NN IP: US economy characterized by better growth but low inflation

Meanwhile, continued labour market tightening, rising net wealth and increasing consumer confidence are supporting household spending even though consumption faced a temporary headwind in the form of an inflation hump early in the year. The result is a pretty balanced and sustainable feedback loop between rising domestic spending and income, which in the case of the US has been supported by a substantial easing of financial conditions in the past few months. The big question in the case of the US is how much longer the rise in demand can go on before it runs into supply-side constraints. Once these constraints start to bite, firms could be faced with a rapid rise in unit labour cost growth. In all likelihood they will try to pass this on in their selling prices, which will lead to accelerated Fed tightening and a substantial tightening of financial conditions, which could slow down growth abruptly. If firms cannot pass rising unit labour cost growth on to selling prices, their profit margins will shrink and they will slash capex and labour demand, which then also leads to an abrupt slowing in demand. This is the essential reason why the Fed aims to slow down demand growth to a level that is around long term potential in the vicinity of full employment.

Still that is easier said in theory than done in practice. First of all, as we argued on many occasions, potential output or its labour market equivalent, the NAIRU, is an unobservable variable which moves over time and aggregate demand could well push aggregate supply onto a permanently lower or higher growth path. On top of this, the sensitivity of wage and price inflation with respect to the degree of slack or overheating could also change over time and may even vary with the degree of slack itself. Also, both the relationship between wage and price inflation, on the one hand, and slack on the other, can be pushed higher or lower by a permanent rise or fall in inflation expectations which themselves are unobservable. The upshot is that there are many combinations possible between developments in the real and nominal sides of the economy which make the life of a central banker complicated.

The big question is thus to what extent there is room on the US supply side to accommodate robust growth, especially if it were to be joined by fiscal stimulus. This depends on the future evolution of labour supply and productivity growth. News from the labour market has been a bit confusing over the past few months. The underlying trend in employment growth is slowing down to around 120K but even this trend is a volatile beast so it is difficult to draw any firm conclusions. At the same time the participation rate declined over the past three months and this has helped push the unemployment rate to a new low of 4.3%. At the same time the U6 unemployment rate has fallen fast to 8.4%, which is around the 2006-2007 average. Circumstantial evidence such as the JOLTS data, initial claims and consumer assessment of the labour market also continue to point to strength. The only safe conclusion we can draw from this is that the labour market is tighter than a year or even six months ago but we do not know how tight. The uncertainty about how tight the labour market is compounded by the behaviour of wages. Wage growth is better than in 2013-2014 but only moderately so.

Despite all this, it seems that the room for labour supply growth to give an additional persistent boost to US potential growth is relatively limited. This then puts most of the burden on productivity growth. The good news here is that productivity growth has increased to 1.2% yoy in Q2, which is twice the average rate in the preceding five years. The big question is to what extent this can last and improve. The honest answer is that no one knows, but we see some room for improvement due to two developments. First of all, capex is an important driver of productivity growth. The economy has been characterized by a capex drought until recently and to the extent that this morphs into a sustained capex improvement we should see lasting improvements in productivity growth. Secondly, the quality of capex matters as well, in particular in the vicinity of full employment producers are likely to step up investment in labour saving technologies which could give an extra boost to productivity growth. After all, such labour saving technologies allow firms to pay higher wages which result from the disappearance of labour market slack while protecting their profit margins.

All in all, there seems to be room for continued US GDP growth at a rate between 2-2.5% for the foreseeable future on the real side. This then leaves the nominal or inflation side of the equation. We already alluded to the fact that wage growth if anything is disappointing on the weak side even though that could be partly due to low productivity growth. As for price inflation it seemed that core PCE inflation was on a moderate upward trend until three months ago. Since then we have had a string of three consecutive disappointments which are likely to push core PCE inflation from 1.8% towards 1.4% yoy. A string of three surprises could still be due to idiosyncratic factors and there is certainly some of that at play in the form of a new methodology for quality adjustments in certain tech goods. Nevertheless, the suspicion is growing that there may be secular trends behind the decline in core inflation as well. The uptick was always narrowly based on two items: healthcare and housing. The first one is pretty much a regulatory issue while the second one is related to the vacancy rate which fell a lot between 2010 and early 2016 but is now starting to rise on the back of the improvement in residential investment. Outside these categories service sector inflation remained pretty flattish, which rhymes well with our former story about some improvement in productivity growth as well as wage growth. Our base case would still be that a tightening labour market should push up unit labour cost growth further which will be translated into higher price inflation. Nevertheless, the downside risks to this base case are increasing. Perhaps the equilibrium unemployment rate is a lot lower. Alternatively, the US economy could have been subject to a fall of inflation expectations to a level that is somewhat below 2%. Both market-based and survey-based expectations still point to this possibility. To the extent that this is the case the Fed will have overheat the economy to re-anchor expectations. Or maybe we are just dealing with a temporary downward blip after all. Only time will tell!

The Fed continues its gradual march towards neutral, but what happens after that?

When Yellen and her colleagues went into the June meeting the main pieces of information on the table were thus:

• Growth is likely to remain at a somewhat 2% plus rate while the labour market could be displaying a soft landing towards rates of employment growth closer to the rate required to keep the unemployment rate stable. The latter combined with circumstantial evidence still suggests the labour market is tightening.
• Labour market tightening increases the Fed’s confidence that inflation will rise towards the target but actual wage and price inflation readings have surprised on the downside. Importantly this highlights the risk that inflation expectations have fallen somewhat below target but it could just as well be due to other factors.
• Financial conditions (FC) have eased substantially and are around their easiest levels in two years. This is important information because FC are a vital part of the transmission mechanism. To put it bluntly the Fed wants FC to tighten in response to a rise in the policy rate and the opposite happened. If easy FC persist it could be a reason to accelerate rate hikes, all else equal.

