Figuring out the Beijing numbers game

PUBLISHED : Saturday, 18 December, 1999, 12:00am
UPDATED : Saturday, 18 December, 1999, 12:00am

This week we had another instance of economic figures from the mainland into which Beijing-watchers tried their best to read some trends but really couldn't do it.

Let's take them one by one.

First we had the retail price index (RPI) for last month, which showed a year-on-year decline of 2.8 per cent compared with a 2.6 per cent decline in October.

This was fairly widely taken as evidence of deepening deflation just at a time when hopes were rising that the mainland could work its way out of its deflationary spiral.

Now look at the first chart and say that the last slight twitch downwards in the bottom line at the right, representing the RPI inflation rate, really makes any difference. It is in fact what we call statistical noise. These figures really don't tell you anything about a trend one way or another.

The only mainland inflation indicator that really shows any trend is the raw materials price index and it has just swung back up into positive territory.

This is in line with the trend elsewhere in Asia and tells you that the mainland's RPI will probably swing up too in coming months.

Next we have retail sales. The raw November figures say they rose 7.8 per cent year over year in value terms, down from 8.2 per cent in October, which presumably represents bad news.

Let's ignore the fact that either 7 or 8 per cent is still higher than it had been for almost a year up to October.

The real issue is what the change has been in volume terms after adjusting for inflation.

To see the trend here you also have to look at the figures on about a six-month average basis to knock out the volatility to which they are prone every month.

The result of doing all this, as the second chart shows, is that volume of retail sales has actually been rising a little recently but once again there is no firm trend, certainly nothing like the big collapse and recovery in turbulent 1989.

Finally we have the announcement that fixed-asset investment by the state sector last month rose 5.9 per cent year over year, which must be good news because the figure was only 0.6 per cent in October and it was negative in September.

Here we have the difficulty that, according to the official figures, the state sector every year does about a third of its investing for the entire year in December alone and, year after year, invests nothing at all during February.

This is almost certainly the result only of a reporting convenience and, of course, it raises a big question mark over every monthly figure.

Once again look at it on a year-on-year growth basis with a six-month moving average to knock out the extreme ups and downs which this unusual reporting produces.

You now get a growth figure for last month of only 2.9 per cent, down in an almost straight line from 28 per cent in November last year and about as low as the figures on record have ever showed it.