This paper proposes a new forecasting framework for the stock market that combines machine learning algorithms with several technical analyses. The paper considers three different algorithms: the Random Forests (RF), the Gradient-boosted Trees (GBT), and the Deep Neural Networks (DNN), and performs forecasting tasks and statistical arbitrage strategies. The portfolio weight optimization strategy is also proposed to capture the model's return and risk information from output probabilities. The paper then uses the stock data in the Chinese A-share market from January 1, 2011, to December 31, 2020, and observes that all three machine learning models achieve significant returns in the Chinese stock market. The DNN achieves an average daily return of 0.78% before transaction costs, outperforming the 0.58% of the RF and 0.48% of the GBT, far exceeding the general market level. The performance of the weighted portfolio based on the ESG score is also improved in all three machine learning strategies compared to the equally weighted portfolio. These results help bridge the gap between academic research and professional investments and offer practical implications for financial asset pricing modelling and corporate investment decisions.

A deep learning method is applied to predict stock portfolio allocation in the Chinese stock market. We use 6 original price and volume series as benchmark model settings and further explore the model's predictive performance with social media sentiment. Our results show that our model can achieve a high out-of-sample Sharp ratio and annual return. Moreover, social media sentiment could increase the performance for both Sharp ratio and annual return while reducing annual volatility. We provide an end-to-end stock portfolio allocation model based on deep neural networks.

本文以2009-2020年针对我国A股上市公司的485,366份分析师报告为研究样本，考察了分析师利空关注对公司投资效率的影响。研究表明：第一，分析师的利空关注会同时抑制过度投资和缓解投资不足，对公司投资效率具有显著的提升作用。第二，当公司面临较大的卖空压力时，或分析师能力和努力程度更高时，分析师利空关注对公司投资效率的提升作用更突出。第三，基于双重纠偏LASSO（DoublyDebiasedLASSO）回归对影响机制的遗漏变量控制后，发现分析师利空关注对投资效率的积极影响同时存在卖空压力的“大棒效应”以及信息增强的“萝卜效应”。第四，基于因果路径分析法，对多个机制间相互重要性和关联性检验后发现分析师利空关注的“萝卜效应”要强于“大棒效应”，而机构投资者的直接作用要强于资本市场的间接作用。本文为探讨分析师在资本市场中的角色提供了新视角，补充了中国情境下分析师关注影响公司投资效率的经验证据。

现有研究发现，中国证券市场存在盈利季节性的股市异象：处于盈利淡季的股票在盈余公告期间会比处于旺季的股票获得更大的累计超额收益。如果管理层在盈余公告前发布业绩预告，能否有效地抑制投资者的非理性预期、降低盈余公告期间盈利季节性的股市异象？本文以2010至2020年A股上市公司为样本，分析发现：首先，A股市场在业绩预告期间也存在盈利季节性的股市异象。处于盈利淡季的公司在业绩预告期间比处于盈利旺季的公司平均可多获得2.1%的累计超额收益；其次，盈余公告和业绩预告期间，处于盈利淡季的公司比处于盈利旺季的公司有更大的股价波动率和股票交易量。第三，在盈余公告前发布业绩预告，能显著抑制盈余公告期间盈利季节性对收益率、股价波动率和股票交易量的预测能力。第四，自愿性业绩预告、与盈余公告的间隔时间越短的业绩预告抑制作用更明显。在更换盈利、盈利季节性、累计超额收益测度指标等一系列稳健性检验后，结论依然存在。表明管理层业绩预告能有效地降低投资者的非理性预期，提高股票定价效率。研究结果对市场建设和投资者决策均有一定的借鉴意义。

We examine the roles of bank ownership and CEO political faction membership in facilitating or hindering the implementation of central bank policy in China. Specifically, we examine the response of China’s commercial banks to People’s Bank of China (PBC) guidelines intended to decrease mortgage lending and to slow down the rise in residential property prices. We find that both bank ownership and faction membership matter. Central government-owned banks whose CEOs are members of the specialist finance faction within the Chinese Communist Party (CCP) respond most strongly to PBC guidance, whereas provincial or city government-owned banks whose CEOs are members of a generalist faction respond least strongly. The implementation of PBC policy has real effects: in those cities where central government-owned banks with specialist CEOs constitute a larger percentage of total bank branches, house prices grew more slowly, as did the number of residential real estate transactions and the number of new listings. Where in contrast provincial and city government-owned banks with generalist CEOs dominate, the number of transactions grew faster; the rate of house price appreciation and the number of listings were however unaffected. We conclude that China’s different levels of government and the CCP’s different factions enjoy some discretion in responding to PBC guidance and that they exploit the discretion they are afforded to vary the strength of their response.

