Overall CPI includes property time warp
Those gloomy consumer price inflation figures for December released a few days ago show a few trends of interest on closer examination.
On first glance, they indicate that deflation is accelerating, with the overall consumer price index (CPI) down 1.6 per cent year over year, which is not good news in one of those rare times that the Hong Kong economy needs its bubble blown out rather than punctured.
But just two components pushed the figures into negative territory. The housing component showed a price decline of 3.3 per cent while clothing and footwear fell a whopping 16.2 per cent. Some people are either taking mammoth losses in the retail rag trade now or are no longer making mammoth profits, probably the second.
It is not clothing and footwear that has exerted the biggest downward pull on the CPI, however. This component accounts for only 6.6 per cent of the total basket. It is rather housing, which has a 28.8 per cent share and which has been skewing the inflation figures for several months now.
Strip housing plus clothing and footwear out of the figures and you get an inflation index that shows prices still rising although by a steadily lower margin than before, as the second accompanying chart shows.
Housing is a thorny problem for people who construct consumer price indices. The idea is to measure the regular monthly cost of housing but it is easier said than done.
Take the case of a homeowner paying down a mortgage, for instance. If interest rates are falling and home prices are rising, as they tend to do when interest rates come down, it does not make his housing costs rise. He locked in those costs when he bought his home. The only variable that affects him is the interest rate on his mortgage and this is falling.
So in a perverse way, rising home prices can actually mean falling home costs for a large segment of the public.
There are other such anomalies that make it difficult to track housing as part of the consumer-price basket. The normal response of statisticians to these is to exclude home prices entirely from the basket and measure only rents and some ancillary costs such as maintenance.
But then you get the problem that the actual rents being paid, the ones that a CPI basket should measure, tend to lag the trends in the property market. The standard two-year lease period for rented flats gets in the way here.
If a tenant has signed his lease just before the property market suddenly drops off a cliff it is just his tough luck. Two years from now he may get lower rents but for the moment he is one of those households that the CPI registers as showing no decline in housing costs.
Then those two years finally come up and he gets his rent reduction just as housing prices bottom out and start rising again. Now the CPI registers the housing component as showing a decline while the spot market is already rising.
This is what is happening in our inflation figures at the moment. We are seeing the belated impact of the property slump that began in 1997. These housing CPI figures do not measure price. They measure time and it never quite fits.
