【英语中国】中国经济指标出错原因何在？

The Conference Board's attempt to replicate in China the respected series of economic indicators it compiles for the U.S. has suffered a fresh blow. After a promising start, the New York business group was embarrassed last month when it had to correct its Leading Economic Index, or LEI, for China after it discovered a mistake in the calculations.

The revision got an unusual amount of attention after some reports blamed the change for a sharp fall in the Chinese stock market that day, even though few Chinese investors are aware of the index. Some economists criticized the Conference Board's procedures, and the whole episode led the organization to re-examine how it produced the indicator.

It turns out there were more fundamental problems than a one-time mistake. In the latest update of the China leading indicator, the Conference Board says it has had to make a 'technical adjustment' to how it uses one of the components of the index, a measurement of supplier delivery times from China's monthly purchasing managers index, or PMI, survey. That component is now being 'inverted.'

In plain English, that means the Conference Board had been putting that part of the indicator into the calculations backwards, so it pointed up when it should have pointed down, and vice versa.

The confusion arose because of differences among countries in how PMIs are calculated. The standard purchasing managers' index survey asks respondents whether deliveries from their suppliers are getting faster or slower. In theory, a faster delivery time means that suppliers aren't as busy, and there is slack in the economy, while a slower delivery time means suppliers are busier and the economy is growing.

The PMI then converts survey responses to a diffusion index, where values above or below the breakeven level of 50 indicates the direction of change. In U.S. surveys, slower supplier delivery times generate a reading above 50, signaling a growing economy. But in the PMIs for some European countries, it is faster delivery times that generate a reading above 50. To make the European-style one comparable to the U.S., it has to be turned around. (The math is actually simple--just subtract the original reading from 100.)

The Conference Board now admits that it had been mistakenly treating the PMI produced by the China Federation of Logistics & Purchasing as a U.S.-style PMI, when in fact it is a European-style one that needs inverting. The Conference Board says it made the change after confirming the methodology with experts at the CFLP and the National Bureau of Statistics.

But the CFLP's methodology was not a closely kept secret. In the English-language press releases it issues each month, the CFLP includes a line explaining the delivery time component of the PMI: 'In general, a reading above 50% implies faster delivery and a cooling-down economy; below 50%, slower delivery time and a booming economy.'

Last month, Merrill Lynch-Bank of America economist Lu Ting publicly drew attention to the Conference Board's handling of the PMI indicator, arguing in a published report that it seemed to be forgetting to properly invert the supplier delivery times component. The Conference Board says those comments helped it discover the mistake.

'Lu Ting was one of several people who's been giving us feedback on the China LEI. As a general principle, we welcome feedback from the analytical community, and inputs and comments on the indicator which feed into our continuous evaluation process,' said Bill Adams, the Conference Board's resident economist in Beijing.

The oddest thing about the error is that it doesn't seem to have significantly changed the signals the China leading indicator has been sending over its short history. The recalculation to properly handle the supplier delivery times 'does not impact [upon] the LEI's historical trend,' the Conference Board said. That seems to mean that the supplier delivery times is not a particularly important signal of coming changes in the economy.

Says Adams: 'This component hasn't had movements that would have a big impact on the composite LEI in either indication. We're including it because it's a measure of dynamics in supply chain and it measures capacity utilization, and these are important factors.'

世界大型企业联合会(Conference Board)编制的美国经济指标被各方所看重，但它推出同等质量中国经济指标的努力却遭受了一次新的打击。在雄心勃勃推出这一系列指标后，这家纽约的商业团体上个月却不得不尴尬地对自己不久前公布的中国领先经济指标进行修正，因为它发现自己在指标的计算中出现了错误。

在一些报导将指标修改当天中国股市的暴跌归咎于这件事后，这一修正行为引发的关注达到了不同寻常的高水平，尽管没有多少中国投资者知道这一指标的存在。一些经济学家对世界大型企业联合会的数据发布程序提出了批评，这一插曲已导致该机构重新评估了自己计算中国领先经济指标的方式。

事实证明，世界大型企业联合会在计算中国领先经济指标方面存在的基本性问题要多于一次性错误。该机构称，在计算最新的中国领先经济指标时，它已不得不对自己使用中国采购经理人指数中供应商配送时间指标的方式进行“技术性调整”。这一指标现在被“反向使用”了。

从英文字面理解，世界大企业联合会在计算中国领先经济指标时曾将供应商配送时间指标用反了，也就是说该指标本该显示下跌时却被当作了上涨，反之亦然。

之所以会出现这种乌龙事，原因是各国在计算采购经理人指数时存在不同做法。编制采购经理人指数时标准的调查方法是，要求受访者回答其供应商的交货速度是在变快还是变慢。理论上说，交货时间加快意味着供应商不像以往那么忙碌了，也就是说经济出现了疲软。而交货速度放慢则意味着供应商变忙了，也就是说经济正在增长。

采购经理人指数编制者随后会将受访者的答复转换成一个景气动向指数，该指数高于或低于50分别代表着商业景气将提高或下降。在美国计算这一景气动向指数时，供应商交货时间延长会导致景气动向指数高于50，从而表明经济在增长。但在一些欧洲国家计算采购经理人指数时，却是供应商交货时间缩短导致景气动向指数高于50。为了使用欧洲方式计算出的指数能够与用美国方式计算出的指数相互比较，就需要对一些指标反向使用。(其数学方法其实很简单──只需用100减去原数值就行了。)

世界大企业联合会现在承认，它曾错误地按美国方式使用了中国物流与采购联合会编制的中国采购经理人指数，而中国在编制这一指数时使用了欧洲方式，即供应商交货时间缩短时景气动向指数高于50，因此世界大企业联合会在编制它的中国领先经济指标时应反向使用中国的供应商配送时间指标。该机构说，它是在向中国物流与采购联合会及中国国家统计局的专家们确认了后者的指数编制方法后作出相关调整的。

中国物流与采购联合会的指数计算方法并不对外保密。在它每月发布的英文新闻稿中，该机构都会附上一行解释其采购经理人指数中供应商配送时间指标的文字：总体而言，指标高于50意味着交货时间缩短和经济的冷却；低于50意味着交货时间的延长和经济的繁荣。

上个月，美国银行/美林(Merrill Lynch-Bank of America)经济学家陆挺的言论引发公众开始关注世界大型企业联合会对中国物流与采购联合会所编制供应商配送时间指标的处理。他在自己发表的一份研究报告中说，世界大企业联合会似乎忘记反向使用中国物流与采购联合会的供应商配送时间指标了。世界大企业联合会称，陆挺的话帮助它发现了这一错误。

世界大企业联合会驻北京的经济学家叶文斌(Bill Adams,)说，陆挺是几位就我们所发布中国领先经济指标给予反馈意见的人士之一，我们总的原则是欢迎分析人士和各界对中国领先经济指标给予反馈、建议和评论，这有利于我们改进工作。

世界大企业联合会所犯这一错误的最奇怪之处是，被重新计算的中国领先经济指标，似乎并未改变该指标在推出至今这段不长时间内所反映的经济总体趋势。世界大企业联合会说，正确使用供应商配送时间指标后重新计算的中国领先经济指标，并未影响该指标所反映的历史趋势。这似乎意味着，供应商配送时间并不是反映中国经济未来变化趋势的特别重要信号。

叶文斌说，供应商配送时间指标的分量还不足以对中国领先经济指标产生明显的推升或下拉作用，我们之所以把这个分类指标包括进来，是因为它是一个衡量供应链动态情况和产能利用率的指标，而被它衡量的这两个因素都很重要。