Historically junk bonds, also known as high yield bonds, correlate quite well with equities. In fact we have built timing systems that use junk bonds as a trend confirmation indicator that have done quite well from a historical perspective. Because of all this we, as well as half the internet, love that the junk bond ETF’s HYG and JNK exist. They are both great trading vehicles and a fantastic aid in our intermarket analysis and to measure investor risk sentiment in both the equity and bond markets. All that said there are two major ways in in which we see them misused all the time.
First lets look at charts of the SP500 and HYG. As you can see in the chart below the SP500 as represented by SPY has been trending higher despite the almost 10% pullback in October. In fact it is at new all time highs right now.
Next we have a chart of HYG. As you can see it has been diverging from the SP500 since the end of June and has been sloppily trending lower ever since.
Many analysts and traders have been posting these charts and saying that since there is a growing divergence between junk bonds and equities that we have to see a move lower in equities to bring things back into line. There are potentially a lot of things wrong with this statement but the one we want to look at right now is the idea that they are diverging. If virtually ALL the stocks in the SP500 were moving up and virtually ALL the bonds in HYG were moving down then we would indeed be worried about a major divergence. Rarely does investor sentiment go all in on equities and all out on junk bonds or vice versa.
So what is the problem? We don’t see a real divergence. The secret is that not all of the stocks or all of the bonds are diverging. If we look at an industry group breakdown we see something very important, namely that almost 15% of it is made up of oil and gas companies. Do you think it might be important to know that the highest industry weighting is made up of a sector of the market that has been gored by the bears in the woods?
In fact if we look at a performance chart of the above groups we can see that there is not a real divergence. Over the past six months stocks have gone higher, junk bonds are basically flat, and oil and gas stocks have been absolutely hammered. Consequently in our view this is NOT a bearish divergence as most junk bonds are not dropping in the face of a rising stock market but instead only the oil and gas junk bonds are dropping.
While the above can be said to be up for debate we think the case is fairly solid that this so called divergence is less indicative than most seem to think. The next issue with how people use HYG and JNK is far less debatable. Have you seen any of these charts showing the ratio between HYG and TLT? Most people use it to gauge investor sentiment. Are investors running to safety in Treasuries or are they going to the opposite end of the fixed income risk spectrum and buying junk bonds? We think the concept is completely valid but the implementation is less so.
Here is a chart of the ratio of HYG and TLT. As you can see it has been trending downward for all of 2014 as long term Treasuries have vastly outperformed their junk bond cousins. The problem is that usually you want to compare first or second cousins and not tenth cousins.
If you just went WTF is this guy talking about we don’t blame you, so let me explain. In the bond world there is a thing called maturity which is simply how long until the bond is paid back. TLT is a 20+ year bond ETF which means that it is full of bonds that end in 20+ years. In fact 90% of TLT is comprised of 25+ year Treasuries. Where the HYG/TLT ratio breaks down is when we then look at the HYG or JNK maturity profiles. As you can see in the table below 95% of HYG is comprised of 1-10 year bonds with less than 3% being more than 10 years.
HYG Maturity Profile
The story is much the same with JNK as almost 94% of the bonds in JNK have a maturity of 3-10 years. And just like HYG, less than 3% is made up of bonds with maturities over 10 years.
JNK Maturity Profile
So while the idea of a junk bond/Treasury bond ratio is great, using TLT as the Treasury component is less than ideal as the maturity mismatch is so large. Lucky for us there are several other Treasury ETF’s and they are made up with ranges of maturity. TLT is 20+ years but did you know that IEF is a 7-10 year ETF and IEI is a 3-7 year ETF. Either of these would be a far better choice for this indicator than TLT as we are able to compare first or second cousins instead of tenth cousins.
So what are these more accurately matched up ratios saying now? HYG/IEF shows that fixed income investors were happy to own junk bonds for the first half of the year and since then junk bonds have drastically under-performed. Of course since the oil and gas sectors took their dive at the same time it is up to you to decide if the under-performance of junk bonds is as bearish as it usually it. We are discounting it but are still wary.
Looking at the HYG/IEI ratio things are even less bearish as the downtrend has been both shorter and less damaging. Like above we think that the under-performance in junk bonds has largely been in the oil patch with other sectors barely affected. Consequently we are wary and yet modestly bullish on risk.
ETF’s, indices, securities, and other indicators are all well and good but it is important to look under the hood as much as possible so that you are drawing the right conclusions. In this case you can decide if you should discount the under-performance of oil and gas and the subsequent divergence between junk bonds and stocks but we think that anyone that looks at the Junk/Treasury ratio needs to be using a proper maturity comparison or they are doing it wrong.