Q:  What is the investment philosophy behind managing the Old Mutual Long Short Analytic Fund?

A: Our philosophy is based on the belief that success in the investment markets requires a style that is disciplined and systematic, but at the same time, dynamic and capable to adapt to changing market environments. We also believe that to be successful, you need to pay a lot of attention to cost and risk controls.

In the current environment, where information is relatively inexpensive, the main challenge is what to do with it. In other words, the most important element of our process is how we weight each of the different data factors within the framework of the respective business cycle and economic environment. Since we have started running this process in 1996, we constantly test ways to come up with a better weighting scheme.

Everyone looks at performance, but going forward, we believe that our edge and core competence is the ability to continuously innovate and come up with ways to incorporate all the new data into a pricing model, and then incorporate that into the ranking process and, ultimately, into performance.

Q:  Why have you chosen a long/short strategy?

A: Based on our research and experience, we believe that structures that allow you to go short provide the client with more consistent value-added. In this fund, for every $100 invested by the client, we buy $120 worth of stock and short $20. As a result, there is more stock selection in the fund than in a traditional long-only fund.

With a long-only structure, if you don’t like a stock, all you can do is not hold or underweight it, which wouldn’t add any value for the client, especially for the stocks with tiny weights in the S&P 500. We estimated that by shorting the stocks that we don’t like, we achieve better results.

We started using that strategy in our institutional accounts and in our separate accounts about five ears ago with very good results. In January 2006 we also transformed this fund into long/short fund to create more alpha, while keeping the fees unchanged.

Q:  How do you translate that philosophy into an investment strategy and process?

A: The portfolio is invested 100% in U.S. stocks, and our goal is to outperform the large-cap U.S. indexes, regardless whether the S&P 500 or the Russell 1000 as both of them are cap-weighted benchmarks. Our universe consists of the largest 1,000 companies in the U.S, or the 1,000 stocks in the Russell 1000 Index.

Since we are a quantitative firm, the most important element of our process is generating the rankings. The process starts with collecting a variety of fundamental and technical data on each of the companies in our universe, weighting the data in terms of importance, and then, on a daily basis, ranking the stocks in our universe from 1 to 1,000.

We collect data on 70 indicators, including valuation numbers, technical information, risk variables, volatility, and profitability indicators. We also collect information on the insider buying or selling and analyst revisions. Then we evaluate the importance and the weight of each of these characteristics.

For example, prior to 2000, investors focused mostly on price momentum and growth characteristics. But when the tech bubble burst, everyone became focused on valuation. In 2002, when we lost confidence in analysts, the focus shifted towards dividend yields. And last year, because of the leveraged buyouts, investors began to focus on sales-to-price as a valuation characteristic.

So our process involves analyzing the data for the last month, the last year, and the last three years, to examine the relationships between each of the different characteristics and returns. Then we use the strength of that correlation to weight the variables.

In the last three years the strongest relationship between variable and return is the sales-to-price ratio. As a result, the companies with higher than average sales-to-price ratio receive a big positive score. Our biggest underweight now is the dividend yield. Investors are actually shying away from companies with dividend yield, which is typical behavior for the late stages of recovery when growth opportunities are scarce. The companies that pay high dividends usually lack growth opportunities within the company itself, so that characteristic gets a negative weight in our process.

Q:  Could you explain in more detail the process of classifying the 70 indicators that you collect data on?

A: To find which indicators are important in the current environment, we examine each of the 70 indicators and the relationship between the indicators and the recent stock returns. If an indicator is important, there will be a big positive correlation or a big negative correlation. If an indicator has no correlation, then it is not that important.

To adapt to market changes, we change the weightings based on the 36-month moving average of the correlations. If something has been in favor for 36 months, it will get a heavy weight in our ranking process. Although we update the weightings every month, we have found that the 36-month window is the best way to forecast what will work next month. The system is exponentially weighted, which means that the last month would get a higher weighting than the months before that.

Q:  How quickly does the strategy respond to market changes? For example, how does your system react to the already three months of a different environment in the subprime area?