09/06/2011: CPI data for May

Consumer Price Inflation data for May is out today. Recall that a month ago, higher mortgage costs and oil prices pushed inflation to a 30-month high, with prices in April up 0.4% mom and 3.2% yoy. This was the second highest rate of annual inflation since 2008. This time around, the catalyst for inflationary pressures was supposed to be mortgages costs, as ECB hike of 25bps in April was expected to feed through to retail rates. CSO is very careful about this aspect of inflation, having issued in the latest release an explanatory note (see below). Market expectation, consistent with my view expressed in December-January issue of Business & Finance magazine, is for inflation to average around 2.8-3.1% in 2011.

Now, on to today's data:

• May CPI rose 0.1% mom - below the markets expectations and below 0.6% mom rise in May 2010. Yoy inflation was at 2.7% in May 2011, again below expectations in the market.
• HICP - omitting, among others, cost of mortgages, car and home insurance, car taxes etc (see CSO note on this in the main release) - posted 0% change mom against 0.3% increase mom in May 2010. Annual HICP rose 1.2% relative to May 2010.

Charts to illustrate - first CPI, then two indices of prices:
In annual terms, largest increases were posted in

• Housing, Water, Electricity, Gas & Other Fuels - up 8.5% after posting 11.8% rise in April and 12.5% in March. Within the category, Rents posted a 1.0% decline yoy and 0.1% increase mom, while mortgages interest costs posted a 0.6% mom rise and 20.1% increase yoy. Electricity, gas & other fuels sub-category posted a 1.0% decline mom and 6.6% rise yoy with Liquid fuels falling 3.8% mom and rising 17.9% yoy.
• Miscellaneous Goods & Services posted a 8.4% increase yoy primarily driven by Insurance (+15.9% yoy) of which Health Insurance (+21.6% yoy, but -0.6% mom) was the biggest culprit. Motor car insurance was up 7.6% yoy and 0.7% mom.
• Communications were up 4.1% yoy - driven solely by 4.3% rise yoy in Telephone & communication services.
• Health was up 4.0% yoy - hospital services up 11.4% yoy (no change mom) followed by Pharmaceutical products (+2.5% yoy and 0% change mom)
  

Deflation was recorded in

• Furnishings, Household Equipment & Routine Household Maintenance (-1.9% yoy and -0.1% mom) with strong deflationary momentum in Furniture & furnishings (-5.7%), and Major household appliances (-4.0%)
  
• Education - down -1.3%- driven by 1,8% yoy decline in Other education and training and -1.4% drop in Third level education. On the opposite side of the spectrum, Primary education costs rose 1.3% yoy and Second level education costs were up 0.8% yoy.
  

Charts to illustrate these trends:

As usual - an imperfect measure of state v private sector controlled prices - first straight forward state-controlled or dominated or influenced sectors:

Next - an index of prices in two broadly defined sectors:
One point worth making - the above chart clearly shows that inflation has moderated in state-controlled sectors. It remains to be seen if this welcome change mom will translate into a longer term trend.

Finally, a point, as promised above, on the issue of mortgages costs. CSO provides a handy explanation of their terminology on page 10 of the main release, from which I quote here:

"... current approach to measuring mortgage interest in the CPI reflects the situation in the base reference period December 2006 when the standard variable rate was dominant. Subsequently, tracker mortgages have become more popular. This did not give rise to any difficulties while the standard variable and tracker mortgage interest rates moved broadly in line with one another, which would be the normal expectation. However, the decoupling that has taken place since August 2009 has resulted in dramatically different trends emerging. For example, between September 2009 and September 2010 the standard variable rate increased from 2.93% to 3.66% whereas the tracker rate did not change. The Mortgage Interest component of the CPI, which is largely determined by the trend in the standard variable rate, increased by 25.1% as a result and contributed +1.25% to the overall change in the All Items index. It is crudely estimated that the latter impact would have been reduced by between 0.2% and 0.5% had the Mortgage Interest component been calculated on a current weighting basis."

So what CSO are saying is that current mortgages costs metric overstates the overall impact of mortgages costs increases on CPI because more mortgages, since 2006, were issued in the form of tracker mortgages. That's fine, but there is also a sticky problem of the weights assigned to all spending categories, which are all based on December 2006. If since December 2006 the following changes took place:

1. Overall costs of mortgages rose relative to other costs,
2. Home ownership proportion in population rose (which could have been due to emigration out of the country selecting predominantly non-homeowners, for example),
3. There have been significant exits from tracker mortgages and fixed-rate mortgages since 2006 (perhaps due to either selection bias in defaults or due to bias in favor of fixed rate mortgages in maturing mortgages, for example)
   

Then the weights used for this sub-category of spending might be below their current levels, off-setting the above effects of tracker mortgages.