Abstract: The Shanghai-Hong Kong and Shenzhen-Hong Kong Stock Connect mechanisms were launched in November 2014 and December 2016, respectively. This marked a significant breakthrough in the liberalization process of China’s capital market, reducing the barriers for foreign investors to participate in the mainland stock market and attracting more and more foreign investors (that is, investors involved in northbound trading) to invest in the Shanghai and Shenzhen stock markets. The mature investment concept brought by these foreign investors not only helps to improve the structure of mainland investors, but also contributes to the full play of the stock market’s price discovery function and enhancing the overall market efficiency. The existing research mainly focuses on the influence of foreign investors on stock price pricing efficiency, stock price collapse and stock price volatility under the Shanghai-Hong Kong Stock Connect and Shenzhen-Hong Kong Stock Connect mechanisms, but does not investigate whether foreign investors can predict the return on stock market. In fact, compared with mainland investors, foreign investors involved in northbound trading have comparative advantages in information collection, information analysis and processing, and mine more information that has not been reflected in the stock market to obtain a higher return. Therefore, from the perspective of information, this paper studies the ability of foreign investors’ capital movements to predict stock market returns.
   This paper takes the Shanghai Composite Index (SHCI) and Shenzhen Composite Index (SZCI) during the period from November 17, 2015 to December 31, 2020 as the sample, adopts in-sample regression prediction models, out-of-sample R² statistics and encompassing tests to investigate the ability of northbound fund movement (NCM) to forecast future excess returns of stock markets. The empirical results show that the movement of northbound funds can effectively predict excess returns of stock markets in next 2, 5, 10 and 20 trading days, and its predictive power is robust to Granger causality test, different funds movement measurement and different test samples. On this basis, this paper uses a combination of dividend discount model and VAR model to analyze the economic source of NCM’s forecasting ability. The results show that NCM’s forecasting ability mainly comes from foreign investors’ ability to analyze and process future cash flow information. Finally, using the predictive power of NCM, we construct a quantitative timing-strategy. It is found that NCM timing-strategy can have good investment performance in the holding period of next 2, 5, 10 and 20 trading days.
   Compared with previous studies, this paper mainly has the following contributions: First, different from the existing literature, this paper finds that northbound fund has excellent forecasting ability for future stock market returns, and finds out the economic source of NCM’s forecasting ability, which expands the existing research on the predictability of Chinese stock market returns. Second, the existing studies mainly focus on the impact of foreign investors on stock pricing efficiency, stock price crash and stock price volatility under the Shanghai-Hong Kong Stock Connect and Shenzhen-Hong Kong Stock Connect mechanisms. This paper starts from their information mining ability and finds that northbound fund contains additional predictive information that is not available to predictive variables such as margin financing and short selling, which can effectively predict the return of the stock index. Thus, it enriches the research on the economic consequences of foreign investors’ information mining ability. Third, this paper constructs quantitative timing strategies based on northbound funds movement, and finds that these investment strategies can have good investment performance in the holding period of 2, 5, 10 and 20 days, which can provide support for investors to construct quantitative strategies and manage stock market risk.

Key words: northbound funds, return predictability, economic explanation, economic value

摘要： 自“沪港通”和“深港通”开通以来,北向投资者的交易行为始终被市场所紧密地关注,对于未来的市场走势具有重要的引领作用。本文构建了反映北向投资者资金动向的代理指标(NCM),并且采用样本内的回归预测模型和样本外的R²统计量以及全包围测试方法,研究了其对于股票市场收益率的预测能力。实证结果表明,北向资金的动向能够有效地预测2,5,10和20个交易日的股票市场收益率,这种预测能力对格兰杰因果检验及不同的资金动向计量方法和检验样本都具有较好的稳健性。进一步研究发现,NCM的预测能力主要来自参与北向交易的投资者分析和处理未来现金信息的能力。最后,利用这一预测能力构建量化择时策略,发现NCM市场择时策略可以在2,5,10和20个交易日的持有期内具有较好的投资绩效表现。

关键词: 北向资金, 收益可预测性, 经济解释, 经济价值