US: Inflation Risks are Rising

[fxstarr]

Despite high commodity price inflation, tight labour markets and rising import prices there has, so far, been little evidence of inflationary pressures feeding through to goods prices and into the core consumer price level in the US. However, in this report we argue that there are clear signs that significant inflationary pressures have been accumulating in the inflation pipeline. These pressures are now on the verge of popping the surface implying rising goods price inflation in the coming quarters.
Firstly, the anchor for core goods prices, unit labour costs, has been slowly climbing as a response to the tightening of the labour market. This slow - but steady - upward trend in labour costs is the real long-term danger for the US economy, and we think it is underestimated in the market at present.
This is partly because the expansion in the 90s left the impression that inflation was and is restrained by globalisation. We show that the low inflation in the later part of the 90s cycle was actually mostly a cyclical phenomenon attributable to the deflationary impact from the Asian crisis in 1997-8.
Secondly, the rise in commodity prices and weakening of the dollar during the past 12-24 months is about to feed through to higher US goods price inflation. This follows the usual lag of about 12 months.
Consequently, we expect core PPI inflation to accelerate above 3.0%. In combination with higher import price inflation this will add 0.3-0.4%-point to core CPI/PCE inflation. However, the overall path for core inflation will be flat - not rising - as the higher goods price inflation will be countered by a normalisation in housing-related inflation.
Hence, core CPI inflation will go sideways continuing in the higher end of the implicit Fed comfort zone during the coming year. However, as housing services make up a smaller share of the PCE price index, core PCE inflation is likely to drift slightly upward during the coming quarters.
Goods price inflation is less dead than it seems
High core inflation, but not yet driven by goods
The pick-up in US core inflation (consumer prices index ex. food and energy) in 2006 was largely a result of higher imputed housing inflation, which boosted the rent-of-shelter component - a development we forecasted and explained in the report Research USA: Core inflation to rebound, January 2006.

Looking forward housing-related inflation is likely to recede somewhat, cf Global Scenarios, June 2007. Indeed, Fed Chairman Ben Bernanke has said that this is part of the reason why he expects US inflation to fall back this year.

So far there has been little evidence of inflationary trouble in goods prices in the US. This is despite high commodity price inflation, tight labour markets and rising import price inflation.



Overall this seems to lead to a simple comforting analysis: The rise in US inflation was entirely due to a rise in housing-related inflation. This was again partly of a technical nature, partly due to a strong housing market. With the housing market tanking, this inflation component should subside, revealing a benign fundamental inflation picture for goods and non-housing services.

However, there is much more trouble brewing, especially for goods prices. Indeed we forecast a substantial rise in goods price inflation over the coming 12 months - enough to offset the drop in housing-related inflation and to keep the Fed worried.

Below we argue that there is upside pressure in the goods price pipeline. The implication will be a pick-up in core PPI in US goods price inflation from 1.5% to above 3.0% y/y. This will imply an upward impact on overall annual core PCE/CPI inflation of +0.3-0.4%-point, broadly countering the downward drag from slowing housing-related inflation.

Why goods price inflation will rise now
To understand why we are getting worried about goods price inflation in the US, we take a step back and take a look at how the US inflation process works for goods, cf. the following chart.



US goods inflation is sensitive to both domestic and external price trends. However, in the long run producer prices excl. energy are anchored by US unit labour costs.



This is not surprising, as labour is by far the dominant cost component for US producers.

However, one would expect swings in external prices competition to also affect US goods price inflation - at least in the short run. These external price swings would come through exchange rates, global finished price trends and commodity prices.

Consequently, we would expect divergences between core finished goods PPI and growth in manufacturers unit labour costs (ULC) to be driven by the impact from the effective dollar, commodity prices and global finished goods prices.

[fxstarr]

This is indeed the case. In the chart below we show that the difference between core PPI inflation and ULC inflation is related to swings in core crude PPI. Core crude PPI reflects the combined impact on US producers’ costs from movements in the dollar, commodity prices and external competition in general. As can be seen, core crude PPI inflation tends to lead the difference between Core finished PPI inflation and ULC inflation by 12 months. This reflects that it takes time for swings in input price pressures to work through the price pipeline to finished goods prices.



These simple charts offer a neat framework for understanding US goods price inflation dynamics, at least at the finished core PPI level. Producer goods prices tend to grow with unit labour costs in the long run. However, in the short run this relationship is often blurred by swings in the dollar, commodity prices and the competitive situation.

Actually, these charts give a very interesting interpretation of why inflation never rose in the US in the later part of the 1990s expansion. In our opinion the lack of inflationary pressures in the US at the end of the 1990s - despite very brisk unit labour cost growth - can largely be attributed to the Asian crisis in 1998, cf. the box below. This is important, as the lack of inflation in the late 90s has probably contributed to the rather relaxed attitude among many investors towards US inflation at present.

The deflationary legacy of the Asian crisis
The 1997-8 Asian crisis implied a steep drop in GDP in many Asian countries. At the time there were widespread fears of global contagion. Global industrial indicators reacted, with business confidence falling in Europe and North America during 1998, with a peak of anxiety after the Russian debt crisis and the related failure of Long Term Capital Management in the fall. This led to the Federal Reserve cutting the fed funds rate by an accumulated 75bp.

The Asian crisis was transmitted to the US via a drop in Asian demand and a very swift strengthening of the effective dollar. This hurt US competitiveness and created a situation, in which US industry was in trouble.



However, contrary to the expectations of most observers, this manufacturing weakness was never transmitted to the overall economy. Indeed the ISM manufacturing index signalled a manufacturing recession, while the economy at large continued to grow strongly through 1998.



