WEEKEND ECONOMIST: CPI won't inflate rates

There were no real surprises in this week's minutes for the Reserve Bank Board meeting in April. However, the theme for the Australian economy of the large schism that has opened up between prospects for the mining sector and the resources boom relative to the domestic focussed and Australian dollar exposed sectors has been accentuated more so in those minutes than we have seen before.

On the mining boom, the RBA uses the all important final paragraph on "Considerations for Monetary Policy" to highlight "the high level of the terms of trade and the prospective further large increase in investment". We cannot recall this theme figuring so prominently in the final paragraph in previous minutes.

The conclusion on the policy stance gave no indication that the RBA has any near term intentions. When the RBA intends sending that signal, the final sentence usually talks about keeping the cash rate unchanged, "for the time being" or "pending additional information". The final sentence in these minutes is "members therefore did not see a case to change the cash rate". Arguably, this statement seems a little stronger than the March statement of the "Board therefore decided to leave the cash rate unchanged".

Next week the market will be firmly focussed on the March quarter Inflation Report, which will be released on April 27. Of course, the policy makers mainly pay attention to the core inflation print.

We forecast that the average of the two RBA measures of underlying inflation increased by 0.7 per cent in the quarter up from 0.4 per cent in the December quarter – with both measures estimated at a rounded 0.7 per cent. That will see annual underlying inflation drop to 2.1 per cent (trimmed mean) from 2.2 per cent.

Despite careful seasonal adjustment in a number of price series this "jump" in the March reading of the trimmed mean is consistent with recent years. In March 2009 the trimmed mean jumped from 0.6 per cent in December 2008 to 1 per cent; in March 2010 it jumped from 0.6 per cent in December 2009 to 0.8 per cent. Furthermore, the trimmed mean print in March of 2008; 2009; and 2010 proved to be the highest quarterly read in each of those calendar years. There is significant seasonality in the CPI components of health; education and utilities. It appears that the seasonal adjustment factors used to calculate the trimmed mean and weighted median may not be fully capturing this seasonal effect leading to some upward bias in the underlying series for inflation in the March quarter.

Taking these considerations into account, we do not believe that the RBA will be particularly unnerved by a near doubling of the core inflation measure in just one quarter. Furthermore, even if the print was to stay around 0.7 per cent for the rest of the year it would still only be consistent with the RBA's current forecast of underlying inflation for 2011 of 2.75 per cent.

The fall in the prices of a range of discretionary goods observed in the December quarter due to soft consumer demand and the high Australian dollar is expected to continue in the March quarter. We expect prices of cars (-0.6 per cent); clothing and footwear (-1.6 per cent); and household contents and services (-0.9 per cent) to all fall again.

Offsetting these falls are substantial increases in a number of food items: fruit (17 per cent); beef (7 per cent); cereals (6 per cent) and vegetables (4 per cent). These increases are largely the direct result of the destruction of food supplies from the floods and cyclone in January.

These increases are likely to be partially offset by a 10 per cent reduction in milk prices, reflecting the current discounting by the major supermarket chains. Overall we expect food prices to increase by 2.2 per cent – remarkably, that is the same increase as in the December quarter when wet weather also affected food supplies. Petrol prices are expected to have increased by 8.5 per cent while utilities are forecast to rise by 3.4 per cent.

However all of these large items will not fall within the 15th to 85th percentile of price increases (weighted) which is used to calculate the trimmed mean. We forecast that the price moves within the trimmed mean will be the –0.6 per cent to 3 per cent range.

Key variables which in the trimmed mean range will include those associated with housing; financial services; holidays; motoring; education (seasonally adjusted); holidays (seasonally adjusted); some health (seasonally adjusted) and some utilities (seasonally adjusted).

Housing pressures are expected to be in line with the December quarter with house purchase increasing by 0.8 per cent (as indicated by the first quarter PPI detail) and rents by 0.75 per cent.

