In our previous RedBlog post, my colleague Dan discussed how to incorporate expected rate rises into the liability hedging decision. In this piece I offer an alternative perspective on yields and liability hedging.
A Little Background
Over the last few years, many investors have delayed hedging as they believed that benchmark yields would increase, meaning that hedging would be cheaper later. These increases were expected as a result of the end of quantitative easing and central banks raising their historically low policy rates.
It has been a long wait.
Today, unemployment is close to its pre-crisis average, wages are rising, and as the oil price shock falls out of the annual CPI calculation, inflation is expected to rise sharply towards the end of the year. While we may not be in a “normal” recovery, these are conditions in which a central bank might tighten policy.
Does the market always get it right?
The market’s expectation of the timing of the first rate rise has varied greatly over the post-crisis period.
In early 2011 the first rate hike seemed imminent. By June 2012 it was three and half years away. Based on current pricing the latter estimate seems sound. Throughout 2015, markets have been expecting the first rate rise to occur sometime in the first half of 2016. As the market gains confidence that the first rate rise is now on the way, I thought it would be interesting to see how the yield curve has behaved in previous tightening cycles.
Limitations to the analysis
This analysis is constrained for two reasons.
Firstly, full yield curve data does not exist for all previous cycles. Meaningful data is only available since the mid 90s, which has constrained the analysis to the last 20 years.
Secondly, defining the bounds of a tightening cycle requires some subjectivity. In August 2005 rates fell 0.25% (25 basis points) after having risen for a couple of years. Should the rate increases on either side be considered separate tightening cycles? Between September 1999 and February 2000, the Bank of England’s policy rate (the Bank targeted the Repo Rate at the time) rose from 5.0% to 6.0%. Is a 1.0% increase over 5 months enough to be considered a tightening cycle?
In the case of the former, I have treated this period as one tightening cycle. Given the lack of available data I have included the latter. (Of course some would argue that we’ve been in one long loosening cycle since the early 80s, but I’ll leave that topic for another day.)
In total, there are three periods in the data set:
- October 1996 – October 1998
- September 1999 – February 2000
- November 2003 – December 2007
What actually happens after the first rate hike?
So how did the yield curve look in October 1996? The first chart shows the spot curve, the expected spot curve one year forward (black), and the spot curve that actually transpired a year later (red).
In the second chart we have the expected spot curve two years forward (black) and the spot curve that transpired at that time (red). The third chart shows the same, but on a three year time scale.
For September 1999 we have the Spot, 1y Forward and the Future Spot that transpired one year later (I’ve only included one year as it was a short cycle):
And finally, November 2003:
There’s a clear pattern.
In each chart the future spot is below the market-implied forward curve. As I explained in a previous blog, this implies that hedging would have improved funding levels, and that delaying would have been more expensive.
The purpose of this post is not to criticise investors who delayed hedging in expectation of higher yields. That is a trite theme. Moreover, hedging decisions are among the most important decisions that investors make and the charts above on their own are insufficient to justify action.
A more robust conclusion that one could draw is that investors needn’t fear regret risk. While current levels may feel expensive, hedging is unlikely to be significantly cheaper in the coming years, even when rates are rising.
Related articles you may find interesting: