宏观经济因素对股票市场收益的协整计量分析
宏观经济因素对股票市场收益的 协整计量分析 内容提要 以经典宏观经济和金融理论为框架，应用协整计量分析，本文分析了代表性宏观变 量与股市收益率之间的关联性，试图为宏观对股市的影响提供一个有意义的探索。 宏观经济和股市的关联一直是个困扰经济学界未解之谜；从 “代表性消费模型”的构建直到今天“股权溢价之谜”也都一直没能在理论上得到真正的解 决。大量的金融研究者从资产定价模型入手加入宏观变量因素，计量了宏观变量对股票 资产的影响，取得了丰富的成果；但在实证中基于资产定价模型的“市场有效性”假说过 于严格，从而使宏观与股市的关联在实证过程中的适用性也成了一个“谜”。中国资本市 场是一个新兴的市场，也是一个有着明显制度转型特征的市场，市场的有效性更弱，宏 观变量对股市的影响也是有限的。然而，随着近年来机构投资人的增加，资本市场化进 程的加快，以及资本市场 “有效性” 的加强，对宏观变量与股市关联的研究开始显现其实际意义了。本文以资产定价模型为 基础，构造了一个实证体系，从总体上利用协整模型对宏观变量和股票市场资产的收益 性进行了关联研究，并得出“不稳定”和有限影响的基本结论；进一步，本文深入到宏观 变量对行业与上市公司绩效影响的微观层面，分析得出了很多符合经济学假说的结论， 为中国宏观变量与股票收益是否关联这一“谜题”提供了实证的探索和理论的解释。 基于中国股票市场弱有效性的特征，本文除了在传统意义上研究了宏观经济与股市 收益率之间的直接关系外，更进一步将宏观经济与股市收益率之间的两个传导环节——宏 观经济对行业和上市公司微观绩效的影响以及上市公司微观绩效对股市收益率的影响—— 隔离开来，并集中分析了前一个环节的作用机制和传导效率，从而使本文的研究有别于 前人的成果。 本文利用协整方法对中国宏观变量与股市收益之间的关联性进行了实证研究，得出 了很多有意义的结论：1、宏观对股市收益影响的有限性和不稳定性。除GDP外，其他各 项宏观经济指标（M2、利率和通货膨胀率）与股市收益率之间的相关性均通不过协整检 验；只有GDP与股市收益率在95%置信区间下显著正相关——对应GDP增长率1个百分点的上 升，股市收益率上升0.1个百分点；但该协整关系在滞后2期以后就消失了，显示在整个 样本期内（1996年初到2004年第三季度）GDP对股市的长期影响并不稳定，模型可能发生 了结构性变化，从一个角度探索了市场的有效性不足。2、宏观变量与上市公司基本面的 数量扩张正相关。本文的实证结果显示上市公司总体的收益规模（主营业务收入增长率 ）与GDP增长率之间存在着显著的正相关关系（99%置信区间）。进一步，通过细化研究 GDP增长率对各产业群收入增长的影响，文章发现，GDP总量增长的好处在各产业群之间 的分享程度是不同的：上游基础工业，如石油、石化、有色金属和钢铁等行业对GDP增长 的敏感度均大于1；而出口型产业和房地产业的敏感度在0.7～0.8之间；基础设施类产业 和农业的敏感度为0.4；相反，消费类产业对GDP总量增长不敏感（对人均收入的增长敏 感）。但由于消费类产业在上市公司中所占的比例很大，接近60%，大于其在国民经济中 所占的比重，因此，对应GDP增长率1个百分点的上升，上市公司主营业务收入增长率上 升幅度小于1（0.76个百分点）。3、宏观变量与上市公司基本面的质量改善在产业群层 面显著正相关。尽管在上市公司整体层面上，本文的实证研究显示找不到一个宏观变量 能够对微观个体的收益质量产生显著影响；但在具体产业群上，本文却发现了不同宏观 变量与不同产业群在收益质量上的显著相关性。不过这种相关性主要体现为宏观变量对 毛利润率的影响，而与ROE之间的相关性不显著——这再次证明了我国投资推动型的经济增 长模式。 综上，本文利用协整模型对中国股市和宏观经济变量进行了实证分析，结果表明中 国股市的成熟度依然较低，市场有效性较弱，但微观层面的传递有效性较强。总之，整 体市场与宏观变量的关联性仍不尽人意，但相信随着市场化改革深入，中国股票市场将 逐步走向有效，宏观对股票的影响也将成为影响股市收益和风险的重要因素，从而在宏 观与微观、符号与实体之间建立一种有效的逻辑对应。只有这样，中国的资本市场才能 服务好经济建设，同时分享中国经济的高速增长！ ABSTRACT In this paper we analyze relationships among selected macroeconomic variables and Chinese stock market basing on the classical financial researching frame, by employing a cointegration Analysis model, try to offer a meaningful grope for the macroscopic affection on the stock market. The correlation between macro-economy and the stock market has been a mystery for years. Since the construction of the Representative Agent Model, the Equity Premium Puzzle has not been solved theoretically. To explain this puzzle many financial researchers add the macroeconomic factors to the capital asset pricing models to calculate the influence of macro variables on the stock assets. The results are affluent. However, the Efficient Market Hypothesis of the capital asset pricing models are too strong in practice, which creates a new puzzle on the application of the correlation between macro-economy and the stock market to the empirical work. As an emerging market with the evident characters of a transition period, the China capital market is weakly efficient and the influence of the macro variables is also limited. Along with the incremental proportion of the institutional investors, the accelerated progress of the marketing reform and the strengthening efficiency of the capital market, the researching work on the puzzle of the correlation between the macro variables and the Chinese stock market has become more and more instructive. In fact, according to this paper, the cointegrated analysis of the correlation between the macro variables and the stock market derives the conclusion that the influence of the macro variables to the stock market is limited and unstable. However, the intensive analysis of the correlations between the macro variables and the micro performance of the industries and companies derived many conclusions consistent with the economic theories. With these two analyses together, the paper constructed an empirical frame for the current study as well as for the further grope on this puzzle. Besides the traditional research on the direct correlation between macro variables and stock market, the paper separated the transmission system into two steps, the impact of the macro variables to the micro performance of real industries and the impact of the micro performance to the stock returns. By focusing on the efficiency of the first step, the paper makes itself distinct with the predecessors. By using the cointegration analysis to study the correlation between macro variables and stock market, the paper has drawn some significant conclusions. 