Sign in
Research link-chevron Created with Sketch.
link-chevron Created with Sketch. Products and Services link-chevron Created with Sketch.
link-chevron Created with Sketch. Products and Services
Economics link-chevron Created with Sketch.
Equities link-chevron Created with Sketch.
Analysts
Analysts
Help and Support
Help and Support
Inside China 06 May 2019

Was Q1 the high point for 2019?

Jeremy Stevens

Many want to know if China’s economic cycle has bottomed. We here provide a summary of the underlying tapestry of data. Our China Activity Index (CAI) has improved in recent weeks but it is tough to tell if this is temporary, or a genuine turning point, owing to the typical distortions in Q1 because of seasonal factors. Nevertheless, we believe that Q1 was likely was the high point for macro data in 2019. 

The question matters a great deal: China’s macroeconomic performance has been one of two critical forces – along with US financial conditions – to shape the path of risk assets over the past six months. When expectations around China’s data softened and US financial conditions tightened in H2:18, global growth and asset prices stumbled. Conversely, when China firmed and US financial conditions eased, as in the last few months, expectations around growth and asset prices improved. Hence, China data in the coming weeks will be closely watched.

However, the latest China PMI confirms our view that downside risks loom large over the better-than-expected March data.  After jumping in March, both manufacturing PMIs fell back in April, disappointing hopes for a further recovery.

We have been using an alternative tracker of Chinese economic activity for some time. Indeed, a plethora of studies has found official reported GDP growth data to be unreliable. It is not just questions of reliability that prompted the development and deployment of the tracker – which we named the China Activity Index (CAI), but rather, the deficiencies in official GDP series are its low frequency – updated on a quarterly basis – and long delay, for a month or two. As such, our alternative series that can serve as a proxy for Chinese economic activity on a more frequent and timelier basis is useful.

To this end, we have identified a set of data series that reflect the most important aspect of the economic fundamentals, such as consumption, industrial production, and foreign trade. Essentially, the CAI extracts and aggregates the most indicative components out of all the underlying set of data series. As a result, we can read through a single index to get a sense of the direction of Chinese economic growth. (refer to PDF for heat map)

The inputs of our model consist of more than a dozen of data series that depict the economic condition of the real economy across different sectors. The heatmap below is an intuitive picture of the trends – recording year-on-year (y/y) growth rates for all except consumer confidence and PMI – in this wide range of macroeconomic metrics. Cells in red stand for below-trend three-year trend growth; yellow for growth close to the average rate; and green for above-trend growth.

Most series turned yellow or red since the middle of last year, which indicated that China’s real economy was losing momentum. That is, until March, when a few variables showed improvement prompting many to declare that the economy has reached its cyclical bottom. In fact, 11/15 variables showed improvement last month.

Of course, the raw data published in China are not perfect for analysis due to seasonality – especially the Spring Festival – so we have attempted as best possible to seasonally adjust the figures. Nevertheless, it is fair to assume there remains still a considerable influence from Chinese New Year. Industrial production accelerated from 5.3% in February to 8.5% in March – obliterating expectations and recording the fastest pace of expansion since 2014. Thus, the data remains very difficult to interpret.

Indeed, in February the CAI was continuing to weaken, coming in at the lowest level in a year, which is why we remain cautious, waiting for April data before we believe that the stabilization has fully crystalized.   

Turning back to our CAI, we deployed Principal Component Analysis (PCA) to the data series. PCA is a commonly used systematic tool to reduce the dimensions for data series. It allows us to extract the indicative information out of multiple time series into composite(s), while maximizing the explanatory power in the composite(s). Then we regress the observed vector of outcomes on the selected principal components as covariates, using ordinary least squares regression to get a vector of estimated regression coefficients. Following which, the estimates are transformed back to the scale of the actual covariates, using the selected PCA loadings to get the final PCR estimator for estimating the regression coefficients characterizing the original model. Herein are the weighting for the index or CAI, which is basically rescaled principal components of the inputs.

The red line and blue lines are the China CAIs we have created – smoothed by 3 and 12 months respectively. The CAI traces back to 2008 against real GDP. Clearly the official GDP growth data is less volatile or mute to cyclical forces. In contrast, the CAI provides a more reliable proxy for economic cycles. The improvement in the underlying data in recent weeks is captured in an improvement in the CAI from 5.95 y/y in February – the lowest reading in a year – to 6.96 y/y in March – a four month high. (refer to PDF for CAI graph)

The question is – at least from a cyclical perspective – whether this is the bottom in the near term. Whilst the dashboard of more than a dozen macroeconomic metrics is showing signs of improvement, they are only just turning from red to amber, and the Q1 data in China is notoriously difficult to rely on.

It is certainly noteworthy that the CAI had its biggest monthly increase in over two years in March. Nevertheless, the change in CAI is rather noisy and it is more effective to use a moving average of 6 and 12 months, respectively, to gauge turning points. And on this score, after the gains in March, momentum loss has stopped and both moving averages are basically flat. So, was March a one-off, or a turning point? (refer to PDF for momentum gauge graph)

The underlying credit cycle seems to have bottomed out. Credit is seen as a precursor to gains in the real economy. Longstanding trends imply that the acceleration in credit expansion leads economic growth with a six- to nine-month lag. Boding well for the stabilization narrative, Total Social Financing (TSF) increased by CNY 2.86tr in March – far higher than the expected CNY1.85tr. Overall credit growth accelerated from 10.1% y/y in February, to 10.7% y/y in March.

That said, some caution is warranted. Take fixed asset investment for example, which accelerated from 6.1% y/y in January and February to 6.3% y/y in March. Local government spending has helped, but they have issued a record CNY1.4trn in bonds to finance the spending.

Meanwhile, manufacturing investment, which has been the focal point of Beijing’s efforts to stabilize the economy, are continuing to struggle, expanding 3.7% y/y in March – down from 5.9% in January and February. In addition, turning to nominal GDP growth numbers, the economy continued to lose momentum, slipping from 9.1% in Q4:18 to 7.8% in Q1:19.

It is also worth stating that even those who believe this is the beginning of an economic stabilization have been cautioning that any improvement in China would be modest. Meanwhile, some are cautioning that such positive data will reduce the government’s willingness to push through economic support measures. In fact, owing to the outperformance in the monthly data in March, it is our view that this is most likely the high-water mark for the economy in 2019, but we will be watching the data for confirmation.


Read PDF