Characteristics of the China Stock Market

As represented by the Shanghai Stock Exchange Composite Index

The Shanghai Stock Exchange Composite Index (SSE) is composed of all eligible stocks listed on the Shanghai Stock Exchange. This index is designed to reflect the overall market performance of companies listed on the Shanghai Stock Exchange. Since the Shanghai Stock Exchange is by far the largest stock exchange of China, and since the SSE is a market capitalization-weighted Index, the performance of its blue chips are the bellwether of the Chinese economy.

Type of investors in the China stock market

A large portion of the investors in the China stock market are mostly retail investors unlike the US stock market where institutional investors dominate the market. According to the Shanghai Stock Exchange, retail investors account for at least 80% of total trading volume in China, whereas this figure only amounts to 10% in the US. In terms of account numbers, retail investors contribute 99.9% of the total accounts in the Chinese market. Thus, you will find that sentiments, word of mouth (rumors) and technicals (Technical Analysis Indicators) are still significant factors that affect the SSE’s performance-and not the fundamentals of the market.

In our opening chart above, you can see that there is a big spike in the trading volume just before and after the 10 October China National Day. Retail investors typically trade more during this period hoping that the government will make announcements on the economy that will have positive impact on the stock market. However, the fact that the China stock market has been down for so long has resulted in many retail investors taking the chance to liquidate, resulting in the Index falling. This can be seen in the Table which represents the Open, High, Low, Close and Volume of the Index on the dates highlighted. The Volume is up but the Index’s Close (highlighted by the Black rectangle) is down on those days.

Statistical characteristics of the market

When the market is driven by non-fundamentals, it will inherently be more dynamic and unstable. This means it will always be in a state of disequilibrium. And this can be shown by comparing its Probability Distribution Function (PDF) with that of the Dow Jones Industrials using 250 most recent data points. In the chart below, the SSE’s PDF on the right is much less symmetrical than that of the DJI.

Correlation with the US market

If you take a longer time frame of approximately 500 data points, the SSE will show a very low Correlation Coefficient of 0.27 against the DJI, due to the Chinese economy being relatively self-sufficient and large enough to be driven by its internal dynamics. But because of the current prospects of high tariffs being imposed by the Trump Administration; if you take just 100 of the most recent data points, the SSE currently has a Correlation Coefficient of 0.66.  You can see this in the chart below where it more closely follows the DJI in the latter part of the chart. Note: the numbers are in standardized Z-Score to enable inter-comparison.

Short-Term Forecast Trend

Nevertheless, using ARIMA, Bagged (resampling with replacement)  Linear Regression and fitting with unbounded Metalog Distributions we are able to fit the data into a form more amenable for forecasting its short-term trend. The results for 1-20 steps ahead autoregressive forecast after 1000 trials of Monte Carlo Simulation is moderately positive as shown below. Note: Bagging, also known as bootstrap aggregating, is a machine learning technique that improves the accuracy and stability of predictive models. It is used to reduce overfitting and variance in noisy data sets. 

The Support and Resistance at the 5% (Support)  and 95% (Resistance) quantiles are shown below. but note that the Standard Deviation of 45 and an Independence Test of jut 0.64 confirms our observation that the China stock market is much less predictable due to its dominance by retail investors.