I recently wrote an article for Financial Planning magazine in which I looked at ways of judging income investments. One of the major take-aways from the analysis–and a conclusion that surprised many readers–is that long-term government bonds now look as risky as junk bonds.
This is not a typo.
Long-term government bonds currently have really low yields because investors are scared of ‘risky’ assets like stocks, so they have flocked to government bonds (a ‘safe’ asset) and long-term bonds, and that’s driven their prices up and their yields down. Long-term bonds, while understood to be riskier than shorter-term bonds, provide more yield.
High-yield (aka junk) bonds are the bonds issued by companies with low credit ratings–the bonds are considered to be below ‘investment grade.’ It is expected that there will be a meaningful default rate on these bonds by the issuers. Junk bonds are considered the riskiest class of bonds.
Long-term government bonds have effectively zero risk of default, but have substantial exposure to interest rate risk. If interest rates rise, the prices of long-term bonds will drop a lot more than the prices of short-term bonds.
The sources of risk may be different, but the total risk is a key measure. I think that investors are signaling that they are really concerned about the default risk, but not about the interest rate risk—which explains why investors are willing to hold long-term government bonds when they have such a poor yield.
How can we get a handle on the true risk of these asset classes–as opposed to investor perception?
The best measure that I know of is the implied volatility on bond ETFs. TLT is a long-term government bond index ETF. How risky is it? There are options that trade on ETFs and we can use the prices of the options to figure out how risky the ETF is. This is a standard process in any quantitative finance class. Even Morningstar provides implied volatility on its website. I am not going to go into detail, but suffice it to say that implied volatility is a standard measure of risk.
The simplest way to get implied volatility data is from iVolatility.com (and its free). The implied volatility on the options on TLT expiring in March 2011 is 16%. The implied volatility on options on HYG, a high-yield bond index fund, expiring in March 2011 is 10.3%. This means that the long-term government bonds are substantially riskier (i.e. more than 50% riskier) than the high-yield corporate bonds.
Now, perhaps the options market has it wrong. This is possible, but there is considerable evidence that the options market is a pretty ‘smart’ way to get risk measures.
The options market is suggesting that the interest rate risk in long-term government bonds is huge–big enough so that the aggregate risk on long-term government bonds is higher than the aggregate risk of junk bonds.
It is worth getting acquainted with implied volatility as a measure of risk because this measure provides a quantitative measure of risk that may contradict our innate inclinations. Inclinations like presuming government bonds are ‘safe’ and junk bonds are ‘risky,’ for example.