- Credit spreads are an important indicator of market risk.
- A recent article proposes using duration-matched ETFs as a risk proxy.
- Bond ETFs however are not an accurate measure of yield spreads.
In a recent article on Seeking Alpha, Returning To High School, Investment Style, Cam Hui discusses a risk measure, junk bond spreads, that we use in our indicators at Williams Market Analytics. The author astutely points out the importance of duration matching. Mismatching bonds of different durations, as demonstrated, can greatly distort the measure.
However we find Mr. Hui's analysis problematic due to the use of ETFs as a proxy for yields. To recall, many use the spread between iShares iBoxx $ High Yield Corporate ETF(NYHYG) and iShares 20+ Year Treasury Bond (NYTLT) as a proxy for market risk. The author proposes matching the 4.0 year duration on HYG with the IEI (iShares 3-7 Year Treasury Bond ETF). The error committed by Mr. Hui is assuming the ETF prices reflect yields on the underlying bonds. Unlike the perpetual 10-year U.S. Government Note contract (TY1) or the perpetual 30-year U.S. Government Bond contract (US1), in which cases the prices do have a perfect inverted correlation with the 10-year and 30-year yields, bond ETFs do not perfectly reflect the corresponding yields. The reason is that bonds ETFs pay frequent (usually monthly) dividend payments, and these ETF prices drop on each ex-dividend dates.
To correct Mr. Hui's analysis, we ran his HYG-IEI spread using full prices (total return). For those on Bloomberg, simple change your DPDF setting to reinvest dividends on each ex-dividend date to obtain the full price. The top chart shows the same chart in the original article which uses market ETF prices and the "forced trendline" inserted by the author.
The next chart uses total return ETF prices, which should reflect a near perfect inverted image of the underlying yields. What we get is both a theoretically better result AND a very interesting and significant (5 perfect hits) 3-year trend line which was indeed broken recently.
Finally, the author also cites another credit spread, Emerging Market Debt over Treasuries. In this case the author uses the EMB (iShares J.P. Morgan USD Emerging Markets Bond ETF) and the IEF(iShares 7-10 Year Treasury Bond). We see that, using total return prices, this ETF spread is at its highest level since at least 2008. We chart both the total return price spread between these ETFs as well as the market price spread here. Once again, the difference is significant.
In conclusion, at Williams Market Analytics, we use Total Return bond indexes in our indicators which offer a theoretically superior result. However, one must not rely solely on credit spreads to judge the level of market risk. We present here the risk measures we use in our work and have aided us in our successful investing.