Bernanke's Short-sighted Rises

• We are becoming increasingly concerned about the US economic outlook for 2007. Fears held by the Federal Reserve about near-term inflation and credibility is easy to understand, and our research confirms that inflation is likely to stay uncomfortably high for a number of months.
• But we fear the Fed might be placing too much focus on the near-term inflation outlook and under-emphasising the headwinds that are now gathering in the US in late 2006 and into 2007.
• Tightening into a weaker economy has historically been unwise. Our GDP growth forecasts continue to be revised down: we expect US growth of 2.75% in 2007 and are increasingly worried about downside risks. Critically, those risks are domestically based.
• Accordingly, although we expect another tightening in June, we have now put three cuts into the second half of 2007, bringing the Fed Funds rate back to 4.75% by the end of 2007.
• More aggressive tightening in 2006 would, in our view, only add to the cutting cycle in late 2007 and into 2008.
• Markets and most forecasters have not yet focused adequately on this issue and its implications.

Forecasting inflation

We are becoming increasingly concerned about the outlook for US growth in 2007, especially in the context of the Federal Reserve’s efforts to demonstrate its inflation-fighting credentials. Given the importance of this issue to the global economy and financial markets, we thought it worth elaborating on our concerns.

Over the past decade or so, there has been much discussion among academics and central banks about the global trend to lower inflation. Equally, most have concluded that our ability to forecast inflation has probably reduced. That said, inflation targeting is increasingly recognised as a desirable policy focus for central banks, and is a key policy of the Reserve Bank of Australia).

US Federal Reserve chairman Ben Bernanke favours moving the Fed in a similar direction. That said, over the past year or so inflation has moved upwards, to the concern of most central banks. For a new Fed chairman trying to prove his inflation-fighting credentials, this lift in inflation at a time when the economy is still travelling reasonably well is clearly a concern '” and he has made no secret of his concerns.

While modelling core inflation is fraught with difficulties, some of the key issues can be seen by looking at the results of some of our modelling of US inflation. Key points include:

• After falling significantly through the 1990s and being lower than most expected in the early 2000s, core inflation has now moved up and is higher than those same models (or forecasters) might have expected.

• Model forecasts for the US suggest that the core private consumption deflator will remain at, or above, the top of the Fed’s target range until the end of the year. Presumably, the Fed’s own models will be pointing in a similar direction. Hence, we have seen many Fed spokesmen stressing inflation concerns.

• Part of the increase in inflation reflects second-round effects of higher oil prices (via increased import prices), at a time of reasonably strong growth. Certainly, our equation moves up as a response to this, together with the lagged impact of a weaker US dollar.

• However, it is worth noting that the US inflation process is not being driven by higher wages (indeed, the Employment Cost index has slowed recently).

• As the economy slows, our models suggest inflation will come back to the middle of the target range by late 2007.

• Our own US inflation forecasts broadly follow this pattern. We have a number of core inflation models (both quarterly and annual growth rates) but they are all essentially driven by wages and productivity (or unit labour costs), demand (gross national expenditure and/or the report of the Institute of Supply Management, formerly the purchasing managers index) and import prices (oil and currency).

Although the graph shows the short-run inflation risks, it needs to be stressed that inflation targeting is not about moving monetary policy to affect near-term inflation results. Rather, the focus should be on where inflation is likely to be in 12 months or so (forecasting difficulties notwithstanding). This is where we have issues with some of the talk coming from the Fed and other financial analysts.

Indeed, the above framework is somewhat different to the one that Bernanke has advocated in the past. Essentially, he argues that there are four relevant factors for the inflation outlook:

• Economic slack (output gap);
• Inflation expectations;
• Supply shocks; and
• Inflation persistence.

A feature of this framework, however, is that the combination of inflation persistence and output gaps will, given lags, tend to focus a lot of attention on current (especially core) inflation rather than future inflation. Indeed, there is some recent literature from the Philadelphia Fed that suggests the current level of core inflation is a better indicator of future inflation than the current level of inflation. All of this, together with our own expectations of core inflation staying high until late 2007, suggests that the Fed may well lift rates again in August after moving in June.

However, we are concerned that not enough focus is being placed on the forces pointing to slower growth in the US over the next year or so and what this implies for inflation in 12 months’ time. Put bluntly, too little focus seems to be placed on what the current stance of policy is doing to the activity profile in 2007. In that context, we would point to:

• The lagged impact of the policy moves that have already occurred; that is, their full impact is still flowing through.
• Part of this process includes the wealth effects of slower growth in house prices (as in Australia, elasticities on consumption spending from changes in US house price inflation are, in our view, large '” an elasticity of about 0.1 implies that a slowing in house price inflation to about 5% would lower growth in US consumption by about 1% over the next year or so).
• Oil prices are assumed to remain high and will continue to cramp profitability.
• There is little scope to expect further stimulus from the fiscal side (indeed, we expect a modest tightening in the fiscal position).

Putting the above together with a few other “reasonable” assumptions, our consumption forecasts point to a significant slowing over the next year or so. For example, assuming house price inflation slows to about 5% a year; real household disposable income grows at about 2.5%; the equity market grows moderately after recent corrections; and oil prices remain around current levels, our consumption function (although by no means perfect) gives some estimates of the likely impact on the spending behaviour of US consumers based on past relationships '” the impact being shown in the following chart.

Our own forecast does not fully take on board all of the predicted slowing, but still results in GDP slowing to about 2.75% in 2007 (and 2.5% in the second half of 2007). It is worth noting that we have allowed for only one further rate rise in 2006 (this week) and a series of cuts next year. Allowing the model estimates to fully flow through would lower growth to about 2% in 2007. A more aggressive monetary policy stance would further substantially lower these estimates.

Summary

We are not saying the US economy is about to fall over. But in our view the economy is facing more headwinds that many expect in late 2006 and into 2007. Inflation will, however, remain high in the near term, which could well see the Fed doing more on official rates. That, however, would in our view only result in the need for more aggressive cuts in 2007 and 2008. This outlook also has serious implications for other central banks '” and, in particular, may ultimately result in less zeal to increase rates in a number of other economies (certainly vis a vis current market expectations). The most obvious example here is the European Central Bank, but it may well also make the Reserve Bank of Australia a little bit more wary about taking out another bit of inflation insurance (or at least a second one).