Adam Hetts, Global Head of Portfolio Construction and Strategy, talks with Head of US Fixed Income Greg Wilensky about the challenges – and unique opportunities – associated with building bond portfolios in a world with historically low interest rates.
- While the Bloomberg Barclays US Aggregate Bond Index was once a good starting point for building diversified bond portfolios, it has increasingly become only that. Greg explains how the index’s inefficiencies can be exploited, enabling an active manager to potentially add value through higher returns, lower volatility, or both.
- Despite the steepening US Treasury yield curve, Greg views a sharp increase in rates or inflation as unlikely. The Federal Reserve’s (Fed) interest in stimulating the economy and investors’ demand for yield – coupled with US rates being among the most attractive developed world government bonds – should moderate the rise.
- Bond investors will need to balance the risks required to generate additional returns against sacrificing too much of the diversity they expect from bonds. Greg and Adam discuss how fundamental research and quantitative analysis may assist in both navigating the changing environment for bonds and building more optimal portfolios.
Adam Hetts: Welcome to Global Perspectives where we feature candid conversations with Janus Henderson’s thought leaders. I am your host, Adam Hetts, Global Head of Portfolio Construction and Strategy. And today we welcome Greg Wilensky. Greg is a Portfolio Manager and our Head of US Fixed Income based here in Denver with me. So Greg, thanks for joining today.
Greg Wilensky: Thanks for having me, Adam.
Hetts: Yeah, my pleasure. Greg, you are an active Fixed Income Manager and there is a lot of talk about fixed income benchmarks being broken, so we should talk about the Bloomberg Barclays U.S. Aggregate Index or what we lovingly refer to as the AG. It is sort of an albatross in your space, although it is ubiquitous, it is starting to feel like it is broken in the sense that the AG is structured in a way that is somewhat opposed to a lot of fixed income investors’ goals. So can you start by helping us understand the structure of the AG and why you think it is becoming easier to exploit as an active Portfolio Manager?
Wilensky: Sure, I will be happy to do that. Although I guess I will start, I will roll back a little, a little before. I think in certain respects it has always been relatively easy to beat for an active manager and we can try to bring that into perspective. But let’s talk about the good part first. You know, I think the AG is a reasonable starting point for discussions for a fixed income benchmark or, you know, an underlying idea of what a fixed income portfolio maybe is supposed to be about. It blends interest rate risk and some spread and credit risk, just to cover the basics. It combines, it was defined essentially to be a broad swath of the investment grade fixed income universe. It combines treasury bonds, agencies, agency mortgaged-back securities, corporates and then a little bit of securitized assets as well. I actually remember when it first changed from being just a government corporate benchmark to the aggregate index adding in the mortgages and other securitized.
Historically, blending interest rate risks and some spread and credit risk has provided a nice mix of some incremental income or return over time, while still having a base fixed income investment that can defend or diversify against the downside risk of an overall portfolio that has a meaningful equity exposure. So from that perspective, I think we should look and think of AG with adoration. But just because it was a good starting point, I don’t think it was really ever a great ending point and I think this has become more true over time. The construction of the index I think was limited by a number of factors, a lot of it being just data availability. This never really captured the full range of fixed income investments out there and that has only gotten, you know, to be more so over time as the markets developed additional sectors or security types and there has been some more complexity around that. There was also limits in computing power.
And then also I think there was some maybe narrow mindedness that drove the construction of the Aggregate Index that people, you know, didn’t want to bring in below investment grade corporate bonds into the Aggregate, because the way the Aggregate was constructed, you would have had a market weight, the high-yield component of the corporate index into the AG as opposed to saying is there something more efficient? What if we just added a little bit of below investment grade corporate bonds to the Aggregate Index? Would this improve the risk-adjusted performance of the Index without kind of negatively impacting its ability to diversify an overall allocation?