In the December quarter the measured price of financial and insurance services fell by 0.4 per cent. We expect this item to increase by 1.5 per cent in the March quarter. This change is dominated by our forecast for deposit and loan facilities (D&L). This component has a weighting of 3.9 per cent in the CPI behind only house purchase (8 per cent); rents (5.9 per cent); and motor vehicles (4 per cent). It is designed to measure the cost of retail financial services by monitoring the spread between mortgage rates and retail deposit rates. In the period from September 2009 to September 2010, which captured the recent rate hike cycle, the average quarterly move in the series was 1.3 per cent. That implied that banks were widening their retail margins at a time of rate hikes.

In the December quarter when we saw the banks increase their mortgage rates by more than the RBA's rate hike of 25bps, the D&L series actually fell by 1.2 per cent. That implies that banks raised their deposit rates by more than they raised their mortgage rates following November's rate hike. That move seemed inconsistent with our analysis of the RBA's estimate of the movements in the bank's retail deposit and loan rates. Accordingly we are forecasting an increase of 2 per cent in the D&L series as a catch up for the unexpected measured fall in the December quarter.

This forecast is very important to our overall estimate of the trimmed mean. An assumption of a zero would push our trimmed mean forecast down to 0.6 per cent and a repeat of the fall in the December quarter would reduce the trimmed mean estimate to near 0.5 per cent. While we think a 0.7 per cent print will be neutral for the RBA a print of 0.5 per cent would make the case for a rate hike over the next 6 months very difficult.

A key theme of the inflation prints over 2010 was the constraining effect discretionary goods prices were having on inflation due to soft household demand and the high Australian dollar. We expect these effects to be sustained in the March quarter but at a less intense pace.

This assessment is supported by both the trend in retail volumes in the December and March quarters (as a measure of the influence of the "cautious consumer") and the movement in the Australian dollar. Retail volumes in the March quarter are expected to increase by 1 per cent (admittedly partly inflated by post flood replacement spending) compared to a 0.3 per cent contraction in the December quarter.

In trade weighted terms the average Australian dollar increased by 6.2 per cent in the December quarter compared to 1.1 per cent in the March quarter.

Overseas holidays can sometimes make a big difference. Markets were surprised by a low CPI read in the June quarter of 2010. That was partly because overseas holidays, which have a 1.7 per cent weighting in the CPI, printed a negative 2 per cent in the quarter. Evidence of airlines raising their prices in the wake of the Icelandic volcano in the June quarter prompted forecasters to pencil in a hefty increase in overseas holidays. There was a sigh of consternation when overseas holidays printed a 2 per cent fall in the quarter – that error alone was worth up to 0.1 per cent on some forecasts of the core CPI.

The errors were made because some forecasters were unaware that the Bureau assesses overseas holiday costs with a two month lag, as the price is recorded when the holiday is "consumed" not when it is "booked". If, for instance, airfares rise in February they are only registered for April. There were widespread increases in fuel surcharges for overseas airfares from February 2011 but these increases will not appear in the CPI estimates until the June quarter. Accordingly we are forecasting a flat outcome for overseas holidays.

Around 40 per cent of the cost of overseas holidays is airfares. Another 50 per cent is package tours. The terrestrial component of package tours is likely to be negative reflecting soft global demand and the high Australian dollar. That should be enough to offset likely modest increases in actual airfares (reported fares net of unofficial discounting).

Domestic holidays are forecast to have fallen by 2.7 per cent in the March quarter following a 3.7 per cent increase in the December quarter. That compares with a 2.5 per cent fall in March 2010. For purposes of calculating the trimmed mean domestic holidays are seasonally adjusted with a 0.3 per cent fall being consistent with the 2.7 per cent estimate.

It is also important to note that the following items are seasonally adjusted for the purpose of calculating the trimmed mean: a range of food items; some leisure items; utilities; health items; urban transport fares; domestic holidays and education. As discussed, there may be increasing seasonality in items which are currently not seasonally adjusted or the seasonal factors currently used are understating the seasonality.

All in all, some key assumptions around deposit and loan facilities; a weaker effect from the cautious consumer and the appreciation of the Australian dollar and some residual seasonality are supporting our view that the underlying measure for the CPI in March will print markedly higher than the December print. However we do not expect that this jump will be enough to significantly change the current "relaxed" rhetoric from the Reserve Bank.

Bill Evans is Chief Economist at Westpac.