Q:  What core beliefs guide your investment philosophy? A : Our belief is that emerging markets are among the most inefficient markets and can be best exploited by a quantitative approach. Our investment approach combines value- and growth-oriented factors, as well as macro-economic and company specific factors. We believe this combination of factors can direct us to investment opportunities that exist in emerging market equities. We also believe that buying attractively valued companies with superior earnings prospects with positive market sentiment will produce consistent returns over all phases of the investment cycle. Lastly, we employ a quantitative investment process using models and optimization to create a portfolio that we expect will outperform its benchmark over an investment cycle. We believe that using a quantitative process is the best way to efficiently exploit the multiple investment opportunities while avoiding the undue sources of risk that exist in the emerging markets. Q:  How does your investment philosophy translate into an investment strategy? A : The fund normally invests at least 80% of total assets in equity securities of companies in emerging markets and other investments that are tied economically to emerging markets. The investment process begins with a liquidity screen for companies which have a daily trading volume of at least $5 million in the 25 emerging markets countries in the MSCI EM index. This screen helps the team invest in companies with adequate liquidity and reduces the universe to approximately 800 stocks. Next, we create a return forecast for every stock in the universe based on our proprietary multi-factor quantitative alpha model. The model seeks to exploit both bottom-up and top-down sources of alpha. On the bottom-up side, we broadly classify our factors into three groups: value factors, earnings growth factors, and technical/price momentum factors. On the top-down side, we broadly classify our factors into three groups as well - macroeconomic factors, country aggregate factors, and sector aggregate factors. Essentially, the model is two-third bottom-up and one-third top down. Once we have created the return forecasts, we use an optimization tool in order to determine the portfolio weights that will maximize the expected alpha of the portfolio subject to our risk tolerance, as well as the diversification constraints that we employ. The level of risk we are targeting in the portfolio is 5% tracking error with respect to the benchmark. We forecast risk with the help of a proprietary cross-sectional risk model which includes country, sector, currency, and style risk factors. In the optimization, we also constrain the portfolio to take no more than a 2% active exposure versus the benchmark at the country, sector, currency, and stock levels. The optimization process uses a transaction cost model to ensure that all trade decisions are influenced not only by risk and return forecasts, but also the cost of entering or exiting a position. The model takes into account implicit transaction costs such as market impact as well as explicit transaction costs such as commissions and stamp tax. Our strategy is rebalanced when the available return opportunities are sufficient to justify transaction costs. In practice, the strategy rebalances about once a month with annual turnover of 85%, but we do execute trades in between rebalances in response to stock-specific events like mergers and acquisitions, for instance, where our quantitative model is not as relevant. Q:  What do you consider emerging markets? A : We adhere to the definition of our benchmark, the MSCI Emerging Markets Index. The eight countries that are leading the index are Brazil, Russia, India, China, South Korea, Taiwan, South Africa and Mexico, and the combined weight of these eight countries is over 80% in the benchmark. Q:  What are some of the analytical steps involved in your research process? A : We use a quantitative approach combining bottom-up, top-down, value and growth factors in our research process. In the top-down portion of the model, we look at relative attractiveness of countries and sectors. In addition, we look at macroeconomic indicators, which help us identify the countries likely to outperform. In the bottom-up model, we look at valuation ratios, and we normalize by country and sector. We also look at earnings growth and technical/price momentum factors, which are not normalized. An integral part of our process is that we do simultaneous picking of countries, sectors, and companies. For example, we might find a certain country that does not look attractive but within that country we may find a very attractive stock. We may own it in the portfolio because its bottom-up characteristics are so attractive that it overrides the top-down negative or neutral characteristics. Typically, the countries that we prefer have increasing GDP growth expectations, strong current account surpluses, low nominal and real interest rates, and increasing industrial production numbers indicating a strong industrial base. These countries will do especially well in a world where global growth is slowing. Thus, we might find that there is a company in a certain country where the country looks good, but the company is not so attractive. But because the top-down characteristics are sufficientlypositive, it more than offsets the negative bottom-up characteristics. We aim to keep the best countries in the portfolio but we also want the cheapest stocks with the best earnings growth prospects in those countries. Therefore, we focus on companies that produce lot of cash flow and earnings. We also look atsales, book value and dividends. We mainly appraise a company relative to its peers. Our belief is that there is more alpha or excess return to the market in the bottom-up side than the top-down side so we can add more alpha by stock picking within the top countries. Once we identify the sources of alpha, we narrow down our universe defined by liquidity. From a quantitative viewpoint, we are interested more in breadth of coverage rather than depth of coverage. We believe that by