Sun, 7 Sep 2014
Though some assets are more sensitive to interest rates, even a conservatively tilted diversified portfolio has historically been able to produce positive returns through rising-rate environments, says David Falkof of Morningstar Investment Management.
Christine Benz: Hi, I'm Christine Benz for Morningstar.com. One of the most vexing questions facing investors is when interest rates will rise and what effect that could have on their portfolios. Joining me to discuss some research on that topic is David Falkof. He is an investment consultant for Morningstar Investment Management.
David, thank you so much for being here.
David Falkof: Thanks for having me, Christine.
Benz: David, I think this question is top of mind for a lot of investors--what rising interest rates will mean for their portfolios. Let's talk about the impetus for this research. It sounds like you were hearing a similar set of questions from your clients.
Falkof: That's right. Most of our clients are plan sponsors who are overseeing a 401(k) plan, a defined-contribution plan. And so they are very worried about some of their older employees. These employees typically have more conservatively built portfolios, which have lots of fixed-income exposure. And [our clients are worried about] how these portfolios may respond in a period of rising interest rates; the concern being that when interest rates rise, bonds tend to lose value. And these employers are worried that if their older employees see a steep loss in their portfolios, they may respond very negatively. They may not have an ideal outcome that they want for them. And so they're asking us to address that issue.
Benz: So, you went back and you looked at historical performances of various asset classes and various performances in various rising-rate scenarios. Let's talk about how you crunch the numbers and how you approach the data.
Falkof: We knew from the outset that if we were going to look historically at how various asset classes responded to different interest-rate environments, we needed to look beyond the 1980s. So, essentially the interest rates on Treasuries peaked in the early 1980s, and ever since then--the past 30 or 35 years--interest rates have fallen.
Benz: 1994 was a rare year when they didn't. But for the most part, there haven't been many observable periods.
Falkof: Exactly. So, if we went before that, we could then see longer periods of time when interest rates rose and that would give us a whole new set of data. Fortunately, there are few indexes out there; the Ibbotson Associates' SBBI [Stocks, Bonds, Bills, and Inflation] indexes go back to 1926. So, this is some of the longest historical yield data, return data, and we also can get some credit-spread data. With that, we decided to choose a five-year Treasury bond as our yield data; so that was our government-bond index. That's roughly in line with Barclays Aggregate Index, a common benchmark for most intermediate-term bond funds. So, we could get yield data and return data from that.
We could also apply credit spreads for typical corporate bonds as well as high-yield bonds. That would give us a corporate-bond and high-yield bond index going back to 1926. And then we had good stock return data back to the 1920s. One other asset class was cash.
So, we could build a reasonable portfolio with multiple asset classes and see how each of these asset classes responded over time. Then, we could also see how a diversified portfolio--sort of building it all together in a typical allocation we would build for a client--how that portfolio would behave over time relative to interest-rate environments that we saw.
Benz: And you looked at six-month periods. Why did you choose that increment of time? I think that, any time we're talking about bond holdings, we would want people to have a longer time horizon than that. Why did you home in on those six-month increments of time?
Falkof: We are very much proponents of owning a portfolio for a long period of time. Certainly, we would never encourage someone to own a portfolio just for six months. But what we were trying to address was, what would happen if there was shock to the market? We wanted to see if interest rates rose quickly or rose slowly over a very short period of time. How would a portfolio respond? And that would inform us as to how likely an employee, a participant, or a typical investor might react to that portfolio. So, we were trying to match the expectations of what a typical investor might see in the market.
Benz: Looking at the results, when you looked at asset-class returns in various rising-rate periods, I think some of the findings are pretty intuitive. You saw government bonds generally behaving the worst. That's sort of what everyone expects--that they will be the most responsive to changes, certainly in Treasury yields. Corporate bonds were somewhat less sensitive. But still, high-quality corporate [bonds] were still quite sensitive. I think one finding that maybe would be less familiar was that stocks weren't all that responsive to interest-rate changes. Let's talk about what you found there.
Falkof: A big part of the study was, first, identifying which asset classes are truly sensitive to interest-rate changes. We try to dissect this in a few ways. One way was just running a typical regression analysis. So, we're looking at the whole history of returns for stock, let's say, and then the whole history of interest rates. And we looked at six-month changes as well as the starting interest rate. We used a statistical term called R-squared, which helps us determine whether our independent variables--interest-rate changes and starting interest rate--are influencing the stock returns--the dependent variable. And we found for stocks that the R-squared was negligible--essentially zero--that interest rates are really not a major [factor]. Since the 1920s, on six-month periods of returns, interest rates are not a driving factor for how stocks return. Now, for government bonds, the R-squared is closer to 70%. So, it's a much more important factor when you're looking at government-bond returns than when you're looking at stock returns.
When we looked at other assets classes, it was also interesting. For cash, if you are looking at interest-rate changes, it doesn't really affect cash returns. But if you are looking at the starting interest rate, that explains essentially 90% of the returns for cash. For high-yield bonds, it is much closer to stocks--fairly low. And then, corporate bonds are in between where [interest-rate changes are] informative, but you have other factors influencing corporate-bond returns, such as spreads, default rates, and liquidity of the corporate bond. So, it was important for us to identify for each of these asset classes how important interest-rate changes were to the actual return of that asset class.
