Q: What is the history of the company and the fund?
The fund was launched in the early 1980s and has always been a conservative core bond option for investors who want to add a low-volatility diversifier to the equity portion of their portfolios. Because the fund invests in investment-grade bonds, it can be effective in controlling exposure to more credit-sensitive sectors like high yield or emerging markets.
Although U.S. investors have moved over time to a slightly more aggressive core-plus style, this fund has stayed true to its relatively conservative roots.
The long-term struggle that puts us in, and a lesson learned which has led the fund to evolve, is having a less aggressive stance on a net-of-fee basis makes it harder to outperform the benchmark and peers while properly controlling risk. In the last few years, the fund incorporated our strategic beliefs in more international markets by using currency and international rate strategies. In some respects, this is unwinding and our focus is moving back on credit as valuations look more attractive.
Q: Would you highlight your investment philosophy?
Our strategic beliefs revolve around three key areas: in credit, in exposure to currency factors, and in global interest rate markets, where high real yields tend to outperform low real yields. These factors drive the structure of our portfolios and how we manage them.
We believe in having a credit overweight, though there is a lot of nuance in credit because it is quite common across the active manager universe. The hardest part is having the discipline to mean reversion in credit spread, particularly when spreads are low. Our portfolios are disciplined so when spreads are wide they have more credit, and when spreads are tight they have less.
Within credit, it is okay to go off benchmark, other than high-yield in this portfolio. Many securities are left out of benchmarks for arbitrary reasons in the bond space, and oftentimes it is possible to pick a liquidity premium or an illiquidity discount by investing in those, as long as the proper due diligence has been done.
Regarding currency, our two decades of research into currency managers has found the relatively simple factors of carry, value, and trend do outperform over time. Most active managers steer toward one strategy or another, but few actually run a disciplined process of exposing themselves to these factors consistently over time. We try to incorporate this into our portfolios.
Real yield explains itself. Countries with global bond markets that have higher yields net of inflation tend to outperform, either due to additional carry or due to compression over time in the international yields as the forces of relative value take over.
Q: What is your investment strategy and process?
The strategy begins with manager selection consistent with our strategic beliefs. Compared to other funds, our process is qualitative, and includes having boots on the ground to meet with managers repeatedly. Our depth of knowledge results from having experienced manager-research analysts as well as an internal database called Radar, which contains notes from every meeting we have had with managers as far back as the mid to early 1990s.
This database allows us to discuss investment ideas and evaluate managers over time. They cannot change stories after the fact because we have a record of their opinions versus what happened, and what actually drove value. It helps us determine who is likely to be able to repeat results going forward. Our focus always remains on future rather than past results, which naturally requires a qualitative assessment as much as quantitative screening.
When evaluating managers, our research side is independent of portfolio management in many respects. They make their own decisions on which managers they think will outperform over time, and will bring forward those who deserve our highest rating. Before a consensus is reached, we use a process called the sounding board where a candidate is critiqued by the head of research, the portfolio managers, and the other members of the research staff.
Ultimately, the research analyst, with approval from the head of research, decides whether to put someone on the buy list for portfolio management and the fund’s clients. Then the portfolio managers determine whether that manager has a role in the portfolio because there are far more buy-ranked managers than are roles available. My role is then to ascertain whether that manager would bring something unique to the portfolio, particularly net of the transaction costs. We remain highly cognizant of portfolio turnover and the cost associated with it.
This fund is a little different from its multi-manager peers, because it will hire standalone specialized managers who, by themselves, would not be sufficiently risk-controlled to fit into a core bond portfolio. For example, if we hired someone for their currency strategy and nothing else, we would use our investment capabilities to hedge their unacceptable factor bets so their job of producing alpha in currency still shines through in the portfolio’s overall results.
A similar process is followed when I want to bring a manager into the portfolio. In addition, the CIO of the firm and the CIO of fixed income evaluate my proposal to determine what it might do to the portfolios and whether it is consistent with where we want to steer our portfolios in terms of our Cycle-Valuation-Sentiment process, as well as the long-term investment proposition with that particular manager.
Q: Would you give some examples of the manager selection?
Right now, Logan Circle is the firm hired primarily for its credit capabilities, and it is a more aggressive credit manager. We like the reliability of their overweight to credit in the portfolios and know in certain market cycles or parts of the market cycle to hold less of them. At times when we do tilt with them like earlier this year, it is to get more BBB-rated credit in the portfolio. I know they are reliably always going to be overweight credit and it is going to be the biggest risk in their portfolios. Also, because the firm is relatively small, they can be nimble and take advantage of smaller issuer sizes to add value to security selection.
The Loomis Sayles team based in San Francisco was brought in more recently and is a more “steady Eddie” manager that generates consistent alpha, having had perhaps one negative year in the last 20. They add value through security selection across sectors and do so in a risk-controlled fashion without taking a long beta bet. This will never be a high-alpha portion of the portfolio because it is designed as a conservative core and is a great option to eke out sufficient value in excess of our fees on a consistent basis.
