Q:  What’s your definition of long-term performance?

A: Our goal is to outperform through the market cycle as opposed to generating a specific return over the near term. Upgrading is a long-term strategy that involves looking at short-term performance. Our experience shows that by monitoring our rankings and moving incrementally as leadership changes, we ultimately align ourselves with the leadership trends as they develop.

Overall, we try to capture 70% to 80% of a trend that lasts for months or even years. Because our system is less sensitive to shortterm movements, implementation is easier. We focus on making sure that we have adequate diversity among the funds to select from and an approach that gets us into things reasonably quickly but doesn’t constantly fluctuate. Fortunately, over the last 36 years that strategy has worked very well for us.

Q:  How do you handle the research process?

A: We have a five-person investment committee and staff that is focused on research only. Since we take a quantitative look at the market, we hold the funds we buy accountable for their performance versus a peer group that we select. A great part of our research is geared towards the risk classes and involves looking at historical data and performance.

We use data from a variety of sources and we meet with fund managers to understand their funds and to make sure that we place them in the right category. Most of the research is done through our proprietary software but we also use data from external sources to maintain our back tests and to monitor the actual performance versus our expectations. Then we monitor the performance of the funds to make sure that they’re in line with our risk expectations.

We look for new funds and we evaluate the implications of the new funds to the various risk categories. Even if a fund is never selected, its presence in our rankings will affect the other funds. Adding a fund of a certain type can change our portfolio weighting more than our investment in it. We also continuously examine our risk classifications to make sure that the metrics we use are the best for the purpose. Overall, the monitoring is a constant process of matching historical experience and expectations with the actual practice.

Also, we always look at the number of funds that we need to own. When we evaluate the sell threshold, we have to factor in minimum holding times to avoid trading for the sake of trading. Having more activity not only generates potentially higher trading costs, but also increases the chance to cause problems for the underlying funds. So the optimal holding periods and the potential impact of the transaction costs are important pieces of the research process.

Q:  Could you give us specific examples that illustrate your process?

A: An example would be the emerging markets area. About a year ago, we had a lot of exposure there and we took a big hit when they sold off. The reason was that while some emerging markets funds fell in the rankings, others stayed highly ranked, and certain areas like China even moved higher. We don’t feel compelled to react to a sell off; typically, we move incrementally. If the sell off is the beginning of a real inflection point, new leadership will emerge and we'll make a thorough change.

But not all the changes are as dramatic. Often funds start to move up and replace other areas, even when they’re still performing well. For example, the REIT funds have done well over the last several years but in some of our portfolios they have been replaced by utilities funds. REIT funds continue to do well and who knows whether utilities leadership will persist, but our system works through incremental moves into the one area and out of the other.

Q:  What’s your view on risk control?

A: The portfolio construction in itself is related to risk control because, by design, we have to make sure that if we allocate money to a risk category, we would be comfortable buying any of the funds there. But once we have placed the funds in the risk categories, we continuously monitor for changes. The other important part is the liquidity. We recognize that sooner or later market leadership will change, and depending on the size of the fund that tops the ranks, we should be comfortable with the size of our position if we have to sell it.

Since we follow a quantitative strategy, we make sure that we don't take risk beyond our tolerance level for a specific piece of the portfolio. We measure the risk based on historical performance, mostly looking at Maximum Drawdown or the Ulcer Index. The standard deviation represents positive and negative deviations, the Ulcer Index measures only the downside. It is the average drawdown as opposed to the maximum drawdown.

Overall, I believe that most momentum strategies have a difficult time with implementation because they try to capture trends that don’t last very long. Others try to predict future market leadership based on valuations or pure subjective judgement calls. We have found that it is best to stick to our unbiased process in a consistent way. Sometimes it may work against us, but we only do the things that have proved to add value on average, over time. By controlling risk through portfolio allocation, we are able to patiently follow our strategy in a disciplined way, which I believe is the principle driver of our longterm success.