二、主讲人：王靖一，北京大学国家发展研究院博士生。主要研究领域为金融科技、情绪与市场、机器学习。研究成果发表在《经济学（季刊）》、《金融研究》、China Economic Journal等国内外著名期刊上。
Abstract: How does news sentiment affect behaviour in a fast-growing innovative segment of the Chinese financial market? This paper addresses this question by making two major efforts. First, employing natural language processing and deep learning techniques, we construct the “China FinTech Sentiment Indices (CFSIs)”, making use of over 17 million Chinese language news articles published between January 2013 and August 2017. The CFSIs cover the dimensions’ attention, positive sentiment and negative sentiment. And, second, we examine impacts of such news sentiment on China’s fast growing and practically unregulated Peer-to-Peer (P2P) lending market. We find that positive sentiment tends to raise trading volume and negative sentiment tends to reduce it asymmetrically. Additionally, we also discover that P2P trading volumes respond positively to the interest rates granted to lenders and negatively to the market liquidity conditions. Hence, basic market mechanisms appear to function normally in this nascent financial market segment.