Kin-Yip Ho

Predicting Stock Price Movements with News Sentiment: An Artificial Neural Network Approach

Behavioural finance suggests that emotions, moods and sentiments in response to news play a significant role in the decision-making process of investors. In particular, research in behavioural finance apparently indicates that news sentiment is significantly related to stock price movements. Using news sentiment analytics from the unique database RavenPack Dow Jones News Analytics, this study develops an Artificial Neural Network (ANN) model to predict the stock price movements of Google Inc. (NASDAQ:GOOG) and test its potential profitability with out-of-sample prediction.

Volatility and Correlation Dynamics of the Mainland Chinese and Hong Kong Stock Markets: Evidence from the A-, B-, H- and Red Chip Markets

This article examines the volatility dynamics of the mainland Chinese stock markets (Shanghai A- and B-shares, Shenzhen A- and B-shares, H-shares, and red chips) by employing the daily returns data from 1993 to 2012 and a multivariate volatility framework that incorporates the features of asymmetries, persistence, and time-varying correlations, which are typically observed in stock markets of developed economies. Our results indicate that, unlike the Shenzhen and Shanghai A-share markets, the B-share, H-share and red chip markets do not exhibit significant asymmetric volatility (leverage effect). Furthermore, return volatility in the A-share markets is relatively more volatile than the others before 1997 and becomes more stable afterward. In addition, there is strong evidence of volatility persistence in all the markets, and this finding is robust to changes in model specification. The Chinese stock markets apparently share a common degree of persistence (fractional integration) in volatility. Moreover, the conditional correlations are significantly time-varying and are strengthening over time, especially after 2002. These findings have important implications for hedging and portfolio management and diversification.

VOLATILITY DYNAMICS IN FOREIGN EXCHANGE RATES: FURTHER EVIDENCE FROM THE MALAYSIAN RINGGIT AND SINGAPORE DOLLAR

The evolution of volatility and correlation patterns of the Malaysian ringgit (MYR) and the Singapore dollar (SGD) are analyzed in this paper. Our approach can simultaneously capture the empirical regularities of persistent and asymmetric effects in volatility and time-varying correlations of financial time series. Consistent with the results of (1997), there is only some weak support for asymmetric volatility in the case of the MYR when the two currencies are measured against the US dollar (USD). However, there is strong evidence that depreciation shocks have a greater impact on future volatility levels compared with appreciation shocks of the same magnitude when both currencies measured against the yen. Moreover, evidence of time-varying correlation is highly significant when both currencies are measured against the yen. Regardless of the choice of the numeraire currency and the volatility models, shocks to exchange rate volatility are found to be significantly persistent.

It Takes Two to Tango: A Regime‐Switching Analysis of the Correlation Dynamics between the Mainland Chinese and Hong Kong Stock Markets

This paper proposes a regime‐switching model to examine the correlation dynamics of the mainland Chinese and Hong Kong stock markets with high‐frequency A‐, B‐, H‐shares and Red Chip indexes. We find significant evidence of volatility persistence and asymmetries in these markets. Our model further suggests all correlations are significantly time‐varying with various patterns and co‐persistence in both low‐ and high‐correlation states. Our findings have important implications for both policymakers and investors, such as understanding the extent and nature of integration between the mainland Chinese and Hong Kong stock markets over time and developing dynamic strategies for optimal hedging and portfolio management.

High-Frequency News Flow and States of Asset Volatility

This chapter examines the relationship between high-frequency news flow and the states of asset return volatility. To estimate asset return volatility and smoothing probability, we first apply the Markov Regime-Switching Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model. Second, the different states of asset return volatility are identified by comparing the previously generated smoothing probability with certain thresholds. Subsequently, we employ discrete choice models to investigate the impact of high-frequency news flow on the volatility states of hourly returns of the constituent stocks in the Dow Jones Composite Average (DJN 65). Our dataset for high-frequency news flows is constructed from the new RavenPack Dow Jones News Analytics database that captures >1200 types of firm-specific and macroeconomic news releases at high frequencies. Estimated results show that different types of news flows have varying significant effects on the likelihood of volatility states of intraday asset returns.

