In the context of portfolio construction, the best thing about bonds is that they are not stocks. Their cash flows are generally far more certain, their lives finite, their terms transparent, and their rank in the pecking order of companies’ capital structures higher than common equity. All these attributes lend themselves to bonds being less than perfectly correlated with stocks, making them good diversifiers of equity risk. But not all bonds are created equal. Their diversification potential occupies a spectrum from great to lousy. The best diversifiers are the most boring.
As stocks continue to chug along and interest rates have lifted off from their recent lows and could climb higher still, now is a good time to revisit bonds’ role in a diversified portfolio. Here I will look at the degrees of diversification potential offered by various segments of the bond market. Then I will share a peek at recent trends in exchange-traded fund flows among the largest categories of fixed-income ETFs to assess how investors are positioning themselves to cope with the confluence of full stock market valuations and the possibility of further interest-rate increases.
The Spectrum of Fixed-Income Securities’ Diversification Potential
While bonds share many common features, they are not created equal. Bondholders are lenders, and as such, bonds’ characteristics will vary depending chiefly on the borrower and the length of the loan in question. If your dependable twin-sister/next-door neighbor—let’s call her Trish Bills—asked you to spot her a bit of money for a brief period, you probably wouldn’t hesitate and might not even charge her interest (she’s blood after all). But if you lent money to an out-of-state college buddy with a bad luck streak—let’s call him Hugh Yeeld—you would likely limit the amount you were willing to let him borrow, ask that he return that money sooner rather than later, and require a higher interest rate—all as means of limiting and compensating for the level of risk you would be assuming. Of course, what I’m describing here are the two risk factors that are the primary drivers of returns of fixed-income securities: credit and duration risk. These same factors also drive the varying degrees of diversification potential across fixed-income sectors.
Exhibit 1 shows the correlations between U.S. stocks (as represented by the Russell 3000 Index) and various bond benchmarks over the 15-year period through July 2018. It is immediately apparent that some types of bonds are better diversifiers of equity risk than others. For example, U.S. Treasuries had a negative 0.30 correlation with stocks over this period, indicating that they tended to zig when stocks zagged. Thus, lending to the U.S. government has proved to be a good way for investors to balance equity risk in their portfolios. On the other end of the spectrum, high-yield bonds had a high and positive correlation (0.73) to U.S. stocks over this span. Lending to less creditworthy borrowers (like your old buddy Hugh) is risky business. Credit risk correlates positively with equity risk. So investing in more credit-risky segments of the bond market like high-yield bonds and corporate credit won’t provide the same diversification benefits as investing in safer bets like Treasuries.
Variation in Correlation
Its important to note that Exhibit 1 reflects a snapshot in time. Correlations between different corners of the global market will vary over time. Exhibit 2 shows the rolling 36-month correlations for the same bond benchmarks featured in Exhibit 1 versus the Russell 3000 Index over the trailing 15 years through July 2018. While the rank order of diversification potential for these indexes remains intact, its clear there is a good degree of variation in correlations over time. For example, the rolling 36-month correlation between the Russell 3000 Index and the Bloomberg Barclays U.S. Treasury Index ranged from negative 0.03 to negative 0.72 over this period. The same figure for the ICE BofAML U.S. High Yield Index went as low as 0.35 and as high as 0.92. In my opinion, the most important takeaway from this exhibit is that correlations between bonds and stocks tend to jump in tumultuous times. This is most evident in the spike seen around the time Lehman Brothers went bust in September 2008. When markets crash, correlations converge.
Buy the Dip
It’s important to understand correlations between segments of the market and how they vary over time, but they don’t tell the full story. Yes, bonds are ballast for your portfolio. But they can also serve as a source of dry powder for use in implementing a simple portfolio rebalance or ratcheting up your equity risk during stock-market drawdowns.
Exhibit 3 shows the rolling 36-month maximum draw¬downs for the Russell 3000 Index as well as the bond benchmarks featured in Exhibits 1 and 2. The intent of this exhibit is to show you that when stocks drop, some bond types will keep your powder drier than others. The starkest example, which should come as little surprise, came during the depths of the financial crisis. The Russell 3000 Index experienced a maximum drawdown of 51.2% during the period. During the same time frame, the ICE BofAML U.S. High Yield Index had a 33.2% drawdown. The draw¬down for the Bloomberg Barclays U.S. Treasury Index was a relatively muted 4.3%. So, while correlations spiked as investors feared the worst, some segments of the market felt the sting more acutely than others. The most important lesson in all of this is that bonds’ usefulness as ballast and a source of dry powder is positively correlated to their credit quality and negatively correlated to their duration.