1. The limited and unstable influence of macro-economy on stock market. Except GDP, the correlation between stock market and other macro variables, including M2, interest rates and inflation rate, cannot pass the cointegration test. Only the GDP growth direct correlates with the stock market return under a 95% confidential interval, a 1% increase in GDP growth leading to a 0.1% increase in stock market return. However, such a relationship will vanish after two terms lag, which shows an unstable relationship between the two variables. In other words, during the sample period (from Q1 1996 to Q3 2004), the model may have changed for some reasons, which implies the weak efficiency of the market. 2. The significant correlations between macro variables and the scale expansion of the listing companies. According to the metric analysis of this paper, taking all the listing companies as a whole, there is a significant positive correlation between the revenue growth and the GDP growth (under a 99% confidential interval). Furthermore, the sub studies show that for different industrial groups, the interest share of GDP growth is different: the upstream heavy industries, such as oil, petrochemicals, steel and non- ferrous metal industries, with a sensitivity larger than 1, benefit most from the GDP growth; the export industries and the real estate builders with a sensitivity of about 0.7 take the second place; the infrastructure industries and the agricultural companies, a sensitivity of 0.4, share less; while consuming industrial groups, not sensitive to the gross growth, are sensitive to the per head growth in income. Since the consuming companies play an important role in the stock market, with a proportion of 60% to the total listing companies, lager than the proportion of the consuming industries to the total domestic economy, the listing companies as a whole is not quite sensitive to the GDP growth, a sensitivity of 0.76. 3. The significant correlations between macro variables and the profitability of the listing companies on the industrial layer. Although taking the listing companies as a whole, the author failed to find any significant impacts of the macro variables on the micro profitability, the impacts on the specific industrial groups are significant. However such impacts are reflected more in the form of the gross profit ratios, but still insignificant in the form of ROE, again, implying the investment activities as the major engine for the economic growth in China. In summary, by using the cointegration analysis, the paper studied the correlation between the macro variables and the stock market. The empirical results show that the Chinese stock market is still an immature market, with a low efficiency. However, contrary to the unsatisfactory correlation between the macro-economy and the index return in stock market, the performance of micro-economy is strong, In the future, along with the market reformation, the Chinese stock market is sure to be more efficient. 目 录 1导言：宏观经济因素和股市的关联之“谜” 7 2宏观经济对股市收益影响的理论和协整分析 8 2.1 宏观经济对微观绩效的传递理论 8 2.2 协整方法和变量选择 10 2.3 宏观对股指收益的影响 11 2．3．1变量的单位根检验 13 2．3．2协整检验 14 2．3．3方差分解 16 3上市公司基本面与宏观经济变量 17 3．1 变量选取 17 3. 2 数据处理 17 3．3 实证结果 18 3．3．1收益规模指标与GDP增长率 18 3．3．2收益质量指标与GDP增长率 20 3．3．3收益质量指标与其他宏观经济变量 20 4产业群业绩变动与宏观经济变量 21 4．1 产业群划分 21 4．2 变量选取 21 4．3 实证结果...
宏观经济因素对股票市场收益的协整计量分析