Why China's official data don't add up

(ATF) Last week China released it’s much anticipated third quarter GDP, and along with it came fresh concerns of data accuracy.

While the 4.9% print missed rather exuberant market expectations of over 5% growth, the real head-turner was Beijing’s announced revisions to its fixed asset investment (FAI) data for September 2019. The revised numbers depressed last September’s FAI by 2.8 trillion yuan, thereby boosting this year’s growth figures as well as the Q3 GDP.

China’s National Bureau of Statistics explained away the change in a short footnote, stating that historical data were adjusted on the basis of the “fourth national economic census unit inventory.” In August, China had reported its first positive retail sales growth of 2020 on the basis of a similar adjustment, which depressed year-ago figures by 50 billion yuan.

Predictably, this latest revision came under immediate scrutiny.

The core statistical problem here is the sample’s changing characteristics and size. The NBS has a longstanding practice of only surveying firms above a minimum annual revenue. Any firm that falls below the threshold is dropped from future surveys and its results removed from past data. Alternatively, if the number of firms meeting the revenue threshold rises in a given year, the sample then becomes larger than last year’s by thousands of firms.

This cherry-picking approach – which can produce trillions of yuan worth of adjustments – invalidates any claims of random, representative sampling. Furthermore, it creates a constant stream of revisions to official statistics without details on methods or the release of underlying numbers. The more fundamental challenge it underscores is that investors cannot rely on official data to reliably capture economic conditions, especially when the economy is under pressure.

My firm, China Beige Book (CBB), was founded in 2010 to serve as a check on these types of behind-the-curtain machinations, which are all too common when dealing with Chinese government data. CBB tracks the Chinese economy in virtual real-time by directly surveying thousands of Chinese firms across all major sectors and regions.

Our most recent results cast serious doubts on Beijing’s upbeat recovery story.

Growth uneven, nowhere near last year

First, despite the announced on-year rebound in official GDP growth, CBB’s recent third-quarter data show no part of the economy registering growth anywhere near year-ago levels. Despite sequential quarterly improvement since the Chinese economy emerged from its Covid lockdown, every sector, sub-sector, region, and growth metric remains in contraction year-on-year.

Second, our surveys of over 3,300 firms showed a two-track recovery in Q3. For large firms and companies based in Shanghai, Beijing, and Guangdong the economy is accelerating. This is the recovery much of the West sees and what has become the public-face of Beijing’s rebound narrative. In the rest of country, however, the climb back is far more uneven and beset with reversals.

For example, while coastal firms reported a double-digit expansion in production, revenue, and profits compared to the second quarter, most interior provinces saw on-quarter declines. Official data are not reporting this recession.

Finally, our data belie claims of a consumer-driven rebound. Instead, the economy continues to be powered by the old industrial sectors, with Manufacturing leading across the board, Commodities rallying, and Property gaining steam.

Hospitality, travel sectors in deep trouble

By comparison, consumer-facing industries continue to struggle.

While Retail has emerged from on-quarter contraction, the Services sector lagged in the data, with hospitality, restaurants, and travel industries in the deepest trouble.

With its revisions to last year’s numbers Beijing is setting the stage to announce full year GDP growth for 2020. This despite many official data contracting through much of this year.

As the political imperative to claim victory over the Covid-induced downturn takes center stage, investors should not bet on murky official statistics adding up.