A Strategy For Investing In China

Includes: CICHY, FXI, SPY, VWO
by: PopperTech


China's government and the US market explain two-thirds of FXI's variation since 2011.

The success or failure of China's reforms will likely drive FXI performance over the next year.

China has become the dominant force over other major emerging economies.

China's reforms are likely to significantly influence VWO over the next year.

Since the expense ratio of VWO is 58 bps lower than FXI, an investment in VWO may be a cheap way of gaining exposure to China.

Introduction: This article discusses the current relationships between emerging markets, the U.S., and China, and suggests how they are likely to change over the next year. First, I define a causal model relating China's stock market, the People's Republic of China (PRC), and the U.S. Next, I suggest that China's economic reforms will likely drive the performance of emerging markets over the next year. Lastly, I sketch alternative scenarios that could derail this thesis.

Model Structure: In this section, I propose that both the U.S. and the PRC drive China's stock market. I use the SPDR S&P 500 ETF (NYSEARCA:SPY), China Construction Bank (0939.HK, OTCPK:CICHY), and iShares China Large-Cap ETF (NYSEARCA:FXI), and Vanguard FTSE Emerging Markets ETF (NYSEARCA:VWO) as proxies for the U.S., Chinese Government, and the Chinese Market, respectively. The period under study ranges from December 31, 2011 to June 1, 2014.

To start with, I assume the causal direction between any two countries is the reverse of net imports. For example, the U.S. is a net importer from China, and consequently, I model the U.S. stock market as driving the China's stock market (Figure 1). My underlying hypothesis is that U.S. consumption influences Chinese GDP.

Figure 1 indicates the U.S. stock market influences the performance of China's stock market.

Source: PopperTech

Figure 2 indicates that the U.S. is a significant explanatory factor of FXI; although it only explains 40% of variation of FXI variance over the last two and a half years.

Figure 2 illustrates the results of an ordinary least squares linear regression with FXI as the dependent variable and SPY as the independent variable. Given the magnitude of the T Stat for SPY (test for statistical significance) and model R-Squared (explanatory power)/Correlation, it is pretty clear that a relationship between FXI and SPY exists.

Source: PopperTech and Yahoo Finance

In addition, Figure 3 shows the correlation between China and U.S. is decreasing. It appears more recent U.S. economic events have diminishing importance concerning the performance of FXI.

Figure 3 displays the 90-day rolling correlations between FXI and SPY.

Source: PopperTech and Yahoo Finance

Figure 4 indicates China Construction Bank (CCB) is also a significant explanatory factor over FXI. Although CCB only accounts for approximately 9% of FXI by weight, CCB explains 40% of FXI's variance.

Figure 4 illustrates the results of an ordinary least squares linear regression with FXI as the dependent variable and CCB as the independent variable.

Source: PopperTech and Yahoo Finance

To explain this, Figure 5 shows that the correlations between China's four major banks range from .82 to .90 over this period. Since these account for 27% of FXI by weight and move in tandem with each other, an investment in FXI results in a large, effectively single bet on China's banks.

Figure 5 displays the correlations between China Construction Bank, Industrial and Commercial Bank of China (1398.HK, OTCPK:IDCBY), Bank of China (3988.HK, OTCPK:BACHY), and Agricultural Bank of China (1288.HK, OTCPK:ACGBY).

Source: PopperTech and Yahoo Finance

The influence of the PRC on all of these banks explains the high correlations. According to the 2013 annual report for China Construction Bank (Page 62), the PRC, through its wholly-owned investment company, Huijin, owns 57%, 35%, 68%, and 40% of CICHY, IDCBY, BACHY, and ACGBY respectively. As a result of the high correlations and ownership structure, I use CCB as a rough proxy for the influence of the PRC.

Lastly, Figure 6 corroborates that both the U.S. and CCB drive the Chinese Stock Market. Both SPY and CICHY are statistically significant, and the model explains about two-thirds of FXI's variance.

Figure 6 illustrates the results of a multiple regression with FXI as the dependent variable and SPY and CCB as independent variables. Please note: The T-Stat of the intercept suggests this model is missing one or more factors. At least two other global events occurred over this period affecting FXI: 1) the European debt crisis; 2) the Russian invasion of Crimea. A four-factor model (not shown) including Europe (NYSEARCA:VGK) and Russia (NYSEARCA:RSX) has explanatory power of 72%, and both factors are statistically significant (T-Stats of 3.92 and 7.3, respectively). Although both of these add to the historical explanatory power, their influences are marginal relative to the two factors listed. Therefore, I leave them out of the model and disregard them in the rest of the analysis.

Source: PopperTech and Yahoo Finance

Correlation Thesis: In this section, I propose that the correlation between FXI and VWO will increase over the next year, because 1) PRC reforms should increase the volatility of FXI; 2) China significantly influences the other emerging economies in VWO.

On May 17, 2014, the National Development and Reform Commission announced China needs faster reforms in nine focus areas. These include investments, market pricing, urbanization, social welfare, and the environment. Then, the State Council stated on June 7 that it plans to send audit teams to inspect the implementation of these policies later in the month. Uncertainty regarding the effectiveness of these changes will likely increase the volatility of China's stock market over the next year.

In addition, China exerts significant influence over the other emerging market countries in VWO. For example, China accounts for a greater percentage of exports than the U.S. for 7 of the next 9 largest countries by weight in VWO (Mexico and India are the exceptions). When combined with China, these constitute 74% of VWO. Therefore, VWO should be more sensitive to FXI than SPY.

Figure 7a and 7b indicate the U.S. influence over the emerging markets is declining, relative to China. Figure 7a shows the correlation between VWO and SPY has trended downward over the past two and half years; whereas Figure 7b shows the correlation between VWO and FXI has stayed relatively stable.

Figure 7 displays the 90-day rolling correlations between (a) VWO and SPY (b) VWO and FXI.

Source: PopperTech and Yahoo Finance

Lastly, if the above correlation thesis holds, then VWO may represent a better way of betting on China than FXI, since its expense ratio is 15 bps, compared to 73 bps for FXI.

Alternative Scenarios: Alternatively, the U.S. Federal Reserve is likely to stop asset purchases over the next year, and may raise interest rates. If these actions significantly affect U.S. consumption, and consequently, U.S. market volatility, then the trends in U.S.-Chinese and U.S.-emerging market correlations are likely to reverse. In this case, the correlations are likely to revert to the levels seen in 2006 to 2013, and VWO may actually increase the risk of a U.S.-focused investment portfolio.

Lastly, if both the U.S. and China steady, then the correlations between VWO, SPY, and FXI will likely decline. As a result, other countries will increase their influence over VWO's performance.

Disclosure: The author is long VWO. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I am long a call option on VWO expiring in 2016.

Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.