Until the 1970s empirical studies generally supported the prediction following from weak-form efficiency: prices cannot be predicted using past prices, and returns are serially uncorrelated.
However inter alia Shiller, Fischer, and Friedman (1984) refuted the conclusion of economic insignificance in the case of negligible autocorrelations by modeling return processes with irrational bubbles where short-horizon returns are not significantly autocorrelated.
Empirically, studies such as Lo and MacKinlay (1988) reject the random walk hypothesis (RWH) for US stocks using the more powerful variance ratio (VR) test.
Similarly, RWH has also been tested for other markets with varying results.
This report aims to examine the Chinese stock market as it stands out insofar that it is segmented by investor type 3 that face identical conditions otherwise.
It also allows to observe consequences of loosening capital controls in terms of return-predictability.
Empirical evidence so far has produced mixed results: for instance Long, Payne, and Feng (1999) do not find significant autocorrelation using a 100-week sample.
On the other hand, Su and Fleisher (1999) reject the RWH when using daily data and for the market for foreign investors when using weekly data. Similarly, Darrat and Zhong (2000) find positive autocorrelation in the composite index using weekly data.
Using daily data until 2000, the RWH is rejected again only for the market for foreigners, but not for locals.
This work will build on empirical studies such as Long, Payne, and Feng (1999), concentrate on the VR test, and include observations following a policy reform in 2011 to test inter alia the hypothesis whether the market for foreigners became more efficient.
Table of Contents
1 Introduction
2 Data and Methodology
3 Results
4 Conclusion
5 References
A List of Previous Studies and Acronyms
B Summary Statistics
C Results
Research Objectives and Key Topics
This report examines the market efficiency of the Chinese stock market by analyzing return predictability for A-shares and B-shares, specifically focusing on the impacts of investor segmentation, capital controls, and policy reforms.
- Analysis of the Random Walk Hypothesis (RWH) for the Chinese stock market.
- Comparison of market efficiency between local (A-shares) and foreign (B-shares) investor segments.
- Evaluation of return predictability using Variance Ratio (VR), Cowles-Jones, Runs, and Box-Pierce tests.
- Assessment of the impact of the 2011 policy reform and SOE reforms on market efficiency.
- Investigation of the relationship between trading volume and return predictability.
Excerpt from the Book
1 Introduction
Until the 1970s empirical studies generally supported the prediction following from weak-form efficiency: prices cannot be predicted using past prices, and returns are serially uncorrelated. However inter alia Shiller, Fischer, and Friedman (1984) refuted the conclusion of economic insignificance in the case of negligible autocorrelations by modeling return processes with irrational bubbles where short-horizon returns are not significantly autocorrelated.
Empirically, studies such as Lo and MacKinlay (1988) reject the random walk hypothesis (RWH) for US stocks using the more powerful variance ratio (VR) test. Similarly, RWH has also been tested for other markets with varying results (see, e.g., Lee, 1992; Urrutia, 1995).
This report aims to examine the Chinese stock market as it stands out insofar that it is segmented by investor type that face identical conditions otherwise (Long, Payne, and Feng, 1999). It also allows to observe consequences of loosening capital controls in terms of return predictability (Rabinovitch, 2011). Empirical evidence so far has produced mixed results: for instance Long, Payne, and Feng (1999) do not find significant autocorrelation using a 100-week sample. On the other hand Su and Fleisher (1999) reject the RWH when using daily data and for the market for foreign investors when using weekly data. Similarly, Darrat and Zhong (2000) find positive autocorrelation in the composite index using weekly data. Using daily data until 2000, the RWH is rejected again only for the market for foreigners, but not for locals (Araújo Lima and Tabak, 2004).
This work will build on empirical studies such as Long, Payne, and Feng (1999), concentrate on the VR test, and include observations following a policy reform in 2011 to test inter alia the hypothesis whether the market for foreigners became more efficient.
Summary of Chapters
1 Introduction: Provides the theoretical background on weak-form efficiency and the Random Walk Hypothesis, and outlines the research objective concerning the Chinese stock market.
2 Data and Methodology: Details the datasets for A- and B-shares, the sampling frequency, and the statistical tests employed to measure market efficiency.
3 Results: Presents the statistical findings of the Cowles-Jones, Runs, Box-Pierce, and Variance Ratio tests applied to weekly and monthly returns.
4 Conclusion: Discusses the empirical results, their consistency with previous literature, and potential explanations for observed market behavior post-reform.
5 References: Lists the academic literature and sources cited throughout the report.
A List of Previous Studies and Acronyms: Provides an overview table of existing studies on Chinese market efficiency and defines technical acronyms used in the text.
B Summary Statistics: Contains descriptive data, graphs, and autocorrelation function tables for the SHASHR and SHBSHR indices.
C Results: Contains comprehensive test summary tables and detailed statistical test results (Cowles-Jones, Runs, Box-Pierce, Variance Ratio).
Keywords
Random Walk Hypothesis, Chinese Stock Market, A-shares, B-shares, Variance Ratio Test, Market Efficiency, Return Predictability, Capital Controls, Autocorrelation, Box-Pierce Test, Cowles-Jones Test, Runs Test, Policy Reform, Financial Econometrics, Investor Segmentation
Frequently Asked Questions
What is the primary focus of this research paper?
The paper investigates the weak-form efficiency of the Chinese stock market by testing the Random Walk Hypothesis for two specific share classes (A-shares and B-shares) which are segmented by investor type.
What are the central themes discussed in this work?
The central themes include the impact of investor segmentation, the consequences of loosening capital controls, the effect of market reforms (such as the 2011 reform), and the relationship between trading volume and return predictability.
What is the main research question?
The study aims to determine whether the Chinese stock market follows a random walk and specifically examines if the market for foreign investors (B-shares) became more efficient following policy reforms.
Which scientific methods are employed?
The author uses several statistical tests for autocorrelation, including the Variance Ratio (VR) test, Cowles-Jones test, Runs test, and the Box-Pierce test to analyze weekly and monthly log-returns.
What is covered in the main body of the paper?
The main body includes a literature review of market efficiency, a detailed methodology description, the presentation of empirical results through various statistical tests, and a concluding discussion regarding the implications of the findings.
Which keywords characterize this paper?
The paper is characterized by terms such as Random Walk Hypothesis, market efficiency, A-shares, B-shares, variance ratio, and return predictability.
How does the author explain the inconsistencies with earlier studies?
The author suggests that inconsistencies with previous research are largely due to the use of different data frequencies, as higher frequency daily data may introduce biases related to nontrading and bid-ask spreads.
What conclusion does the author reach regarding the A-share market?
Contrary to expectations, the author finds the A-share market to be less efficient and notes that for both A- and B-share types, weak-form efficiency appears to have decreased when comparing the first to the second sub-sample periods.
Why was the Variance Ratio test prioritized?
The author prioritized the VR test because it is considered more powerful than the Augmented Dickey-Fuller (ADF) and Box-Pierce (BP) tests, especially when dealing with heteroskedastic samples.
- Quote paper
- Maximiliane Brecht (Author), 2015, Tests of Random Walk Hypothesis. Evidence from China, Munich, GRIN Verlag, https://www.grin.com/document/303724