So I think some of those things existed over time and that is why when you look, if you look through long periods of history, the Aggregate, the people manage against the Aggregate Index using a core or core plus investment style have typically outperformed this Index over time. In fact, through over long periods of time, the Aggregate Index has often been in the third, you know, closest to the bottom of the third core tile. So I think this means there is definitely room for active managers to consistently add performance against that index. And as you noted, or you were implying, this has gotten to be I think a little bit even easier over time for a couple of reasons. A few of the things I mentioned before, more sectors of the, more sectors or security types available in the fixed income universe that aren’t covered by the Aggregate Index, because of those data availability issues. Also, frankly, the rise of passive investing and the fact that a lot of money is going into the allocations just as the Aggregate lays it out, means this creates inefficiencies or mispricings in the market, because money is flowing in or out. And an active manager who has got deep research, research, resources and expertise can exploit those. And then just more broadly, what we can do with the broader sectors, we can come up with a smarter, almost a smarter Index to start with, a smarter beta rather than the beta of the Aggregate, we can walk across the full range of fixed income, the fixed income universe and pull together sectors that not only provide incremental expected returns, which is nice. But since these sectors also provide diversification, we can actually create a better portfolio, one that has not only higher expected returns, but one that should overall do this without materially increasing risk, because it will benefit form increased diversification.
Hetts: I am glad your benchmarks are getting easier to beat, means more time for podcasts, right?
Wilensky: Yeah, don’t tell my boss. I have always looked at managing against an Aggregate Index with a broad opportunity set was one of the easier things to do, not without risk, of course. And there are times where you can be on the wrong side of things.
Hetts: And that index over market cycles, it has somewhat fixed, you know, rules-based construction, but it doesn’t mean that the performance of it doesn’t change over time. So I think over the last, I don’t know, decade, or even more since the global financial crisis, I think probably the key characteristics of the AG would be its level of duration or interest rate sensitivity and its yield. So can you kind of track the history the last ten years or so and how the duration and yield profile of that index has evolved?
Wilensky: So over this time, and it is the duration yield profile and, if you will, there is also that blend between interest rate risk and spread risk. And I think what we have seen over the last ten years or so is, well a couple of things, the amount of interest rate risk, you know, roughly let’s talk about measured by the mathematical duration. We have seen that increase over time. That can be a function of several things. One, just as rates have fallen, that naturally increases the duration of security, so a 30-year bond will have a higher duration in a low-yield environment than it will in a higher-yield environment, because the coupons, the coupons will be lower. There has also been a feedback loop as rates have gone down, it has made sense for lots of investors, whether we are talking about corporate investors or the US government in the form of treasuries, etc., to actually issue more long, maturity bonds, so that also increases the duration. And then even the mix, if you will, between treasuries and corporates and securitized has evolved over this period of time. I mean if we look at what has happened even recently, as the Federal deficit increases, that means the US government issues more debt and they have been frankly issuing over time extending the maturity distribution of their issuance, so that is increasing the amount of duration.
And then just as, on the flipside, the yield level has fallen as treasury yields have fallen. So when you look at things like yield per unit of duration on the index, that is certainly something that has deteriorated through time.
Hetts: So duration, or interest rate risk has gotten higher and you are getting paid historically low levels of yield to take on that high level of interest rate risk, especially in the face of such low rates. It sounds like a bit of a dangerous combination. One of the things I think you hinted at, Greg, was the mathematical duration, like you might see on paper versus what you can call maybe empirical duration or what happens in real life in different sectors when interest rates rise. So I guess for our listeners, what advice would you have when it comes to them interpreting those duration numbers they might see on paper versus what happens in real life across different sectors?
Wilensky: Yeah, unfortunately, this is one of those areas where fixed income does come with a certain amount of complexity and to do things right I think it adds, it adds performance or it increases the chance that we are going to generate better risk-adjusted performance, but it does make it a little more complex to explain. The mathematical duration calculation, which is something that all of us in the fixed income world. You know, it is kind of almost like the first thing we learn. It does do a great job of answering a very specific question, that is how much will the price of that security change if the yield on that security changes by a specific amount? But it turns out that question is not very relevant, especially when you are talking about a multi-sector fixed-income portfolio like the typical portfolios and strategies managed by whether it be passive managers or active managers. What we really want to do is think about how will a security or more importantly how will a whole portfolio react to a change in interest rates? And for this we can’t necessarily rely on just the closed form formula that can be used to compute the mathematical duration of the typical fixed income security.
So what we need to do, and understanding the differences between these two things is something that I have spent a lot of time thinking about over the last 25 years. I actually started thinking about this in some level of earnest, almost 20 years ago and thinking about how empirically different securities, whether we were looking at mortgages relative to Treasuries or corporate bonds relative to treasuries or two-year treasuries relative to ten-year treasuries might move. So I started to think about that and use that tangentially in my approach to portfolio management. But it was really after the GFC, the great financial crisis in 2008/2009, where this became a major focus. And that is because I observed firsthand the shortcomings of relying on mathematical durations for managing multi-sector fixed-income portfolios. The portfolios did not behave in the way that the duration would have said they should have behaved and that meant I really wanted to change the way I looked at that to make sure I had a more robust framework. And the other thing that had evolved, if you will, from kind of 2000 to 2008 was much more availability of data and much more significant computing power, which would allow for more sophisticated ways of looking at these empirical durations. So we could now look at them using risk models, we could use regression analysis looking at how a corporate bond’s price performance evolved for changes in treasury rates. We could look at the relationship between mortgages and treasuries or two-year treasuries and 10-year Treasuries.
A lot of this revolves around looking at the correlation in the change in yield of the security you are interested in versus the change in yield on the treasury bond. And we are really trying to think about what that relationship will be in the future and that is difficult to do. But what we need to recognize here is we know our forecast for that relationship will not be perfect, because it will change based on various market factors, but the important point is we don’t need to be perfect in our forecast, that is not the true hurdle. All we need to demonstrate is by incorporating these correlations or other tools that we can do a better job than using the mathematical duration.
So what does that mean for a multisector portfolio manager? It means things like when we are overweight spread product, things like corporate bonds or securitized assets versus the index, it is likely the case that the mathematical duration of our portfolio will overstate the true empirical behavior that we will experience. Because when treasury yields rise, typically the spreads on corporate bonds or the difference in yield between the corporate bond and treasury will compress. That means the yield on the corporate bond will not go up as much as the yield on a treasury bond and hence it will have a, its price will move less and it will display a lower empirical duration than its mathematical duration.
Hetts: That is great, so that empirical duration or in other words what investors will actually experience as the markets are moving around, you are talking about a lot of it depends on the inner play of interest rate movements and interest rate risks and then also credit spreads and credit risk in a portfolio. So maybe let’s dissect those a little bit and see where you think each of those are moving and maybe we could start on the rate’s piece for a minute. So on the rate’s markets, you have already seen a bit of curve steepening, meaning the longer-term rates have been increasing, but the short end hasn’t moved much. So what is your outlook for this year and where we go on the long end, call it the US ten-year versus the short end of the curve?
Wilensky: Actually, some of this builds very nicely on that empirical discussion. While I know you had just highlighted the relationship between the corporate spreads and the treasury yields, there is a similar empirical relationship that I think is incredibly important right now and how, and will drive how kind of the yield curve shape will evolve over time. And this comes from the fact that as most people probably are aware, the Federal Reserve has brought policy rates down to zero. They are actually, and they are going beyond that by buying $120 billion of fixed income securities a month. And they pretty much committed that they are going to keep that policy rate at a very low level for a number of years to come. What this means is the two-year, the front end of the curve is a fairly lockdown. We do not expect to see a lot of yield movement there relative to what can happen out the curve. So this means when, let’s say good economic news comes out, the 10-year is going to rally more than the two-year and the yield curve will flatten. And just as we have seen earlier this year, when certain things happen that let’s say cause us to view, strengthen the economy or that economic growth will be faster in the future, we are seeing the yield rise and steepen. So the 10-year will move more than the two-year. In other words, we are saying the movement in the two-year yield, the empirical duration of the two-year is going to be much less than its mathematical duration relative to the 10-year. And I think that is going to continue to happen. We are going to see much more volatility at the 10-year and the 30-year part of the curve than the front end of the curve, for example, two-year Treasuries.
Our view, while there are some factors that will push in either direction, we do believe that we will see a modest, a further modest to moderate increase in Treasury yields, especially the intermediate and longer-term Treasury yields as 2021 evolves. We don’t think this is going to be a real sharp increase in rates and that is because there are some factors on the other side, such as the aforementioned fact that the Federal Reserve is buying $120 billion a month in fixed income securities and that U.S. rates, while they are absolutely low in historically context, and frankly lower than they should be given where we believe the US economy will go, we should not lose sight of the fact that U.S. yields are much higher than maybe, and U.S. Treasury yields are much higher than many other global treasury yields. So there are a lot of investors outside the U.S. who are selling their local treasury bonds and coming to invest in whether it is U.S. treasury bonds directly or other U.S. fixed income securities. So this will, I think, limit the amount of increase that we will see in US treasury yields, but we definitely think the curve will display a steepening bias as rates rise.
Hetts: Yeah, it is a sign of the times that a 1% U.S. Treasury feels rich on a global standard. And on the short end, especially with the new average inflation targeting, like that response mechanism is a lot fuzzier than it used to be. So to get the short end to move, like what is it going to take? Is it inflation north of 2% that persists for a couple of years? Or is there any way that the short end could actually start increasing sooner than that?
