股票，金融資產和市場影響的價格信息實證分析：
實證解析自計量經濟學面板數據和人工智能模型

在瞬息萬變的交易市場中，信息化一直是投資者和學者關注的焦點。隨著越來越多的投資者進入市場，並越來越強調定期進行公司治理，這使得新聞所喚起的股票信息和投資者情緒對研究變得越來越有價值。第一篇研究遵循 Roll (1988)的方法，調查股票價格信息是否影響債務比率並形成對台灣股票市場的監督作用。根據 Morck, Yeung, and Yu (2000)的研究中，台灣市場在價格同步性中排名前三位，因此我們以台灣市場為背景，以價格信息來檢驗債務比率，以檢查價格的透明變化和市場效率。結果表明，更高的透明度和開放性對市場具有積極意義。在第二篇中，我們直接使用人工智能從公共互聯網網站捕獲新聞進行自然語言處理，以比特幣為標地證明順序信息到達假設及混合分佈假設。簡言之、本研究的創新貢獻是利用文字探勘及大數據方法取得新聞文章資料在導入人工資慧模型計算及量化新聞情緒數據取代交易量，使得研究可以用真正的消息面指標數據以便探討其與報酬波動度的關係，從而深入探討比較兩個財金學說假設 SIAH 及 MDH) 在比特幣的預估可行性。這樣的做法可以彌補過去研究缺乏消息面指標數據的文獻缺口為投資人及政策制定者及學術界提供更大的啟發及未來研究的作用。

關鍵字

價格訊息性；文字探勘；機器學習

並列摘要

In the change rapidly trading market, informativeness has always been attention by investors and academics. As more and more investors engage the market and form more regularly highlight corporate governance, it takes the stock information, and investor sentiment invoked by news becomes more valuable to research. The first part, This research follows Roll (1988) to investigate whether existing stock price informativeness affects the debt rate and forms a supervision effect in the Taiwan stock market. According to Morck et al. (2000) study, Taiwan market rank prior three number in price synchronize; hence we use the sample of Taiwan market as context to test price information with debt rate for checking transparent price change and market efficiency. As a result, the result shows that more transparency and openness have positive significance in the market. In the second part, we directly use artificial intelligence to capture news from public internet websites for natural language processing and use Bitcoin as an underlying to prove the iii hypothesis of sequential information arrival and the hypothesis of mixed distribution. In short, the innovative contribution of this research is to use Text mining and Big data methods to obtain news article data, introduce Artificial intelligence model calculations and quantify news sentiment data to replace transaction volume. Therefore, this research can use actual news indicator data to discuss the relationship between it and the volatility of returns to compare the estimated feasibility of the two financial hypotheses (SIAH and MDH) in Bitcoin. This approach can make up for the lack of indicator data of the information field in the past research literature and provide investors, policymakers, and academia with greater enlightenment and the role of future research.

並列關鍵字

Price Informativeness ； Text Mining ； Machine Learning

參考文獻

Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3-56.

Part 1

Google Scholar

Abubakar, A. (2015). Relationship between financial leverage and financial performance of deposit money banks in Nigeria. International Journal of Economics, Commerce and Management, 3(10), 759-778.