taking a disciplined approach we can narrow stock selection down to the valuation and earnings growth criteria. Then, as we apply those criteria across a large number of stocks, we can offset the lack of depth. Additionally, we use a backtesting engine to select valuation factors such as price-to-earnings ratios, price-to-book ratios, price-to-cash flow, enterprise value-to-EBITDA and, furthermore, earnings growth factors like estimate revisions, as well as upgrades and downgrades. Once we have the factors that individually backtested well over the last fifteen years, we combine them using an optimization technique to come up with the best weighting to each of those factors - one that maximizes historic information ratio or risk adjusted return. Then, the final step of the research process is to put these into a backtesting platform where it actually creates the portfolios historically based on the alpha, risk model and various constraints. As soon as we verify that these factors add value historically, given real life parameters and assumptions, we can put the model into production. The ongoing research effort is to verify that the factors are behaving the way we expect them to behave and also to think about new factors that we can add to the model to improve its efficacy. Q:  Do you strictly follow the models? A : No, we do not always follow the model. At times, we apply certain subjective assessments and our objective in doing that process is really identifying what the model is not capturing. For instance, it is not easy to capture political risk in a quantitative model. To give an example, two years ago in Russia, President Vladimir Putin identified one of the companies in our portfolio as a company that he felt cheated on taxes. After that happened, the company’s stock price fell, but subsequent to that he retracted his statement, as it was motivated by politics rather than facts. The incident did serve to remind us of the political risk present in Russia so we responded by decreasing our active exposure to the country. Political risk is a serious issue with some of the emerging market countries. Q:  Do you constantly improve your models? A : Yes, and that is certainly something that we are keenly trying to do. One of the things that we have recently done is create a value distortion timing aspect to the model. That demonstrates how we are explicitly measuring the difference between cheap stocks and expensive stocks. For instance, we saw in the beginning of 2009, after the sell-off of 2008 in global markets, that the cheap stocks in the universe became even cheaper while expensive stocks became very expensive. The price-to-earnings spread between cheap and expensive became very large at the end of 2008 and that was a perfect time for value investing, because value investing is all about buying cheap stocks and selling expensive stocks. So, when the dispersion narrows we make money and that is exactly what we experienced in 2009. We want to take advantage of that type of phenomena to help us in timing factors. Thus, when the disparity between cheap and expensive increases, we increase the weight to valuation in our model. Another factor to consider is that we are currently implementing a quality timing model, which looks at the difference in valuation for high quality stocks minus low quality stocks. When quality is cheap we steer the portfolio towards higher quality stocks. Conversely, in the beginning of 2009 we found that quality was very expensive and so we steered the portfolio towards lower quality stocks. Q:  Do you hold cash in your portfolio? A : We do not have large cash balances in the account. When we see significant opportunity we will concentrate the portfolio more, but when we there are no opportunities we will maintain it less concentrated. We simply do not believe we can add value to market timing so we keep cash at low balances. Q:  How do you construct your portfolio? A : Our portfolio consists of anywhere from 70 to 120 stocks, and the MSCI Emerging Markets Index serves as our benchmark. The typical characteristic of our portfolio is that it will be cheaper than the benchmark on a number of valuation ratios but it will also have greater earnings growth expectations. We use an actively managed, tracking-error oriented, quantitative investment approach to select investments for the portfolio. The fund will generally invest in companies with market capitalization of $500 million or greater at the time of investment. The fund is constrained to take no more than a 2% active exposure to countries and sectors relative to the benchmark at the time of rebalance. The basic idea behind portfolio construction from a quantitative perspective is assessing how much alpha and risk is added to the portfolio by buying a particular stock. Portfolio construction weights are proportional to our risk adjusted alphas. Q:  Do you plan for unforeseen events? A : Yes, and one of the reasons why we construct our process with a focus on risk management is precisely because emerging markets are a very uncertain asset class with various sources of risk. We constrain the portfolio to have no more than 2% exposure at the country level relative to the benchmark. And conversely, if we did very much dislike a country, the minimum we can own in that is benchmark minus 2%. Since it is difficult to forecast political risk, we do not think it is possible to add alpha in that dimension consistently. Our approach to unforeseen events is to look at it from a risk perspective and identify sources for alpha generation in a way where we hedge out a lot of these sources of risk that are present in emerging markets. Q:  What are the risks that you focus on and how do you contain them? A : The biggest source of risk is the country exposure so we tend to limit exposure at the country level, as well as regarding currency, to no more than plus or minus 2%. Furthermore, we monitor the beta of the portfolio, earnings uncertainty, and sector exposures. And finally, we look at volatility, cyclicality, financial leverage, size exposure, as well as our exposure to China from a regression perspective.