The elusive pursuit of inflation

As we entered 2017, bond investors were fearful that inflation and therefore bond yields had reached an inflection point. Forecasters and investors had heard the promises of fiscal policy easing by the new US administration. In addition, in the wake of a rate hike in December 2016 many anticipated that higher inflation and a change in monetary policy was on the cards for 2017 and beyond. At the start of the year, consensus expectations were for 10 year US Treasury yields to end the year at 2.70% (according to Bloomberg). It has to be said that consensus expectations for bond yields have almost always been higher at the beginning of every year since the millennium! Once again this bearish consensus view on inflation and bond yields was turned on its head, with 10 year Treasury yields trading to a low of 2.04% as recently as 7 September 2017.

Inflation matters for bond investors

The declining trend in bond yields, that started in the 1990’s, is closely correlated with the declining trend in inflation over the same period. Further reinforcing this correlation is the fact that as core inflation gradually began to consolidate (Core Personal Consumption Expenditure has averaged 1.71% since 2000) the bond market too has slowly began to move into a range (where 10 year US Treasury yields have averaged 2.17% since 2012).

US bond yields and inflation

Source: Bloomberg, September 2017. * US personal consumption expenditure index.

Phillips Curve

The most recent Federal Reserve Open Markets Committee minutes (from the 20th September meeting) highlighted the confusion and complexity around forecasting inflation from several committee members.

We explored some of the reasons behind this at our recent Strategic Forum, where we analysed one of the pillars of modern US Federal Reserve (Fed) policy setting, the Phillips Curve. Our analysis determined that the traditional Phillips Curve relationship (a downward sloping curve illustrating that as unemployment falls wages/inflation rises) has steadily flattened since the mid 1990’s. In addition, we found that this remains the case even if we widened the definition of unemployment to include those employed part time but wanting full time work. Our analysis went one step further to determine that if we step beyond the crisis years, which is after 2008, the relationship is no longer relevant.

Phillips Curve 1907-2017 - Phillips Curve post 2009

Source: Macquarie, US Federal Reserve.

As our research has shown in recent years, deflationary forces are being unleashed from demographics, digitalization, too much debt and a dependency on low rates. These forces when combined with the dismantling of relationships such as the Phillips Curve are the starting point to understand that the long term history of inflation is no longer that relevant to explaining the outlook for inflation. We will explore our views on the main drivers of inflation in a soon to be published series of research papers.

Investment implications

In our view, bond markets are embracing this new thought process even if central banks seem less convinced. This can be observed by the steady flattening of the US yield curve through this Fed tightening cycle, where 10 year yields are little changed while short term yields are gradually moving higher.

The implications for our investment strategies going forward are that we do not fear inflation in the next 12 months, rather see it as a scenario risk, albeit a cyclical one not a structural one. This means that while we will continue monitoring inflation very closely, our duration strategies are moving from underweight target/benchmark to neutral, as yields have pushed higher. Looking ahead, our current plan is to further increase duration (to overweight) if yields move higher.

Strategically, modest inflation risk does not deter us from seeking a balance in multi sector strategies, where duration can provide an offset to overweight credit positioning where valuations are getting tight especially in high beta sectors such as high yield credit.