Tuesday, June 2, 2015

Phase II of the global bond bear market has started

Following the sharp rise in yields between mid-April to mid-May, 10y Bund yields have fallen moderately back to around 50bp. This movement lower was probably caused by the previous technically oversold condition for bond prices coupled with the ongoing uncertainty relating to Greece, weakish US data as well as doubts with respect to the inflation outlook for the Eurozone as evidenced by the sharp drop in Eurozone inflation expectations over recent weeks.  However, this should only be regarded as a consolidation/correction within a multi-year bond bear market which started this spring (and at the end of January in the case of US Treasuries and Gilts). With signs of an improving US fundamental backdrop becoming stronger, Eurozone inflation surprising to the upside and some movement in the Greek negotiations, global bond yields have just started on their next leg higher.
First, US growth should improve again. US economic data has surprised negatively in recent months due to several factors. The harsh winter weather has held back the economy temporarily while the US West Coast port strike distorted the supply chain. Both factors, however, are of temporary nature only and should have started to reverse already. In addition, given that the US economy is a very large oil & gas producer and given that the fracking technology allows for an almost instantaneous adaption to changing market circumstances, the negative effects of lower oil prices on the oil & gas sector have for the past few months exerted a downward pressure on growth. The growth positive effects of lower oil prices – via increased purchasing power for households and reduced price pressures for energy-consuming corporates – usually take much longer to develop and should only start to become visible. Finally, seasonal adjustments continue to play havoc with US economic data releases. For example the San Francisco Fed estimated in a recent working paper that the residual seasonality of the US economy during Q1 amounts to 1.6% this year (i.e. growth in Q1 has been underestimated by this amount). The chart below highlights that this problem has grown larger over recent decades. 
Average quarterly US GDP growth 
Source: San Francisco Federal Reserve
This problem of seasonal adjustments is also clearly visible in the US ISM index which surprised positively yesterday. The chart below compares the seasonal adjustment factors applied to the raw data in 2007 with those applied this year. As can be seen, April saw a significant upward revision of seasonal adjustment factors, resulting in lower levels of the adjusted data. However, from May onwards this reverses and in June (as well as in September) the seasonal factors are below those applied in 2007. This is all the more remarkable as seasonality should have dropped in the US economy over the past 8 years and not grown. The highly seasonal construction sector as well as the manufacturing sector have become smaller relative to GDP while the less seasonal service sectors such as education and healthcare have grown in importance. Be it as is, the May ISM report out yesterday which showed an improvement in the index of 1.8 points now paints a more reliable picture of the underlying state of affairs than in April. Importantly, the new orders component – seen as a leading indicator – has risen for the second month running and now stands at a healthy 55.8. 
US ISM seasonal adjustment factors 2015 vs. 2007
Source: DoC  
As a result, the outlook for the US economy remains favourable. The monetary transmission mechanism has been restored, the housing as well as the labour market have been improving and the growth negative effects of the previous fiscal tightening have been absorbed while the level of pent-up demand remains substantial. As such I remain convinced that the US economy has embarked on a self-sustaining upswing and that US economic data will improve substantially in the weeks and months ahead. Therefore, the FOMC remains on course to start hiking rates in September.
Second, Eurozone inflation continues to increase faster than anticipated by the markets as well as by the ECB. I have already previously highlighted that the underlying inflation pressures have been increasing already since May last year. The chart below separates the developments in Eurozone inflation into three components. Commodity price effects are marked in green (defined as the difference between headline and core inflation), fiscal policy effects in red (derived from changes in consumption taxes and prices for administered goods and services) and the residual in blue. This residual can be thought of as the underlying price pressures emanating from the private sector. As can be seen, disinflation was mainly caused by falling commodity prices. Also a reduction in the price effects of fiscal policy from 0.7% to only 0.1% in March was an important driver. Underlying price pressures fell as well, from 1,0% at the end of 2011 to 0.1% in May 2014 with likely the strong Euro (reaching its high in March last year) being responsible for approx. half of this drop. However, these underlying price pressures have increased steadily since then. That core inflation stayed at 0.6% until April was almost exclusively due to the reduction in the price effects of changes in consumption taxes. As these inflation reducing effects have come to an end, core inflation has now also started to increase as evidenced today with the preliminary release of +0.95 for May.
Eurozone HICP and its components
Source: Eurostat, ResearchAhead
Inflation should continue to rise into 2016, be it on the core as well as the headline measure. Moreover, the Eurozone growth outlook continues to improve and growth should rise further over the course of this year, likely hitting approx. 2.5% on an annualised basis before year-end. While structural challenges remain, cyclical forces are adding up to a very strong growth tailwind. The weak Euro, low oil prices, reduced negative growth effects from fiscal tightening, record low nominal and real yields – at last also in the periphery – as well as a turnaround in credit creation all act to lift growth. Furthermore, amid the time lags involved, the positive impetus will get ever stronger in the months and quarters ahead.
For bond markets, the structural bond bull market of recent years has ended. Over the long-term, developments in sovereign yields and nominal GDP growth are closely linked. This can be seen in the chart below showing the history of US nominal GDP growth and 10y UST yields for the past 55 years. The long-term bond bull-market since the early 80s was fuelled by several components: a drop in long-run trend growth, a structural reduction in inflation rates as well as an evaporation of term premia embedded in longer-term bonds with the latter mostly at play in recent years due to the central banks’ bond buying programmes.  
Long-run 10y UST yields vs. US nominal GDP growth
 Source: St. Louis Federal Reserve
Also in the Eurozone, nominal growth and 10y Bund yields are moving in sync over the longer run. However, while the fall in bond yields until last year was mirrored by fundamental developments (lower inflation and lower growth), the discrepancy between the level and direction of yields – with 10y Bunds falling to almost 0% - and the level as well as direction of growth – with growth having improved to 2% in 2014 from the record low 0,8% in 2013 – has become substantial. Also here this discrepancy should largely stem from the speculation on and subsequent decision by the ECB to engage in a large scale asset purchase programme. 
10y Bund yields vs. Eurozone nominal GDP growth
Source: Bloomberg
As a result, we are in a state where cyclical nominal growth improves – due to both higher real growth as well as higher inflation rates – while risk premia are extremely low due to the central bank purchases. This results in an extremely challenging environment for bond markets. As nominal growth improves, the support provided by central bank for bond markets decreases over the medium term. The FOMC should be the first to hike rates, however, also the ECB’s QE days are numbered. While purchases will last until September 2016, they should be reduced and potentially even terminated before the end of next year. Moreover, the ECB will likely increase the depo rate to 0% during 2016 as it ends the emergency state of negative policy rates.  As a result, we have entered a new cyclical bond bear market – due to cyclically higher growth and inflation. Moreover, while trend growth rates should stay depressed and long-run inflation should not deviate much from the central banks’ stated targets of around 2%, the reduced central bank support should lead to a repricing of term premia over the next years. 
10y UST yield takes another attempt at breaking above their long-run downward trend
Source: Bloomberg