|
|
| Journal of Asian Business and Economic Studies |
|
Vol. 29(2)
, June 2022, Page 91-104
|
|
| Forecasting stock price movement: new evidence from a novel hybrid deep learning model |
|
| Yang Zhao & Zhonglu Chen |
DOI: 10.1108/JABES-05-2021-0061
Abstract
Purpose
This study explores whether a new machine learning method can more accurately predict the movement of stock prices.
Design/methodology/approach
This study presents a novel hybrid deep learning model, Residual-CNN-Seq2Seq (RCSNet), to predict the trend of stock price movement. RCSNet integrates the autoregressive integrated moving average (ARIMA) model, convolutional neural network (CNN) and the sequence-to-sequence (Seq2Seq) long–short-term memory (LSTM) model.
Findings
The hybrid model is able to forecast both linear and non-linear time-series component of stock dataset. CNN and Seq2Seq LSTMs can be effectively combined for dynamic modeling of short- and long-term-dependent patterns in non-linear time series forecast. Experimental results show that the proposed model outperforms baseline models on S&P 500 index stock dataset from January 2000 to August 2016.
Originality/value
This study develops the RCSNet hybrid model to tackle the challenge by combining both linear and non-linear models. New evidence has been obtained in predicting the movement of stock market prices.
Keywords
Stock price movement, RCSNet, ARIMA, CNN, LSTM, S&P 500 index
|
|
|
|
The underpricing and long-term performance of Chinese IPOs listed on the Hong Kong exchange
2025, Journal of Asian Business and Economic Studies
More
Abstract
Purpose
This study aims to inform prospective listing firms, investors and regulators of the unique drivers of Chinese initial public offering (IPO) pricing on the Hong Kong Exchange.
Design/methodology/approach
Using a hand-collected IPO dataset, we investigate whether information uncertainty or investor exuberance drives underpricing and Chinese IPOs’ performance from 2002 to 2015, including 114 state-owned enterprises (SOEs).
Findings
Contrasting with the “listing bubble” in the China domestic stock market, generated by the overoptimism of retail investors, we highlight a “placing bubble” among Chinese firms listed in Hong Kong. This is driven by institutional investors’ buoyant demand for Chinese IPO shares, particularly those of SOEs. Chinese listing firms employ discreet earnings management strategies with their working capital accounts to smooth pre-IPO earnings, which becomes apparent to the market only in the long term.
Originality/value
This study is the first to examine the pricing of sought-after Chinese IPOs among international investors, who face various restrictions when investing in the Chinese domestic stock market. Additionally, it is the first study to measure earnings management using hand-collected pre-IPO data in IPO underpricing studies.
Credit spread drivers and cross-country connectedness: a study of emerging economies in Asia
2025, Journal of Asian Business and Economic Studies
More
Abstract
Purpose
While the existing literature lacks a holistic approach to determining credit spreads and is limited to mostly developed countries, this study investigates credit spread determinants and their cross-country connectedness in the context of four emerging economies in Asia by incorporating bonds, market risk, macroeconomic and global factors.
Design/methodology/approach
This study utilizes principal component analysis for dimensionality reduction and variable representation. Furthermore, we employ the dynamic conditional correlation–generalized autoregressive conditional heteroskedasticity model to capture the cross-country credit spread connectedness between the variables.
Findings
The findings indicate that market volatilities are the most significant drivers of credit spreads, while global factors play a moderating role. Furthermore, the results provide compelling evidence of cross-country credit spread connectedness, with China as the primary transmitter and Malaysia as the primary receiver among the selected emerging economies.
Originality/value
This study addresses the limitations of previous research by extending the analysis beyond the commonly studied developed economies and focusing on emerging economies in Asia. It also employs a comprehensive approach to determine credit spread and explores cross-country credit spread connectedness in developing economies, thereby shedding light on financial risks and vulnerabilities within interconnected global financial systems.
