Investing by Duration
By: Heather Rupp, CFA, Director of Research for Peritus Asset Management, the sub-advisory firm of the AdvisorShares Peritus High Yield ETF (HYLD)
It was hard to ignore the call in the fixed income space for “short duration” investing over the last couple years. Duration is a measure of interest rate sensitivity (the percentage change in the price of a bond for a 100 basis point move in rates), so the lower the duration the theoretically less sensitive those bonds are to interest rate movements. Lower duration bonds would not eliminate the interest rate impact, just lessen it. We see this as a good strategy broadly speaking if you are talking the high yield asset class versus the investment grade asset class, with the high yield market naturally having a much lower duration due to its higher starting yields and generally shorter maturities. However, we believe this strategy is lacking when it is used to parse out the high yield space itself, investing in only the lower duration names within the high yield category, irrespective of other considerations.
This gets back to the concept of yield. In a box, this sounds like a good strategy, but you need to factor in the starting yield on the portfolio to mathematically assess if practically speaking this is the right strategy. If you were to invest according to a “short duration” strategy in the high yield debt market, let’s hypothetically say you could achieve a portfolio with a duration of 2.0 years, so a 100 bps change in rates over 6mos would mean that the price of your portfolio would theoretically decline by 2.0%. If your starting current yield on the portfolio was 6.5%, meaning you theoretically generate 3.25% of income over that 6mos, then you are looking at a theoretical net gain of 1.25% (3.25% – 2.0%) over the period of rising rates. However, if you can build a portfolio in the high yield bond and loan market investing according to both maximizing yield and considering duration, let’s say you can build a portfolio with a duration of 2.5 years and a current yield of around 9%. In this case, your theoretical sensitivity to a 100bps movement over 6mos would be a price change of 2.5%, but you would be theoretically generating 4.5% of income over the 6mos, so your net theoretical gain would be 2.0%. If that 100bps interest rate movement is over a year instead of 6 months, that yield benefit gets even larger, putting you at a theoretical net gain of 4.5% for the hypothetical short duration portfolio versus a theoretical gain of 6.5% for the higher yielding portfolio.1
But you also most consider what if rates don’t rise, then what? It seemed the short duration investing was the big trend heading into 2013 as virtually everyone thought rates were going to rise, and the trend hasn’t abated as the year has progressed. However, that rate increase hasn’t played out, as rates on the 10-year Treasury ended 2013 at 3.04%, and now sit around 2.2% today. As noted above, duration is a measure of price change based on a change in interest rates, and that price move works in both directions: a theoretical price decline if rates rise as we have profiled above, but also a theoretical price increase if rates decline. So in an environment such as we have seen so far this year, with rates declining, then the higher yielding portfolio would not only benefit from the higher starting yield but a theoretical positive price movement per the duration calculation.1
So we see this as compelling evidence that investing purely according to a short duration strategy and not factoring in yield is not necessarily the wisest way to approach this market and the seemingly ever present interest rate concerns. At the end of the day, yield matters. A higher yield can go a long way in making up for relatively small differences in duration. Furthermore, even if rates do rise, it very well can take longer than many expect (many were certainly wrong on how 2013 would play out!), making the argument for the higher yielding portfolio versus the purely short duration portfolio even stronger.
1 The duration and price movement relationships are approximates and calculations are provided for illustration only. These calculations assume that credit spreads remain constant and do not factor in any fees or expenses or changes in price movements for other reasons, including security fundamentals, etc. Actual results may be materially different.