Besides these complicated cross-currents in nominal GDP and financial space the Fed is also getting ready for cross-currents in instrument space even though the FOMC will do its utmost to limit the latter. The issue in instrument space is of course that balance sheet roll-off will have to start at some point. There are many estimates of the effect of the Fed’s USD 4.5 trillion balance sheet on the term premium but they tend to centre on a compression of the term premium on 10y yields by 50-100bp. If one believes, as we do, the ECB and BoJ QE had a significant effect on the US term premium the estimate would be at the lower end of this range. The most important effect of shrinking the balance sheet would thus to put upward pressure on the term premium. In addition to this, the market could in theory also interpret balance sheet roll-off as a change in the Fed’s reaction function in which case the expected path of future policy rates could move upward. This is essentially what happened in 2013 during the Taper Tantrum. Back then the Fed was still firmly holding on to the asymmetric reaction function – “Display a vastly underwhelming response to improvements in the data while providing a put option against downside risks”. Taper talk was seen as a signal that the put option was becoming less strong and that the Fed might move at a faster face towards a normal pre-crisis reaction function which displays a symmetric and positive response to trend changes in the data. As a result, risk premiums including the term premium surged and the expected policy rate path moved higher.

With all this in mind we can turn to the outcome of the June FOMC meeting. The most important message is that for now the Fed holds on to its game plan of a gradual but steady march towards normal pastures. This march applies both the policy rate but also to the Fed’s reaction function which moves away from the asymmetric reaction function described earlier towards a symmetric pre-crisis version. An important feature of the latter is that the sensitivity of Fed policy to downside surprises in the data and/or a decline in risk appetite in financial markets will be a lot less than it was over the past few years. This is the essential reason why the FOMC chose to downplay the softness in core inflation over the past three months. A year or so ago this would have been reason enough to change the outlook for policy but at this point the Fed is prepare to look through this for now. The FOMC believes that soft core inflation is mostly caused by idiosyncratic factors and continues to expect inflation to converge to the 2% target over the policy horizon. This was also very clear from the updated forecasts where 2017 core PCE inflation was downgraded from 1.9% to 1.7% while the 2018 and 2019 forecast remained at 2.0%. In support of this hypothesis Yellen mentioned the rebound in Q2 growth, continued labour market tightening and the easing of financial conditions. The FOMC may well be right in making the assumption that inflation weakness will prove transitory but as we argued before there is a considerable tail risk that the weakness is driven by something more fundamental (below target inflation expectations and/or a very flat Phillips curve) in which case the Fed will have to adjust its projected rate path downwards. At any rate the Fed nudged its forecast of the NAIRU down again by 0.1pp to 4.6% which about 1 pp lower than the central tendency back in 2012-13. If the weakness in wage growth persists we would not be surprised to see further future downgrades. For now the Fed expects the unemployment rate to stabilize at 4.2% which validates its forecast of a gradual rise in underlying inflation.

The changes to the FOMC statement were relatively minor. The Fed noted that employment growth had cooled down somewhat but that the labour market remains solid overall. Also, the improvement in consumer spending alongside a continued expansion of capex was noted which suggests somewhat more confidence the robustness of domestic demand growth than back in March. This increased confidence was further underlined by the omission of the sentence that “global economic and financial developments” warranted close monitoring. This makes sense since global growth is on a firmer and more stable footing while the dollar, EM capital flows and global risk appetite remain well behaved in the face of Fed tightening. As a result of all this, the risks to growth remain “roughly balanced”.

In addition to the statement the Fed released a plan for balance sheet roll-of which confirmed the caps mentioned in the minutes of the May meeting. The Fed will set initial run off caps of USD 6 billion per month for Treasuries and USD 4 billion per month for MBS. These caps will be increased by USD 6 billion and USD 4 billion respectively every quarter until the cap for UST stands at USD 30 billion and the MBS cap is at USD 20 billion. It is important that the caps are intended to be a maximum reduction in asset holdings and that the actual reduction could be less. The statement also says that the new normal for balance sheet size will be “appreciably below that seen in recent years but larger than before the financial crisis”. During the press conference Yellen explained that balance sheet size is likely to remain an active policy instrument in the future. One reason for this is that r-star (the equilibrium real policy rate) may well remain below pre-crisis levels for some time to come because of which a recession could push the policy rate back towards zero. In that case altering size and composition of the balance sheet will come into play again.

The final question is what all this means for our Fed outlook. First of all, we have changed our base case for balance sheet roll-off to start in September because Yellen said it could start “relatively soon”. After that we expect this process to remain on auto-pilot unless there is a really big change to the outlook. The FOMC indicates that the hurdle for deviating from the balance sheet course is pretty big indeed. It would only happen in case of a “material deterioration in the economic outlook” which triggers “a sizeable reduction in the federal fund rate”. The next rate hike should then follow in December. For next year we have reduced our Fed call from four to three rate hikes. This actually puts our Fed call right in line with the median dots (which did not change in June). The reason is that we now expect somewhat less fiscal tightening which may also have relatively small demand boosting effects. The risks to our view are even somewhat on the downside because of the possibility that something fundamental is weighing on inflation.