An opportunity to take the guesswork out of investing in China; Nobody is sure whether they believe numbers supplied by China

December 1, 2013 2:40 pm

An opportunity to take the guesswork out of investing in China

By John Authers

Nobody is sure whether they believe numbers supplied by China

For the legions of China-watchers within the world’s investment community, there have always been two difficult questions. One, which applies everywhere, is to look at the figures and to ask whether growth can continue, or if a bubble is forming. A second question, much more specific to China, is whether the figures themselves are genuine, and whether words can be taken at face value.For an example, take the reforms announced at the Communist party’s plenum. They were very well-received, but followed by much analysis asking if the reforms were what they seemed.

It is not a new problem. Nobody is sure whether they believe the numbers coming out of China. Even Li Keqiang, now China’s premier, once commented that the nation’s gross domestic product figures were “man-made” and not to be trusted.

For most China-watchers, unable to speak the language and unfamiliar with the country, an asterisk has to hover over the figures. They tend to look for numbers which are hardest to massage, such as electricity generation (which showed an outright contraction heading into the crisis months of late 2008).

But there are now attempts to deal directly with the most important data, including inflation and growth. The tentative conclusion of a group of Columbia University economists led by Emi Nakamura – who formerly worked on inflation numbers for the US Bureau of Labor Statistics – is that inflation has of late been understated, while growth in consumption was overstated. In the 1990s, however, when inflation appeared to be much lower, the indications are that inflation was overstated and growth understated.

That does not come from independent checking of the data, but rather from checking the public data for their own internal consistency.

Those official data show remarkable growth, much of which is evidently genuine. But the stability of the growth is arguably even more startling than the overall numbers. For two decades, average output growth has ticked along at more than 9 per cent per year.

Meanwhile, Chinese inflation has been low and remarkably stable, given such fast growth, averaging less than 2 per cent and never exceeding 6 per cent since 1997. Can we believe this?

Ms Nakamura’s analysis rests on the concept of the Engel Curve, named for the Nobel laureate economist Robert Engel. He found a stable relationship between a person’s wealth and the proportion of their wealth that they spent on food. The wealthier you are, the less you spend proportionately on food.

On a chart, with total expenditure on one axis and the share of food on the other, there will be a curve – as total expenditure increases, the share devoted to food will decrease. This relationship will be constant over time. As total expenditure increases, so the share devoted to food will decrease.

However, when Ms Nakamura’s team produced Engel Curves for 1995, 1998 and 2000, they found the curves shifted downward over time. In other words, the official statistics showed that a household whose real (post-inflation) total expenditures stayed the same decided to spend much less on food over time.

That was true at almost all levels of income, and it was not because food had become relatively cheaper, because Columbia’s statisticians controlled for this. So the most likely explanations are that households were richer than reported in the official statistics, or that inflation in food prices was weaker – or, most likely, a combination of these.

Then, from 2006 to 2008 – the era immediately before the huge Chinese stimulus that coincided with the Lehman Brothers’ event and subsequent crisis – the Engel Curve shifted upwards. That means that people with unchanged wealth were spending more on food each year. The implication: that inflation was higher, and growth weaker, than the official data stated.

Further statistical detective work suggests that this might have been due to a shock in the supply of pork, a Chinese staple, as a result of an epidemic at China’s pig farms in 2007. That forced up food prices, reducing consumption of non-food items, and it may well have forced the government to act. The official data offered a “smoothed version of reality”.

Deep issues of statistical measurement are involved, but a battery of other tests, considering clothing, and considering the share of different foodstuffs within food, suggests that these results are robust.

There are caveats. The data only apply to consumption, and do not consider the economy through the lens of investment.

This kind of work needs to be done. To invest in China need not be a matter of guesswork. With similar analysis, it might yet be possible to explain the motives behind the latest reform drive.