This paper explores the relationship between corporate social responsibility (CSR) and timeliness of goodwill impairment. Goodwill is the premium that is paid when a business is acquired. If the value of the business declines, goodwill impairment occurs. Deliberately delaying goodwill impairment (timeliness) is a widespread ethical issue. Based on all the mergers of Chinese listed companies during 2010–2019, we study the motivation of corporate charitable donations when facing the risk of goodwill impairment. Our results suggest that long-term (consistent) charitable donations reflect more altruist social responsibility than short-term (suddenly increased) donations. In particular, firms that make more long-term donations tend to report goodwill impairment timely, while firms making excessive short-term donations are more likely to delay goodwill impairment. Furthermore, we find that short-term donation is motivated not only to cover up the goodwill impairment delay, but also to provide insurance-like protection when delayed impairment is announced. Our results also suggest that moral licensing plays a role in inducing such opportunistic behaviors. To address the endogeneity problem, we use the number of provincial charitable funds and the number of provincial deaths due to natural disasters as instrumental variables for short-term excessive donations.

We study the potential negative externalities of mandatory environmental, social, and governance (ESG) disclosure. Our analysis exploits a unique regulatory change in China that requires a subset of firms to report their contributions to poverty alleviation—on top of reporting general ESG issues—using a difference-in-differences design. We find that treated firms significantly increase their anti-poverty spending, but also increase their pollution, after the regulatory change came into force. The negative environmental externality is more concentrated in firms that are more financially constrained, as well as firms that are facing fiercer market competition. We further show that this effect is driven by a firm’s incentive to strategically cater to politicians’ agenda in order to obtain preferential treatment. These findings suggest that mandating ESG disclosure in selected areas may induce firms to trade off different ESG goals by prioritizing more conspicuous ESG issues at the cost of trivializing other, longer-term, issues.

Using both the onset of the US-China trade war in 2018 and the most recent Russia-Ukraine conflict and associated trade tensions, we show that government-linked firms increase their importing activity by roughly 33% (t=4.01) following the shock, while non-government linked firms trading to the same countries do the opposite, decreasing activity. These increases appear targeted, in that we see no increase for government-linked supplier firms generally to other countries (even countries in the same regions) at the same time, nor of these same firms in these regions at other times of no tension. In terms of mechanism, government supplier-linked firms are nearly twice as likely to receive tariff exemptions as equivalent firms doing trade in the region who are not also government suppliers. More broadly, these effects are increasing in level of government connection. For example, firms that are geographically closer to the agencies to which they supply increase their imports more acutely. Using micro-level data, we find that government supplying firms that recruit more employees with past government work experience also increase their importing activity more – particularly when the past employee worked in a contracting role. Lastly, we find evidence that this results in sizable accrued benefits in terms of firm-level profitability, market share gains, and outsized stock returns.

We decompose the difference between a firm’s market and book values into two components: intangible assets that can be created by competing firms through SG&A/R&D expenditures, and the residual denoted as franchise value (FV). The estimated parameters in the model for creating intangible assets by capitalizing R&D/SG&A expenditures vary significantly across industries. Consistent with FV being a measure of economic rents and quasi-rents, ceteris paribus, higher FV firms face fewer product market threats, have higher markups, and their investments are less sensitive to their total Tobin’s Q. In contrast, firms with higher capitalized intangible assets, face higher product market threats.

Using a unique list of 528 fake state-owned enterprises (SOEs) exposed in China, we examine whether and how investors react to the government’s property rights protection actions. Our results show that real SOEs with more subsidiaries, pyramid layers, and popularity are more likely to be targeted by wrongdoers. We find that when fake SOEs were exposed, it caused a significant increase in the stock prices of listed central SOEs controlled by the State Council. Further analysis shows that the stock price rise is driven by both the cash flow and risk effects. We also find that the value impact of the crackdown is more pronounced for listed central SOEs with less media coverage, located in weaker legal protection regions, and facing more competition. Overall, our findings provide empirical support for the effectiveness of exposure, as a non-litigation channel of property rights protection, in enhancing firm value.