The charts in the main text illustrate why. While the drop in Asian goods demand and the fast dollar appreciation inflicted serious pain on US producers, the crisis also led to a steep drop in global commodity prices. Moreover, the drop in the dollar was a gift to US consumers. Goods inflation dropped to one of the lowest levels ever. And as wage growth was already brisk due to a tight US labour market, real wage growth was among the strongest for the last 20 years.

Interestingly the positive (terms of trade) effect experienced through this channel more than out-weighed the slightly slowing job creation. The net effect being an acceleration of consumer spending, which kept overall US growth going and even accelerating into 1999 (eventually leading to the final tightening cycle from the Fed).

The interesting point is however, that inflationary domestic pressures in the late 90s in the US in the form of high and rising unit labour cost growth, was muted by a very large external deflationary shock from the Asian crisis, as clearly illustrated in the chart in the main text. This is why inflation never picked up very much in that cycle. However, the lack of inflation in the late 90s also gave birth to the myth of ever lower inflation over the business cycle as a consequence of globalisation. As the chart shows the lack of inflation in the 90s cycle was due to a cyclical - and not permanent structural - deflationary event in Asia. Asia’s inflationary impact today looks much different, it is starting to add to US inflation - rather than subdue it.

Looking forward the chart above gives a scary signal. If the relationships are to hold going forward, the surge in core crude PPI over the past year should start to materialise as an upside risk to core PPI inflation very soon. Actually the wedge between core finished PPI and ULC should reverse from slightly negative to very positive. We would expect this wedge to reach 1-1.5%. As ULC inflation is currently running at 3% y/y, this implies a core finished PPI running at above 4% over the coming 12 months. This will by far be the highest core PPI goods inflation in the US since the early 1990s

[fxstarr]

Implications for consumer price inflation
However, what we really want to know something about is the outlook for goods prices at consumer level. This implies linking the PPI to the CPI or more precisely the core goods component in the CPI.

From an overall perspective there are two main inflation pipelines into core goods CPI; core producer prices and prices of imported consumer goods.

The core PPI should in theory be a reasonable guide for the core goods CPI inflation, as domestically produced consumer goods still make up more than 80% of total goods consumption in the US (remember that PPI includes many goods of which parts are imported). Consulting history, this is partly confirmed cf. the following chart. However, since 1998 the two inflation series have been diverging.



On the surface this could look like a globalisation effect, with inflation in domestically produced goods being higher than in imported goods. However, this theory is probably wrong. US import prices excl. energy have actually been rising faster than US core producer goods prices since 2002. So, mysteriously, US consumer goods inflation has continued to be substantially lower than inflation in both domestically produced consumer goods and imported goods.



Rather, the explanation of the mysteriously low consumer goods price inflation is likely to be methodological in our view. Conceptual changes made to the CPI index in the last half of the 90s related to the Boskin report (i.e. the introduction of quality adjustments etc.) have probably been causing a systematic downward bias in the CPI (see "Research USA: Facts and fiction in US inflation data", January 2005). This broadly explains the permanent 1.5% yearly drift between the core PPI and the core consumer goods CPI index setting after 1998, as we showed in the abovementioned paper.

Taking this into account, it suggests that core goods CPI inflation could be accelerating to a level around 1.5% y/y over the coming 12 months. Translated to overall core CPI inflation, this corresponds to 0.5%-point y/y.



To quantify the dynamics between core PPI and core CPI more properly and to take into account the import price channel, we have estimated a simple regression on core goods CPI inflation.



Importantly, this is a pure forecasting model meaning that it does not include any assumptions. In fact, the model confirms the pictures from above by suggesting that core goods CPI is heading for a rebound to 1% y/y in the coming quarters.

Outlook for core CPI and core PCE
We think that the US will see a substantial pick-up in goods price inflation putting upward pressure on both core PCE and core CPI inflation in the coming quarters. We estimate that this would add 0.3-0.4%-point to the core CPI as well as core PCE inflation.

However, this does not imply a new acceleration in total core CPI inflation. In fact, we expect core CPI inflation to remain broadly flat around the 2.5% level.

The reason is that the rent-of-shelter inflation (i.e. housing service inflation) will ease back from its current 4% annual growth pace towards its long-term equilibrium around 3%. Accounting for 41% of the core CPI this will broadly counter the pick-up in core goods inflation.

That said, the change in the underlying composition of the core inflation is not likely to be an insignificant change. As a large part of the rent-of-shelter inflation (owners equivalent rent) is not directly observable, but imputed by the BLS, many have argued that the rise in core inflation last spring was merely a technical - not a real - issue.

Hence, the change toward more goods price inflation and less housing inflation will smell more of "real" price pressure, even if the aggregate core CPI inflation stays broadly flat near 2.5%.



Another important consequence is that the core PCE index - Federal Reserves preferred core inflation measure - will tend to rise. This is because the PCE holds a substantially smaller weight on housing services than the core CPI. While housing services account for 41% of the core CPI it only makes up 18% of the core PCE index. In contrast, the two indexes have equal weights on goods prices.



Hence, the increase in goods inflation will not, as in the CPI, be countered fully by lower housing service inflation. Therefore core PCE inflation is likely to rise slightly in the coming quarters, cf. the chart above.

This will matter for the Fed. Another upward move in the core PCE - even if ever so slight - will ring alarm bells at the FOMC. The coming rise in core goods price inflation will make inflation look like a (slow) upward trend since the late 90s. Not something a central bank would like to see.