Q:  Would you give a brief overview of the fund? A : We manage three funds in our team – the Core Bond Fund, a U.S. Government Trust, and the Limited-Term Government Fund. The way we think about these funds is that they are supposed to act as a ballast in an investor’s portfolio. Q:  What are the main beliefs that constitute your investment thinking? A : Our belief is that if an investor makes an allocation into a short-term government bond fund and it does not do well in a down market, the entire reason for the allocation is negated. Accepting this thesis has numerous implications with respect to the investment horizon. Therefore, a fund like ours cannot be a deep value strategy or have an investment horizon of about three years, because if an investor has allocated in this fund and it goes down 10% when it was expected to hold value, then the individual is foregoing the opportunity to invest in equities, which they could have done otherwise. As a result, such a move shortens the investment timeframe quite a bit. All an investor wants is to hold value in a down market in one part of the portfolio for a reasonable investment horizon, which we consider to be between six and twelve months. Q:  What is your investment process? A : The internal benchmark for this portfolio comprises 85% one to three year U.S. government securities and 15% U.S. government agency mortgage-backed securities. While we do not exclusively focus on those sectors, the primary universe that we invest in is U.S. government bonds, U.S. government agency bonds and agency mortgages. That is really the opportunity set for us. We can invest about 20% of the portfolio in commercial mortgage-backed, non-agencies or asset backed securities, although we generally limit that exposure to 10%. Yet, the underlying principle of balance applies to them just as much as with any other asset types. For every security in our portfolio, we calculate the exposure in terms of various points on the interestment curve using proprietary techniques like principal component analysis and a historical data set. We also determine its exposure to two-year rates, five-year rates or 10-year rates and the overall duration curve calculation. We roll it up to the overall portfolio and on the benchmark level, and then, as we compare the two, we try to see if there is any slippage that may need adjusting with the help of our daily hedging program. Our process is different in how we go about figuring out what the logic for our exposures is. The best way for us to do that is by taking data out of a tool like Yield Book, an industry standard, and manipulating that data to get comfortable with what our exposure level is. By doing that we figure out what the duration and curve profile of the portfolio is going to be, which concludes our first process item. The second process item is premised on how much we want in agency mortgage-backed securities and, within the agency space, what types of securities we need at any one particular point. Here, we can play in the entire gamut of agency mortgage-backed securities, beginning from pass-throughs, specified pools, collateralized mortgage obligations, interest-only securities or inverse IOs. The first order of business is to figure out how much exposure we want in agency mortgages, before figuring out within the agency mortgage space whether we want it in 30 years or 15 years, or to invest in collateralized mortgage bonds or other asset backed securities. We will arrive at that decision by looking at where the valuations are today and if there are any overarching macro themes that may have implications for one particular part of the portfolio. For instance, coming out of 2009, we made a determination that, first of all, agency mortgages were very cheap because mortgage spreads had widened dramatically. Additionally, we thought that because we were going through a credit crunch, banks would do whatever the agencies tell them, which would result in prepayment rates being at very modest levels. In our view, that proved a solid argument for buying lots of agency mortgages and higher coupon agency mortgages. In fact, that line of reasoning was especially true for specified pool agency mortgages where we got a good burnout after what had been prepaid was treated at the same valuation level as generic higher coupon mortgages. All in all, finding such types of securities in this space was very profitable in our view, we implemented that and the portfolio performed well in 2009. By 2010 it was very clear to us that mortgage spreads had tightened, and as the slow prepayment story had played itself out, there were prospects of a government mandated refinancing program. Before mortgages started actually widening, we sold all our higher coupon bonds, replacing them with lower coupon bonds to adjust for that particular strategy. To sum up, our strategy is about picking how much we want in mortgages, which part of the mortgage market we want to be in, and what individual security we like to hold. To some extent this methodology is driven by the valuations in the marketplace, but it is also overlaid with our assessment of the housing market and macro policies as well as what the Federal Reserve is doing with respect to different parts of the mortgage market. Q:  What drives your research process? A : The drivers of how much we want in mortgages are our expectations of prepayments. The Yield Book has its prepayment model, but we go on from there by adjusting it to what we expect based on our experience and data. The process involves using prepayment modeling and valuations along with figuring out the Federal Reserve’s policy and taking into account what that means for agency portfolios and for a specific part of the portfolio. As a further step, we evaluate the net and historical result in order to understand how close they are to what we were expecting based on the course of rates. At that stage we also analyze what that means for prepayments and how things should have performed, as we adjust our assessment based on reasonably short-term experience. We understand that it is irrelevant to track five years’ worth of data because in such crises the regime changes quite dramatically. Therefore, what we thought was the right regime for a six months’ span may be very different over the next six-month period. To react in an adequate and timely manner, we ensure that we consider long-term data for our evaluation of the prepayment model but we also adjust the results for a shorter timeframe, so that we could adapt our model relatively quickly before we find out we are underperforming severely. Q:  What are some of the steps you take to avoid losses induced by macroeconomic events? A : We do recognize that this is a government bond fund and in buying a government bond fund the investor has taken on U.S. sovereign risk, so cutting it off altogether is going to be an impossible task. The interesting thing is that if a sovereign crisis manifests itself and it is U.S.-centric, the likelihood that a government bond fund would be immune to it is not practical, but that also implies that we can adjust our duration in a certain way to take advantage of the situation. One aspect of our research process is based on ways of incorporating that thinking into our macro analysis. We figure out and evaluate some strategies that we can implement for tail hedging. For instance, in 2011, when the debt negotiations were going on in the U.S. Congress, we reacted to those developments with a longer duration in the portfolio. We felt that it made sense because while people would be concerned about U.S. credit, they would be far more concerned about equities and consequently running out of equities into government bond funds. Q:  How do you build your portfolio? A : We have specific limits in the portfolio in how much can be in mortgage-backed securities or asset-backed securities. However, the big risk that we take in the portfolio is primarily associated with agency mortgages. After the backstopping of agency mortgages, it is duration risk and curve exposure risk that comes about rather than the underlying credit risk of the agency. Of couse, we apply strict diversification targets and risk controls at any of these levels. Our diversification in a bond context depends on how much duration we can take relative to our benchmark or how much exposure we can have relative to various points on the curve. Q:  When do you decide to sell a security? A : One of the reasons to sell a security would be a decision to reduce our allocation in a particular sector. The next driver of our sell decision is valuation. The final incentive to sell would be a significant underperformance in a particular sector of the market. We do not want to have such exposure and we are going to exit that sector completely or lower the exposure. Even if there are attractive securities that are good IOs and even if we like the sector, the subsector and security valuation, we may decide to sell if we want to reduce our overall exposure due to our macro policy. Q:  What are the primary risks that you focus on? A : The biggest driver of our entire risk process is our balance thinking. The fact that we are going to invest a large portion of the portfolio in agency mortgages has to be very tightly specified and adhered to, and there cannot be any room for discussion as to whether we have violated the risk limit or not. To address risk-related issues effectively, we have established a separate risk management department. We have all agreed upon the framework for risk calculation and risk models that we are going to use in line with decisions by the risk management committee. Furthermore, we execute a more detailed risk control process within the investment team with regard to any concept in the portfolio. If at any one particular point it becomes clear that the risk profile of the portfolio is becoming too uncertain and we cannot define it specifically enough, we establish a stop loss regime for a particular security. The first and foremost driver of risk control is the level of comfort with respect to duration exposure and the various curve exposures of the portfolio. What makes this critical is the fact that we invest quite substantially in agency mortgage-backed securities, which may provide good income but their duration characteristics are somewhat ethereal and curve exposures are far from certain.