Why economists often failed to predict CPI peak, look at the following report from WSJ:
Rewind 12 months, when consumer prices were rising around 3%, and the message was similar. The headline inflation reading ‘is likely to rise again in July due to rising food prices,’ Mr. Lu wrote on July 15, 2010, ‘but it’s very likely that it will just peak in July…and then will trend down on a rising comparison base.’
He was hardly alone. Almost all major economists at international banks, research houses and agencies missed the mark. Although ‘inflation continued to accelerate, we believe underlying inflationary pressures are dissipating quickly,’ wrote Yu Song of Goldman Sachs in June 2010. ‘Inflation might have risen again last month but the peak is probably not far away,’ said Mark Williams of Capital Economics on July 8 of that year. ‘Barring an unforeseen supply shock to food prices, consumer-price inflation should peak in midyear and begin falling in the second half of 2010,’ wrote economists at the International Monetary Fund in a report that month.
In fact, inflation jumped to 5.1% in November, stayed high for months, and is now climbing again.
For investors, executives and others around the world trying to guess where China’s economy is headed now, understanding why analysts got it wrong last summer can help to gauge how much confidence should be placed in the current forecasts.
Predicting prices and other factors is always tricky, especially in an economy as large and complicated as China’s, where there is limited transparency in the official data. And bucking the consensus is always difficult for analysts.
Chinese leaders have been among those underestimating inflation: The average CPI surpassed the government 3% full-year target for 2010, and Premier Wen Jiabao last month acknowledged it will likely exceed this year’s target of 4%, too. The influence of government pronouncements on forecasts can be pernicious in China, because they are often seen as reflecting inside information.
‘Forecasting inflation is always much more difficult than forecasting growth,’ says Merrill’s Mr. Lu. ‘So take all these inflation forecasts with a grain of salt. This is something where we should all be humble, especially for forecasting a specific point in time for peaking or bottoming.’
An examination of the past predictions suggests several reasons that analysts were caught by surprise.
For one, they underestimated the role played by the explosion of cash in the economy resulting from government-driven stimulus spending in 2009 and 2010 aimed at fending off the global recession. China’s total stock of outstanding bank loans rose by
nearly two-thirds during that two-year period, a deluge of credit that sent money sloshing around the economy, putting cash in consumer pockets and eventually driving up the prices for everyday goods.
‘I guess you could say that the persistence of inflation has lent credence’ to the view that surging money supply has driven inflation, says Mr. Williams of Capital Economics in London. ‘I still struggle with that, though, because when you look at where the inflation is coming from, it’s still pretty much about food, and specific foods as well, and I still struggle to see how increasing loans to state-owned firms pops up a few minutes later in cabbage prices.’
Indeed, food prices have been perhaps the biggest driver of China’s inflation, rising 14.4% in June. Food prices are inherently volatile, depending on unpredictable factors like weather conditions.
Floods and droughts occur almost every year in China, and in such a huge country with a largely underdeveloped agricultural sector, those factors can cause significant temporary disruptions. Floods were behind a surge in vegetable prices last year, and drought conditions in wheat-growing areas pushed up wheat prices early this year.
‘The key problem in forecasting CPI in China is that food has been responsible for about two-thirds of the overall CPI rise, and food prices are notoriously difficult to predict,’ says Andy Rothman, China macro strategist at CLSA Asia-Pacific Markets.
Adds Wang Tao, China economist at UBS: ‘I am only an economist and could not forecast weather and natural disasters.’
Despite their earlier missed calls, economists remain confident that inflation is set to slow in the second half of the year. Money-supply growth has slowed considerably in response to tightening measures by the central bank. Forecasters also expect a recent surge in pork prices that has driven up the CPI to be temporary, as farmers respond by breeding more pigs.
A high inflation rate late last year also means China’s inflation rates will appear lower in coming months when compared against the year-earlier period, which is how inflation in measured in China. For that reason, it will be much more important to look at how prices relate to the previous month, rather than a year ago.
But wholly unpredictable factors, such as bad weather, could once again intervene. Given the regularity with which such shocks occur here, forecasters shouldn’t count on trouble-free farming from June onward.
At the end of the day, economists acknowledge that inflation forecasting in China is imprecise. Says Mr. Williams: ‘Although we are still expecting inflation to come down, the confidence with which we hold that view is probably lower than it was a few weeks ago.’
My takes:  Forecasting CPI is just like predicting crises, difficult but possible;  the key is to establish a model that capture key info;  for China those key factors are food prices, # of live pigs in farmers backyard, and corn price;  another constant factor is money supply.
via China Price Watchers Predict Another Peak-华尔街日报.