Understanding how investors respond to different social responsibility communications: an empirical analysis of Japan
2025, Journal of Asian Business and Economic Studies
More
Abstract
Purpose
This study explores the variance in investor responses to the corporate social responsibility (CSR) performance of firms, as influenced by information sources and investor types.
Design/methodology/approach
This study applies a short-term event study and cross-sectional analysis with unique CSR datasets obtained from newspaper articles and the Dow Jones Sustainability Index.
Findings
Investor reactions are significantly shaped by their sources of information. Individual investors are found to predominantly respond to accessible news announcements, whereas institutional investors show heightened sensitivity to adverse news from both scrutinized sources. Foreign investors, mirroring institutional investors' patterns, uniquely react positively to index additions.
Research limitations/implications
Investors’ assessment of CSR activities varies due to the differing sources of information obtained; further, it is affected by the type of investor.
Practical implications
The findings guide public relation managers in strategizing CSR communication toward diverse investor types. This includes recommending targeted approaches for Japanese individual investors through newspapers and TV, exercising caution in disseminating adverse news to Japanese institutions, and promoting and justifying CSR actions to foreign investors. It underscores the need for a strategic investor relations frameworks that considers accessibility, literacy, and investors' interests.
Originality/value
This study examines the relationship between sources of information for CSR activities and investors’ responses, an area under-represented in the literature. The author uses CSR announcement data, collected from newspapers to make the results more accurate and relevant.
Music sentiment and the stock market in Vietnam
2025, Journal of Asian Business and Economic Studies
More
Abstract
Purpose
By extending Edmans et al.’s (2021) music sentiment measures to the Vietnam market, the authors aim to investigate the impacts of music sentiment on stock market returns and volatility.
Design/methodology/approach
The authors adopted Edmans et al.’s (2021) music-based sentiment to proxy for investor mood. The current study uses linear regression analysis.
Findings
The authors find that music sentiment is significantly and positively related to both stock returns and stock market volatility. The authors also show that music sentiment has a contagious effect: Global music sentiment and those in the United States, France and Hong Kong are significant drivers of the Vietnamese stock market. The authors also examine the effect on different industry returns and find that returns on stocks of firms in the communication services, consumer discretionary, consumer staples, energy, financials, healthcare, real-estate, information technology and utility sectors are significantly related to music sentiment. In addition to valence, the authors find that other Spotify audio features can be used to quantify music sentiment.
Originality/value
This study contributes to the behavioral finance literature that focuses on investor sentiment. The authors address this topic in Vietnam using high-frequency data.
Earnings forecast disclosures and oversubscription rates of fixed-price initial public offerings (IPOs): the case of Malaysia
2025, Journal of Asian Business and Economic Studies
More
Abstract
Purpose
The main purpose of this study is to examine the disclosure of earnings forecasts in firms' prospectuses to explain investor demands or, in other words, oversubscription rates of Malaysian initial public offerings (IPOs).
Design/methodology/approach
Ordinary least squares and robust methods were used to examine cross-sectional data comprising 466 fixed-price IPOs reported for the period from January 2000 to February 2020 on Bursa Malaysia.
Findings
The results showed that IPOs with earnings forecasts obtained higher oversubscription rates than those without earnings forecasts. IPOs with earnings forecasts provide value-relevant signals to prospective investors about the good prospects of firms, resulting in an increase in the demand for IPO shares. For the IPO samples listed during the global financial crisis (GFC) period, IPOs with earnings forecasts had negative impacts on the oversubscription rates. These results were robust to quantile methods and the two-stage least squares method.
Research limitations/implications
The research findings provide fresh information for investors regarding the importance of earnings forecasts as a trustworthy signal of a firm’s quality when making share subscription decisions.
Practical implications
The regulator is advised to encourage issuers to include earnings forecasts in their prospectuses since such forecasts help to increase the demand for IPOs.
Originality/value
This study contributes to the literature by offering empirical evidence regarding the signalling impact of earnings forecast disclosures on investor demands for Malaysian IPOs. Moreover, this study provides evidence demonstrating the impact of earnings forecast disclosures on oversubscription rates of Malaysian IPOs during the GFC period.
|