Sterling bond investors have done very well, but it’s time to consider the inflation risks

2nd September 2016 by Chris Iggo

Inflation has been non-existent in the developed world this year but inflation-linked bonds have performed very well in total return terms, and nowhere more so than in the UK. This has been driven mostly by lower real yields, reflecting both the grab for duration and the need to hedge future inflation risks. The returns surely can’t be sustained.

But inflation in the UK should pick up given the 13% drop in the value of the trade weighted exchange rate index over the last year. The most recent survey of manufacturers reveals the impact of sterling’s fall – export orders up sharply but input prices rising at the fastest pace for 5 years. Quantitative easing might pour fuel on the fire. So my view is that investors should continue to seek some inflation protection but avoid taking on too much duration risk – short dated inflation linked bonds provide a partial solution for this. Meanwhile, the pantomime of the Federal Reserve continues with another chorus of ’will they, won’t they’. Audience participation is dwindling, with the market far from pricing in a rate hike in the remaining months of this year. If inflation does come back in the US, things could get very ugly in the bond market.

Missing the target – As of July the UK consumer price index had risen by a mere 0.23% this year. This means that the Bank of England (BoE) is missing its inflation target by a large margin. Indeed, inflation has been undershooting since mid-2013. If you look at the performance of inflation relative to target over two year periods starting when the Bank was made independent by the incoming Labour government in 1997, the most recent period has seen the biggest misses. In the first decade of independence, inflation came in relatively close to, but a little below, the target which meant that the actual price level increased at a steady and relatively predictable rate. Then from 2008 to 2012 inflation overshot the target, mostly because of the impact of the sharp decline in sterling in 2008. The subsequent rally in sterling since 2014 and the global decline in inflation has seen inflation undershoot the target in the most recent period. This is interesting in light of the recent debate in the US about monetary policy and the potential for price level targeting as an alternative to inflation rate targeting. Going all the way back to May 1997 and looking at the actual evolution of the level of the consumer price index versus what the price level would have done if inflation had been steady at a 2% annualised rate, today’s actual price level is not very different to where it theoretically should be. The difference is a mere 1.8%. So long term, inflation targeting has been successful. However, the starting point is very important. If we repeat the same analysis but with March 2009 – the month that quantitative easing (QE) began- as the starting point we see that the actual price level has spent most of the time over the target price level even though, today, the two are less than 1.5% apart. Doing it again with a start date of May 2013, however, gives a very different picture. The actual price level is more than 4% below where it would be if the BoE had successfully targeted inflation at the 2% annualised rate since the middle of 2013.

Sterling drives inflation – This is a rather academic exercise because the BoE does not follow a price-level targeting regime. If it did, and this goes for all major central banks that have a medium term inflation target as the key anchor of monetary policy, there would be a need to try to generate much higher rates of inflation in the short term to bring the actual price level up to the target – especially if the reference starting point for such a policy is in the recent past. The thought process here already highlights a pitfall of price level targeting – what is the appropriate starting point? There is also the obvious question around the ability of central banks to directly impact the inflation rate. Above I show that concurrent to the first phase of QE, inflation in the UK was above target (as was the price level). However, since 2013 both measures have significantly undershot. A very superficial assessment of this would say that it is the impact on the exchange rate that is key and little to do with QE at all. Sterling’s decline in 2008 came before QE got underway (although interest rates were cut sharply) and the more recent but less dramatic fall in sterling occurred before the BoE decided to re-instate QE as an insurance policy against a potential Brexit-related economic shock. A rational conclusion would be that inflation in the UK is going to rise again over the next year because of the exchange rate effect with QE playing a role of locking in a lower level for the pound, as it did in 2009-2011. If the consumer price index level is to reach where it should be by May 2017 then annual inflation needs to step up very quickly to somewhere around 3% by the new year. That seems a bit of a reach at the moment, but the BoE’s trade weighted exchange rate index has fallen by 13% compared to a year ago.

Linkers beat everything – The index-linked bond market certainly thinks UK inflation is set to rise. The 10-year break-even inflation rate – the difference between yields on conventional gilts and index-linked gilts – has risen to 2.6% in recent months, from a low of 2.154% at the end of February. For holders of inflation-linked bonds in the UK, 2016 has been a bumper year even though inflation has effectively been zero. The Bank of America/Merrill Lynch index-linked gilts index has delivered a total return of 29.8% year-to-date, beating an equivalent gilt index by 6.2%. The 1-year to 10-year part of that market has delivered a return of 9.8% and the 15-year and over index has delivered 39%. Yes, that’s right – 39%. At the long end of the market the real yield has fallen from -0.68% to -1.80% since the beginning of the year but break-evens have generally fallen until recently. Almost all the return has come from price (coupons are pretty skinny on linkers and accrued inflation has been low this year) with that price being driven by the decline in yields.  The demand for long dated assets in the UK in response to the BoE re-starting QE together with concerns about inflation picking up have delivered these astonishing returns. They are unlikely to be repeated.

Did Carney panic? –  Inflation will pick up in the near term in the UK because of the exchange rate effect. The August Markit/CIPS manufacturing survey not only showed an unexpected bounce in the headline balance for the manufacturing sector to a reading of 53.3 from 48.3 in July, but also indicated a sharp increase in input prices related to the weakness of the sterling. According to the survey, input prices rose at their fastest rate in 5 years with 44% of firms reporting an increase in purchasing costs and output prices. For bond investors sitting on outsized returns in 2016 it is time to think about the risks. For most sectors in the fixed income market the returns have mostly been driven by both lower rates and tighter credit spreads but it is the rates part that is at the most extreme in terms of valuation, certainly in the investment grade world where the decline in underlying rates has been larger than the decline in credit spreads. Of course, any shift in the level of rates depends a lot on what central banks do and say but there is a risk that the BoE got it wrong – the UK economy is not falling apart and inflation is going to jump. Both real and nominal rates could bounce back quickly leading to negative returns in anything that has longer duration. All sterling fixed income assets have performed extremely strongly because of the fall in underlying rates and it is clear that there is a significant risk of reversal at some point. Having said that, I still think inflation protection is necessary in a fixed income or multi-asset portfolio, thus having exposure to break-even spreads without the exposure to duration is an interesting strategy for the next few months.