Benz: Backing up to the equity question: You looked at broad equity-market performance; you didn't drill into any sectors. I would imagine that it kind of depends on what slice of the stock market you own. Last summer, when we had a little bit of an interest-rate shock, we saw certain types of equities perform very, very poorly. So, investors need to really think about what's in their portfolios when they are thinking about how responsive they might be to this sort of interest-rate shock?
Falkof: It's true. We could have dissected the market into REITs, financials, technology, and various other sectors. And we would have seen essentially the same sort of differences; some sectors would have been very sensitive to interest-rate changes, others not so much. When we looked at the S&P 500, as a whole, it wasn't very sensitive to changes in interest rates.
Benz: One interesting finding, I thought, in your research was that high-yield bonds, as you said, are not particularly closely correlated with rate changes. But you said that even though they did not respond to rate changes, when they underperformed they performed much worse than government bonds did in this rising-rate environment. So, in what sorts of circumstances would one expect to see high-yield bonds have big losses, if not in rising-rate environments?
Falkof: The way that we set up our table is that we split it up into based off of the starting interest rate and then based off of the rate of change. One interesting thing that came out was that high-yield bonds have periods of time where they do perform very poorly. For a couple of those periods of times, there was overlap with when equities were also performing poorly. Now, this is not to say that that particular interest-rate environment caused high-yield bonds or stocks to perform poorly--because we already identified through the R-squared that interest rates are not a major influence.
But what was important for us to recognize was that if you're thinking about the way you are constructing your portfolio--if you're thinking about lowering your interest-rate exposure by going into high-yield bonds and going into another asset class--you are adding other risks. And those risks are coming through by the potential for steep losses. So, the factors that could have caused this may have been the credit risk of these asset classes. These may have been periods of times when investors just decided to flee high-yield bonds that became very illiquid, so the losses were particularly steep in those periods. But we were not trying to identify exactly what the other causes were so much as whether interest-rate environments really influence these. And for us, we found that the answer was no.
Benz: One other interesting dimension of your research is that you then looked at portfolios. You looked at portfolios with various asset allocations and kind of stress-tested those over varying rate environments. Let's start with conservative, move on to moderate, and then aggressive. Just talk about your general findings about how responsive these portfolios were to changes in interest rates and velocity interest rates.
Falkof: We looked at three different target-risk portfolios: a conservative, a moderate, and an aggressive. The conservative portfolio was a mix of 20% stocks and 80% bonds. The bond exposure included cash, high-yield corporate bonds, and government bonds. And that was actually an important part of the study in that we wanted to see for a diversified portfolio--particularly a conservatively tilted one--how that would behave in different interest-rate environments.
And we found that that portfolio was the most sensitive to interest rates, as to be expected. And that during periods when you're beginning at a low interest rate [followed by] rapidly rising rates, the [conservative portfolio] had performed worse than it did when there was a high interest rate and falling. Now, the important part that we found was that even across the worst potential environments that this portfolio could have gone through, it had still on average had positive returns. While the government-bond exposure and the corporate-bond exposure may have had some losses, the high-yield bond exposure, the cash, the stock exposures--all of those allocations sort of working together--still generated a positive outcome on average.
For the more aggressive portfolios, they behaved more like our stock asset class had, where interest rates slightly impacted a portion of the portfolio. But on the whole, they were behaving more distinct from how we would predict an interest-rate environment would affect it.
Benz: I think the thing I want to close on, David, is just talking about how you looked at a lot of historical data and it's really valuable research, but how predictive do you expect these past historical performance patterns to be? We see starting yields so very low, and I think a lot of bondholders think that when rates begin to rise, that could go on for a good long time, making bonds an unattractive asset class for a long period. How should one put these data points in your research into the context of what they might expect going forward?
Falkof: I think that's a really important question. When we were looking at this, we were looking at the contemporaneous response of each asset class to each interest-rate environment. We were not trying to have some sort of a predictive approach. We didn't say, if interest rates rise for six months, what happens in the next six months? That's a different study that could be done. But what we were trying to look more at was, during these environments, what can investors expect to take place? And so, it's really important to understand how things have happened historically and to take pieces of information from that.
We recognize that government bonds are very sensitive to interest-rate changes, and so that likely will persist. Now, how high-yield bonds, how cash, how some of these other asset classes respond--they are very loosely correlated with changes in interest rates. There are other factors out there that may be very important for stocks and other asset classes. So, you can look at economic factors, you can look at valuation for stocks, you can look at credit spreads for high-yield bonds. These are all other considerations that we didn't address in this study but are still important for investors to recognize [when determining] what sorts of risks they are exposed to in their portfolios.
Benz: David, thank you so much for being here. People can see your full research on Morningstar.com running in a separate article. We really appreciate you being here today.
Falkof: Thanks for having me, Christine.
Benz: Thanks for watching. I'm Christine Benz for Morningstar.com.
The Morningstar Investment Management group of Morningstar, Inc. includes Morningstar Associates, Ibbotson Associates, and Morningstar Investment Services, all registered investment advisors and wholly owned subsidiaries of Morningstar, Inc. All investment advisory services described herein are provided by one or more of the U.S. registered investment advisor subsidiaries. The Morningstar name and logo are registered marks of Morningstar, Inc.
The information, data, analyses, and opinions presented herein do not constitute investment advice; are provided as of the date written and solely for informational purposes only and therefore are not an offer to buy or sell a security; and are not warranted to be correct, complete or accurate. Past performance is not indicative and not a guarantee of future results.