Once we have a shortlist of managers that we can talk to, it comes down to whether the manager fits within our objectives. How the managers correlate with the other managers that we may utilize within the portfolio is important. Then we optimize the allocations based on which managers will produce the best risk-adjusted return for the portfolio.

All the managers run separate accounts, so we’re not giving our money to a hedge fund or a limited partnership. We’re actually opening up accounts at our custodian, Citigroup, and the money is managed within that sub account.

Q:  How many fund managers do you monitor regularly and how often do you change them?

A: We’re continually looking for new managers and strategies to improve the optimal mix of the portfolio. We’re trying to be fairly passive with our model once we have it set, but we are looking for strategic or opportunistic changes that we can make going forward.

So we’re constantly monitoring our current managers and strategies for style drift, alpha, and beta across the portfolio and relative to various factors. We're monitoring the capacity, the fees or other significant changes, such as a change in a portfolio manager. Across the portfolio, we make sure that there’s not much overlap between managers. We’re pretty happy with the number of managers that we currently have, but we may add strategies as the opportunity arises.

Q: What are the prospectus limits in terms of the number of managers or strategies?

A: We’ve a maximum of 20% of the assets in a single sub-advisor’s strategy. Currently, we’re allocating to 12 of our managers across the portfolio. These allocations range from few percent up to 15%. When we were building the portfolio, we had a manager with 18.5% allocation because it had a very high investment minimum. For these largely institutional type separate accounts, investment minimums tend to be very high, up to $10 or $15 million. Some of the minimums are higher, but we’ve negotiated those down so that we can use them in our portfolio.

Q:  What would lead to a change in the mix of strategies?

A: It comes down to risk management because we’re trying to maintain a fairly low beta relative to the overall market. We don’t necessarily benchmark to any index, but we look for an overall portfolio beta of 0.2 or less to the S&P 500, as well as a volatility benchmark of 4% to 6% in terms of standard deviation. This is where we want to be in terms of our portfolio risk management.

Anything that may affect the portfolio in terms of beta or volatility, will ultimately change our optimization. There’s a variety of beta factors that we look at, such as the exposure to large-cap or small-cap versus international markets, the exposures to any certain asset class or sector, the single security exposure, etc. Then, if any of the trends in the beta relative to these risk factors is upward, that may be a reason to re-optimize the portfolio.

Q:  Who makes this decision? Is it a team or an individual decision?

A:We have an investment team that’s made up of four people: myself and the other co-portfolio manager, Alec Petro, along with our two quantitative research analysts. We talk about the portfolio and monitor it on a daily basis. Quarterly, we look at any potential strategic or opportunistic allocation during our investment committee meetings. Monthly, we check if there are any real exposures that we may need to act on immediately. But in terms of the performance of our sub advisors, we hold a more long term view on whether the managers are performing relative to our alpha forecast and expectations.

Q:  How do you approach portfolio construction?

A: Most of it goes into our quantitative model and we don’t have many traditional qualitative views on the construction, other than certain constraints, such as the allocation limits for single sub-advisor strategies. Then it comes down to a mix of alpha that’s available to us. Our quantitative model takes the track records of those return streams, tries to forecast alpha going forward, and then mixes and matches the strategies in a way that produces the allocations of the portfolio. So the construction really comes down to our selection of sub advisors and once we have those, the model decides how much is given to any one strategy.

Q:  Could you describe that model in more detail?

A: Because we have diverse sources of alpha, we can’t really maximize return per se, except to select different managers. It's up to the managers to produce return. What we can do is manage the risk. The model attempts to maximize risk-adjusted return or Sharpe Ratio; that's its ultimate goal. We try to improve the Sharpe by attempting to minimize risk and volatility of the overall portfolio through allocation optimization.

Q:  So you focus on risk and volatility and the managers focus on returns?

A: Well, they are focused on risk as well. One of the selection procedures involves their proven risk-adjusted investment performance and their adaptability to various market conditions. We’re not looking necessarily for the managers that can shoot the lights out, but for those who have a risk-adjusted return focus or an absolute return focus. That's why we examine their performance in a variety of market conditions, such as the bear market in 2002. We're trying to avoid the managers with significant draw-downs when the market does the same.

We don’t want the managers that simply behave like the market. As you know, when the market is down 20%, if mutual fund managers lose 15%, they usually consider it a phenomenal year. For us, that would be poor performance. Certainly, we can’t guarantee an absolute return performance, but we can focus on risk-adjusted returns as a primary approach.

Q:  What kind of risks do you perceive and how do you monitor them, other than the volatility that you mentioned?