Wilensky: I think the short end is going to be extremely tied to what investors believe the Federal Reserve will do. And as you were talking about, their new average inflation targeting regime and the fact that they are indicating, they are now looking for two things to occur before they will raise rates, both that inflation starts to get up above that 2% target. At the same time, we are achieving full employment. I think this means the runway for when they will raise rates is much longer than what would have historically been the case, where it was a little more of maybe even an either/or exercise. If we receive the prior, if you will, prior to 2019, they started to raise rates because they felt like we had reached full employment, even though this had not started to pass through to inflation. So I think we are going to have to see movement on both the inflation front as well as on the employment front and when that starts to happen, people will start to, as that starts to evolve, we will see people start to reevaluate when they think the Federal Reserve will start to raise policy rates. I think until that becomes clear, it is hard to imagine the two-year yield moving very significantly from the levels that it is at. Because you can think of it as an average, if you will. The two-year Treasury yield should be somewhat of an average, a weighted average of what you expect policy rates to average in the following two years.
Hetts: So there is something there that I am also trying to wrap my head around. So the two-year can be relatively anchored to your point. In the meantime, inflation could kick up even just some short-term kind of headline inflation. If that does happen, you know, nominal rates kind of anchored and then real rates would have to decrease. So we have this environment where nominals are anchored and real rates start decreasing, because inflation is increasing. I assume TIPS benefit from that. Do you like TIPS in the face of dropping real rates? And then outside of TIPS, as maybe an obvious answer, are there any other sectors you like a lot in the face of real rates going lower?
Wilensky: So on TIPS, I guess the framework I would use for thinking about TIPS with around how we expect inflation to evolve is that break evens, break even inflation rates, which is the difference in the yield of the TIPS security, the real yield on the TIPS security and the yield on the nominal treasury security, you can think of this as what would inflation have to average over the life of the TIPS or treasuries so that you would be indifferent between owning TIPS or treasuries. So if Treasury yield is 50 basis points and the TIPS yield is minus 1%, then the break even inflation rate is 1.5%. If inflation comes in over the life of that security at 1.5, you are indifferent between holding the TIPS security or the similar maturity treasury. If you expect inflation to come in above that 1.5%, you will be better off, well, if you expect and it is realized, you will be better off owning that security over its life. So as we look at, as we try to evaluate the decision of adding something like TIPS to a portfolio that has a nominal benchmark, we are trying to evaluate both our long-term inflation forecast versus that break even inflation rate, as well as how we think kind of market expectations for that will evolve over shorter periods of time.
And certainly as we were looking later, in the latter half of 2020, as we gain more confidence about the rebound in the U.S. economy, our ability to generate vaccinations, etc., we did begin to add TIPS positions in lieu of treasury bonds into our portfolio. The break even inflation rates have moved up quite dramatically actually in the last two or three months over that time. But we still think there is some, there is still additional value to be had that these break even inflation rates will rise in the future as people’s inflation expectations do rise, as the Fed is able to demonstrate, essentially able to demonstrate that they are able to get inflation moving towards their target over time.
Hetts: Okay, so Treasury is at 50 BPS, like in your example, or even at 1%, that feels expensive for U.S. treasuries on a historical basis. But it is actually relatively cheap compared to global sovereigns. And I think that is the big irony in fixed income asset allocation for all the clients my team works with is that historically you saw safety in price’s protection in treasuries. But now with rates so low, those treasuries might actually be the source of a lot of risk in fixed income portfolios if rates increase. So Greg, as your team is building multi-sector fixed income portfolios, how are you striking that balance between using high-quality fixed income for crisis protection, but then also minimizing losses from rising rates?
Wilensky: That is the robber of the challenge that I think we always have to be looking at in an environment where we may be facing rising yields. Because especially when we are talking about kind of a core fixed income allocation. We are always asking ourselves what is the right blend, if you will, of interest rate risk and spread risk? We want to generate that extra, we want to generate income in returns, but we also recognize as I noted earlier that a big part of the role of the core fixed income allocation is to provide that downside protection in case let’s say equities were to sell off dramatically. So that does mean we want to squeak or blend interest rates and spread risk.