Price Discovery and Dynamic Correlations: The Case of the Chinese Renminbi Markets

The Chinese renminbi (RMB) currency system has undergone several major developments in the past two decades, including the adoption of a managed-floating system since July 2005, the expansion of an offshore RMB non-deliverable forward (NDF) market, and the relaxation of certain regulatory controls to promote the increased use of the RMB in the region. These developments have sparked intense debate on the potential of RMB internationalization and its pros and cons. In view of these developments, this paper provides a comprehensive analysis of the dynamics of the RMB in both the spot and NDF markets since 2005 by examining the role and significance of the RMB NDF by quantifying the contributions of the RMB spot and NDF rates in the price discovery process and the volatility dynamics of the RMB markets by adopting two different frameworks with multivariate Student’s t-distribution and time-varying conditional correlations. The results indicate that asymmetric volatility effects are significant for several NDF contract maturities and the spot-NDF correlations are significantly time-varying. Moreover, shocks to the volatility levels are highly persistent. Causality tests on the spot and NDF volatilities further suggest that the NDF markets impact the future fluctuations of the spot market, but the spot market does not have predictive power for the volatility of the NDF markets.

Foreign Trade and Economic Growth in Recent China: A Granger Causality Analysis

Many studies have investigated the relations between China’s foreign trade and economic growth. Most of these studies found bidirectional causality between trade and economic growth. In this study, we analyze the causal relationship between foreign trade and economic growth in China using quarterly data from 2005 to 2013. The empirical results show evidence of unidirectional causality between China’s GDP and China’s export. Our results are different with most existing literature and show the new economic growth pattern in China.

Public news flows and Chinese renminbi: A volatility regime-switching analysis

In recent years the Chinese foreign exchange rate policy and the internationalization of Chinese renminbi (RMB) have been hotly debated both in the academic world and in practice as the growing weight of China’s foreign trade and investment flows in the world economy. Most of the existing studies tend to review the process and impact of the RMB internationalization issues. The great interest of RMB internationalization is probably with financial institutions doing international business in the currency and also the Chinese firms conducting international business. This paper examines the impact of public information flows on the volatility of the Chinese RMB spot and non-deliverable forward (NDF) exchange rates using a regime-switching framework. Two major types of news items – articles that report on the RMB and those that report on the US dollar (USD) – are used to measure information flows. The results indicate that the impact of news flows is more significant in the calm (low volatility) regime rather than in the turbulent (high volatility) regime. These findings have important implications for policy-makers in both the public and private sectors and for business firms who wish to hedge their exposures.

News Sentiment and High-Frequency Volatility Dynamics in the Japanese stock Market

This chapter investigates the impact of high-frequency public news sentiment on intraday return volatility of the constituent stocks in the Tokyo Stock Exchange over the period from January 2000 to December 2012. By using textual and linguistic analytical techniques, we compute the various sentiment scores for all the intraday firm-specific news releases obtained from Dow Jones Newswires . The results show that intraday volatility persistence is significantly reduced after incorporating the effects of firm-specific news releases and their sentiment scores. Compared with positive news, the impact of news releases with negative sentiment on future intraday volatility levels is higher. These findings highlight the importance of public news sentiment on examining high-frequency volatility dynamics.

What Drives the Time-Varying Performance of Japanese Mutual Funds?

In this chapter, we review the performance of Japanese mutual funds with the most recent data and examine the time-varying volatility and the leverage effect of Japanese mutual funds over the business cycle by using a Markov Regime-Switching GARCH model. The results suggest that volatility persistence of Japanese mutual funds are generally quite large and vary significantly with their business cycles. Moreover, the significant leverage effects of shocks on volatility are observed, and positive shocks generally have greater positive effects than negative shocks. Also, we find that contemporary news sentiment and flow can reduce a considerable proportion of the volatility persistence. The effects are different, depending on the states of business cycle. Finally, the marginal effects of negative and positive news on volatility are roughly symmetric in both states of business cycle. All the results are robust when mutual funds are modeled within the proxied global business cycle. Our results have important implications for investors seeking opportunity of portfolio diversification.