Exploring the High Yield Spread’s Predictive Power

What can we learn from bond investors’ fear of default risk?

Riskier borrowers have a higher chance of defaulting than safer borrowers, so their bonds have higher yields. Generally, the more concerned bond investors are about default risk, the more yield they demand for riskier bonds compared to safer bonds. A credit spread measures the difference between the yields of bonds with different credit qualities; investors often use it as a proxy for the market’s perception of default risk.

For this post, we’ll look at the difference between the yield of high-yield bonds and U.S. Treasuries, which is captured by the ICE BofA US High Yield Index Option-Adjusted Spread data series, downloaded from FRED. As you can see in the image below, this spread tends to spike during recessions (shaded regions). In economic downturns, bond investors’ fears about defaults tend to increase, at least initially.

We’ll see what it can tell us about future market conditions. We’ll look at the S&P 500, high yield corporate bonds, investment grade corporate bonds, VIX short-term futures, and small-cap value stocks.

Relationship with SPY (S&P 500 ETF)

From 2000 to present, there’s a positive correlation between the HY spread and SPY’s volatility over the next 30 days; higher spread values are associated with higher levels of future volatility. Interestingly, this relationship has been weakening over time. Between 2000 and 2007 the Pearson coefficient was 0.789, but from 2014 to present it fell to 0.258.

Instead of future volatility, what happens if we look at the difference between future volatility and trailing volatility? We find a negative correlation that has been strengthening over time. Higher spread values are associated with more negative values of (future 30-day SPY volatility — trailing 30-day SPY volatility). The correlation could be stronger though. From 2014 to present, its Pearson coefficient was -0.33.

The spread’s relationship with SPY’s future 30-day returns has been changing over time. The two variables were negatively correlated early on but have been positively correlated as of late.

Let’s see how the spread relates to (future 30-day SPY returns — previous 30-day SPY returns). These variables have a positive correlation that has been strengthening over time, currently standing around 0.359.

The spread is also related to SPY’s future behavior relative to VBR (small-cap value ETF) and IVW (large-cap growth ETF). The spread is negatively correlated with (SPY future 30-day volatility — VBR future 30-day volatility). From 2016 to present, this correlation had a Pearson coefficient of -0.46. This means higher values of HY spread are associated with higher future volatility in VBR relative to SPY’s future volatility.

In contrast, the HY spread has a positive correlation with (SPY future 30-day volatility — IVW future 30-day volatility). From 2016 to present, this correlation had a Pearson coefficient of 0.313. This means higher values of HY spread are associated with lower future volatility in IVW relative to SPY’s future volatility.

Relationship with HYG (High Yield Corporate Bond ETF)

From 2007 to present, there’s a positive correlation between the HY spread and HYG’s return over the next 30 days. This relationship has been strengthening over time, and its Pearson coefficient currently stands at around 0.439.

When we compare the spread to (future 30-day HYG returns — previous 30-day HYG returns), we see a similar relationship.

Next, let’s compare the spread to HYG’s future volatility over the next 30-days. These variables have been positively correlated from 2007 to present, though the relationship is weakening a bit. From 2016 to present, its Pearson coefficient was around 0.489.

The spread also relates to the difference between the future 30-day volatility of HYG and LQD, the iShares Investment Grade Corporate Bond ETF. Higher values of the spread are associated with higher values in (future 30-day HYG volatility — future 30-day LQD volatility). This relationship has been weakening somewhat over time. From 2016 to present, its Pearson coefficient was 0.483.

Relationship with UVXY (VIX Short Term Futures ETF)

From 2012 to present, there’s a negative correlation between the HY spread and (future 30-day UVXY returns — previous 30-day UVXY returns). This relationship has been strengthening over time, and from 2018 to present it had a Pearson coefficient of -0.473. Higher HY spread values are associated with future UVXY returns being lower than previous UVXY returns.

Relationship with VBR (Small-Cap Value ETF)

From 2004 to present, there’s a positive correlation between HY spread values and the future 30-day volatility of VBR, but the correlation has significantly weakened over time. From 2004 to 2009 its Pearson coefficient was 0.834, but from 2015 to present its coefficient was 0.364.

In contrast, the HY spread’s relationship with (VBR’s future 30-day returns — VBR’s previous 30-day returns) has been strengthening over time. From 2015 to present, that relationship had a Pearson coefficient of 0.395.

Conclusion

The HY spread is definitely related to future market conditions. Investors can benefit from using this indicator to shape their expectations, but they should be mindful that its relationship with other assets has been changing over time.

Special Note

My dream is to work on Wall Street as a trader, analyst, or quantitative researcher of some kind.

This May, I will complete my Master of Quantitative Finance program at the University of Maryland, College Park. Feel free to reach out if you know of any upcoming full-time opportunities in finance or data science.

You can reach me on LinkedIn, or by emailing me at MitchellRosenthalOfficial@gmail dot com.

Cheers!