In 2012, there was a lot more to be lost than gained in credit. So we brought in the macro-currency group from the London, U.K. -based division of Principal Global Advisors, as a best-in-class currency manager to add value without exposing the portfolio to more credit risk. Currently, because the fund has been taking gains on the value added in currency by Principal, it has been steering away from them, but it has worked well for us.
Q: When selecting a manager, what key factors do you look for?
Whether in a securitized, currency, or credit marketplace, having managers who are best in class in one area is the ultimate aim, rather than others who are good at everything. Determining that is a difficult aspect of what we do. Abilities and willingness must line up with some sort of edge in one of those areas—perhaps a process others have overlooked.
Logan Circle, for example, offers much talent at the analyst and portfolio manager levels relative to their size of assets, which when combined with their willingness to be quick and nimble, makes the firm a rarity.
Loomis Sayles brings a singular process and a discipline few others are willing to implement in a portfolio. They do not do a lot of credit research in comparison to a Logan Circle. Instead, the firm systematically disintermediates the Street and takes the nickels and dimes from every single trade, which, at the end of the day, adds up to quarters and dollars. This takes a lot of discipline and takes a corporation willing to accept they will never knock the lights out but will be consistent. With them, it is as much their process as the people aspect we wanted to bring into the portfolio.
Another thing we always consider is past performance: does the performance stream match up with what we have assessed qualitatively? If not, it raises red flags and we will reassess.
We do not see the portfolio as a collection of managers but as a combination of exposures that must be managed. Although our high manager turnover has been criticized, it reflects our time-tested process. You cannot just set it and forget it. That does not work, or at best, results in a bumpier ride than necessary, so it is better to rotate through managers. They have their own cycles; and usually later in their respective cycles, managers tend to get more conservative and that is generally not good in a multi-manager context. A transition management desk keeps us abreast of the transaction cost to get managers in or out of a portfolio, and knowing this cost with accuracy is crucial to timing decisions.
Although we are not afraid to make changes, all our managers have been in the portfolio since 2012. Neuberger, for example, dates back to its inception in the 1980s, and Metropolitan West Asset Management was added in 2008.
Q: When do you make a manager change?
An example of why we might change managers is our 2014 replacement of Pimco with Loomis Sayles. We wanted a more conservative stance on credit exposures in the portfolio, and our research had already downgraded Pimco. It had organizational concerns, was a very punchy manager, and took a lot of active tracking-error risk—something we did not want in the portfolio at that time. The decision to bring in Loomis Sayles was driven by a market view as well as the relative research ranks.
Another change that was somewhat fortuitous was when Met West replaced Bear Stearns. We had fired Bear Stearns two weeks prior to the announcement of the collapse of its hedge funds. However, we did not know that was going to happen. The research staff had a buy rank on Met West, and we had been concerned about the organizational environment at Bear Stearns, so we made the decision to upgrade.
Q: Do you give managers a framework to follow in the portfolio construction process?
Ultimately, I am responsible for all the portfolio’s exposures. Practically, security selection rests with the managers, and we do not materially interfere 99.9% of the time because they are best in class and we hired them for this reason.
In a bond portfolio, macro calls can make a lot of money but also lose a lot of money quite quickly. Because this portfolio is conservative, we consequently limit managers’ flexibility. They are given guidelines, and in some instances the leeway to make macro calls, though certain managers receive more guidelines than others do. Logan Circle and Loomis do not have a lot of liberty to make duration calls, Principal can make currency positions, and Met West has latitude from a macro perspective.
We are wary of overly customizing the guidelines and basically playing a game of whack-a-mole, which creates unknown risk. Instead, being aware of the known risk a manager has allows us to transact directly in our portfolios and hedge that risk out if it’s one we don’t believe makes sense for the total portfolio.
Occasionally managers overlap, even though we hire for different areas of specialty. If two views on a security align, there will be more risk behind it and that’s what we want, to a point; if diametrically opposed, maybe that is a good reason to be neutral and find other places to take risk.
Q: How do you define and manage risk?
There are three ways we think about risk in a bond portfolio. Measuring tracking error around the Barclays U.S. Aggregate Bond Index is foremost. Also important are our total return and excess return in correlation to equities, and our drawdown against the benchmark. Because tracking error can hide non-normal distributions in fixed income, we remain aware of what our modeling is suggesting about how—and to what degree—non-normality is in the portfolio.
With tracking error, lower is not always better. Obviously for the same return, lower can be better, but when a return is less than fees, it is not. Our goal is to maintain the tracking error at a level where we can generate a credible net-of-fee return to clients. Currently we are trying to increase the tracking error, adding to credit risk. However, in 2014, we took down the tracking error by introducing more diversifying strategies to generate a higher information ratio to reach our net-of-fee targets.
