, while successful short positions tend to see fairly dramatic price adjustments. Q:  The housing sector has divided some of the fund managers both on the long and the short side. Do you view housing stocks as an opportunity? A: Housing is a great sector to talk about because we have been long in housing stocks for the majority of the life of the fund. They were a big part of our long holdings in 2003 and through most of 2004 largely because they were attractive across so many different stylistic factors. They were attractive from growth, valuation, and momentum perspective. In a number of different environments they continued to score well, but starting towards the second half of last year, our housing positions have been primarily short. That’s an example of how a stock can be the right stock to own in some environments and then can change to the stock to sell. Often what has really changed is not the particular fundamentals, but the attitude of investors toward those fundamentals. Q:  What's your approach to portfolio management? How many positions do you have and what are the reasons for rotating them? A: It’s very intuitive and straightforward. We select the position sizes based on liquidity. After we take a position, we don’t trim the holdings until we’re ready to exit. The positions that do well can drift up to about 2.2% of the portfolio, but rarely, if ever, get bigger than that. In the smallest and least liquid stocks, we’ll take a 0.40% position. On the short side, our largest positions are smaller, about 1.25%, because there is more vulnerability to news-based events in these stocks and we tend to spread our short exposure across more holdings. The total number of holdings fluctuates between 75 and 150 positions, depending on the number of large-cap versus small-cap positions. If we have more positions in smaller stocks, our number of holdings tends to increase. Entering or exiting positions is all driven by our model that is run daily to rank the universe. When we see stocks falling out of the top quartile, we begin to look for exit opportunities. Q:  Is market cap a consideration in the selection process? A: Absolutely. One of the big factors in 2006 has been that the smallestcap stocks have performed better and there's a real preference for small over large caps so far. I wonder if that will persist, but I’m confident that if it begins to erode, the model will give us fair warning to reposition the portfolio before we start seeing significant large-cap outperformance. Our periods of greatest excess return versus the market have come generally on the back of smaller-cap positions, either on the long or the short side. When the portfolio moves more towards holding long positions in large-cap stocks, you’re basically moving towards the mean of the market indices. In such periods we’re more correlated with the market and the traditional managers, while at other times we diverge significantly. Q:  What risks do you perceive and how do you monitor them? A: In terms of individual position risk, it is my experience that stop loss orders or other protective measures can often backfire. We don’t have specific loss-threshold exits on individual positions partially because, in the case of underperformance, our model will take us out pretty quickly. That's the way the model is structured. Being a highturnover fund with short holding periods, we cannot be slammed by a single position underperforming for a long time. It is a different issue whether a certain market environment would make this philosophy irrelevant. One of the benefits of this philosophy and strategy is that it should perform well in any environment. The risk is that during periods of low volatility, when there is little difference between the top and bottom performing styles, the potential for outperformance (by being in the right style) disappears and the fund’s higher trading costs become a bigger deal. There also is the potential for a style to quickly become out of favor and then in favor again. In those environments, the high turnover approach can produce underperformance versus the market indices, but I've never seen that happen for two quarters in a row.