Alistair Wilson explains why duration is an important consideration when investing in fixed income
In the last few months since the Fed began its tapering talk, fixed income investors have really found out why the duration of their bond funds is so important. Duration itself is a fairly simplistic topic; it is basically a measure of sensitivity of a bond’s price to a change in interest rates. So, generally the longer the bond, the more price sensitive to interest rate moves it becomes.
Consequently, the opening question that we tend to receive from investors these days is “what is your duration?” Whilst the absolute number is clearly very important, there are additional questions that are equally important but far less frequently asked, such as “what are the main components of
your duration exposure?” or “what is your interest rate duration versus your credit duration”?
Why are these important?
Fund managers will generally have reacted to the changing landscape over the course of the year, and most will tell you that the duration they run is now significantly lower than it used to be. How have they done that? In many cases they have portfolios of bonds consisting of several hundred securities, the aggregate of which will give them the largest component of their duration exposure. In the last few months most would not have merely sold the old portfolios of longer dated bonds and then bought a new portfolio of shorter ones. These days fund managers have a broad range of investment tools at their disposal - such as futures and derivatives - that are far more efficient for managing large portfolio; however it does mean that two portfolios with exactly the same duration can perform very differently.
Portfolio A holds 100 £ bonds all in the five to 10 year maturity sector, which contributes a duration of six. Against this the manager has decided to take out a ‘short’ position in the 10 year gilt future covering half the portfolio. The short will bring down the duration of the portfolio to around 2.5. If we then had an overnight rise in rate expectations across the yield curve of 100bp, the portfolio would be much better protected and the fall in price which would have been 6 per cent becomes just 2.5 per cent.
However, rate moves never occur perfectly along the yield curve, yield curves flatten or steepen. Long ends of yield curves are very inflation sensitive whereas the zero to five year parts are much more a reflection of interest rate expectations. So you could have a situation where the 10 year yields stayed where they were and all the shorter dated bonds were hit, resulting in a bigger drop in price of the bonds and no move on the hedge, so the hit to the portfolio is much larger than the duration number alone would imply.
Portfolio B holds a portfolio of much longer dated bonds that are reflective of what the fund managers view used to be and a duration of 10. Against this the portfolio manager has decided to hedge via an interest rate swap targeted at the five year point of the yield curve. Given it is less sensitive at the front end, he has decided to do this in much bigger size to also bring the duration down to 2.5 so it is the same as portfolio A. Once again, if we had a ‘parallel shift’ upwards in rates by 100bp across the yield curve, the portfolio would be expected to take the same 2.5 per cent hit, but as we know, these parallel shifts never happen, so depending whether the curve steepens or flattens will materially affect how the hedges perform versus the bonds.
Very long dated hedges bring duration down quickly and cheaply, but are also often a long way from where the bonds sit, so these hedges carry a much bigger risk to those that are focused on parts of the curve that are similar to the bonds that are held.
Now that everyone seems comfortable with the concept of duration, it’s time to get into the detail!
Written by Alistair Wilson, head of institutional business at TwentyFour Asset Management