Right now, given our view that rates will modestly or moderately increase over time, even given the fact that as you noted, U.S. Treasury yields do look attractive relative to many of their global counterparts. We think as the U.S. economy strengthens as we come out from the COVID pandemic, as the global economy strengthens, we will see some upward pressure on yield that can, I am not going to say overwhelm, but will cause rates to move from this point higher over time. But we don’t want to do so much out that we lose that downside protection. So what we are going to rely on at that point, we are going to rely on our research to not only help forecast where rates are most likely to go, but we spend a lot of time thinking about the range of the outcomes, not just the most likely outcome or the average outcome.
We will also spend time trying to forecast what the correlation will be from the change in interest rates and change in other market factors, such as the spreads on corporate bonds or securitized assets. And by combining that information, our views on rates as well as how in different rate scenarios, we might see things like corporate bonds or mortgages or commercial mortgage-backed securities evolve. We can use our risk tools to develop a more robust forecast of the expected returns in their distribution for various portfolio structures. And we look to strike the right balance of returns and protection by looking across, looking across those outcomes.
Hetts: All right, well said. Yeah, I always reason to watch and adjust level of interest rate risk in a fixed income portfolio, but at the end of the day, you want your bonds to feel like bonds, especially when you need them to feel like bonds. So that is the role of interest rate risk ultimately. And then outside of the interest rate discussion, let’s get back to the other part of what you laid out earlier on the spread markets and credit spreads. We have got investment grade running with tighter spreads and high yield, but then again, investment grade now is this implicit backstop from the central bank, even if it is not an actual backstop at the moment. So how are you looking at the trade outs between investment grade and high yield and how does that investment grade backstop end up affecting the high-yield markets?
Wilensky: Really good questions, Adam. How we were thinking about this, because I think it provides a good base for us to talk about our current views. When we, as we were exiting the first quarter of 2020, spreads had really blown out. There was a massive amount of uncertainty about what would be happening from an economic and from a health perspective. But we took a lot of confidence in the fact that the Federal Reserve and also the fiscal authorities were being extremely aggressive in supporting the markets, liquidity and the economy. the fact that as you noted, the Fed had provided a more explicit backdrop for investment grade securities that they had rolled out programs that were focused, said they could buy investment grade bonds or eventually some amount of high yield, but only things that had been downgraded from investment grade after the crisis began. And also the fact that investment grade bonds on a per unit of risk basis actually underperformed during the crisis.
This is something that I don’t know that everyone really focused on. During the crisis, while high yield bonds may have gone down by a higher absolute amount, we expect that to happen. They are, they inherently are riskier investments. So we try to evaluate these securities on a similar risk basis. It might be the fact that you would buy two and a half units of an investment grade bond and sell one unit of a high-yield bond in order to get similar risks. So when looked at on a risk-adjusted basis, those investment grade bonds actually underperformed, because it was the investment grade bonds that people could sell during the crisis. These were the things that were still liquid, so essentially an illiquidity premium almost built up in the investment grade bonds, because these were the things that everyone was looking to sell in order to raise funds for a variety of reasons. investment grade bonds for the better ability to withstand uncertainty, plus the fact that we felt they had underperformed on a risk-adjusted basis and they had that more explicit or more implicit Federal Reserve support. As the year evolved, we became more confident in our economic forecast that things like the vaccinations would be effective and things like that. And in addition, that underperformance on a risk-adjusted basis, that we highlighted, began to reverse and in the second, especially the second quarter and part of the third quarter, investment grade bonds outperformed on a per unit of risk basis versus high yield.
So with that backdrop, we began to rotate our exposure from investment grade bonds into the highest quality below investment grade sectors, for example, BB rated corporate bonds. And we still, at this point, we still think that trade looks attractive. If you look at the ratio of spreads between BB bonds and BBB corporate bonds, even as both BB and BBB spreads have narrowed quite a bit over the last nine months, the ratio of those spreads has expanded. So right now the ratio of BB spreads to the BBB spreads are about 220%. Over the last five to ten years, that ratio has typically averaged around 175% or 1.75, the BB spread is 1.75 times the spread of BBB spreads. So even though there may be a little more implicit support for the investment grade, investment grade corporate sector, we believe this spread gap between BBs and BBBs, means that on average, the BB bonds are more attractive. And by relying on our research analysts who know these companies and industries and can help us determine which are the securities or issuers that are attractively priced.
Hetts: All right, Greg, thanks again for your time. You are certainly a wealth of great information, a ton going on in your space, and like you said, an excellent environment for active investors in fixed income. And as always, the views of Janus Henderson’s investment teams and thought leaders are freely available within the Insights’ section of our websites. We look forward to bringing you more conversations in the near future.