报告主题：1、Funding Liquidity and Arbitrage Efficacy （陈景智博士）
2、交易行为建模：交易型股价操纵研究（刘珂博士）
时间：2018年4月12日，下午15:00-17:00
地点：yh86银河国际教学楼B310
报告人简介
陈景智博士，2011年毕业于西南财经大学获得金融学学士学位，2012和2017年分别于英国约克大学获得金融数学硕士学位和经济学博士学位，现为英国艾克赛特大学兼职讲师。陈景智博士主要研究领域为行为金融、市场摩擦和金融危机等。
刘珂博士，2004、2007和2010年分别于北京理工大学、清华大学和香港城市大学获得学士、硕士和博士学位，具有飞行器设计、机械工程、商科三个学科的符合背景，有超过 15 年的量化建模及数据分析经验 (多个研究项目)，以及多年各类金融产品交易实战，并具有在多家证券投资公司和私募基金从业经历。主要研究关注于量化建模及数据分析，股票、股指及商品期货、股指衍生工具交易实战，以及交易策略及投资者行为建模。
内容简介：
1、 Funding Liquidity and Arbitrage Efficacy ——In a model where arbitrageurs capitalize on mispricing opportunities subject to endogenous leverage constraint, we explore the arbitrageurs’ funding liquidity (i.e., the ability to obtain funding) and arbitrage efficacy (i.e., the ability to correct mispricing), and their linkage. We define the funding liquidity (arbitrage efficacy) as the marginal leverage (error correction) against one more unit of mispricing, and find that funding liquidity affects the efficacy of arbitrage, and more importantly, the binding funding constraint makes arbitrage inefficacious. To test these predictions, we design an empirical strategy to estimate the implied arbitrage efficacy of the S&P 500 index future arbitrage relationship over 1997-2015, and we find that 1. the implied arbitrage efficacy is significantly associated with other broad measures of funding liquidity; 2. The periods of inefficacious arbitrage coincide with the major market turmoil during the sample period; 3. Innovations in the implied arbitrage efficacy predict market volatility, especially the volatility risk premia; The predictive power is most prominent conditional on the inefficacious arbitrage, which is consistent with the amplification effect under the binding funding constraint as suggested by Brunnermeier and Pedersen (2009)
2、 交易行为建模：交易型股价操纵研究——Ramping tricks of trade-based stock manipulation have evolved greatly in the fight with stricter market regulation, and can be extremely complicated nowadays. Despite the rigidity and soundness, theoretical models proposed in extant literature can hardly be applied directly to real market data, due to their assumptions being far away from reality. On the other hand, empirical studies of ramping manipulation still lack guidance and support from theories that can better reflect ramping details in practice. This paper addresses this gap by constructing a theoretical model that is closely linked to practical detection, in the framework of behavioral finance. Built in the framework of behavioral finance, our manipulation model describes a game in which the manipulator artificially produces certain patterns of price movements to impact trend-following investors’ expectations and trading actions. In real ramping cases, the manipulator is likely to muddle price information with combinations of tricks. As against to the ‘pump and dump’ assumed in literature, we suggest that a typical and complete stock ramping scheme consists of four basic phases (or tricks): Absorbing Shares, Shaking, Pumping, and Releasing Shares. These insights on concrete ramping tricks would enrich academic theories of price manipulation. Further, in order to circumvent the complexity of manipulating behaviors in practice, trend followers’ performance can be developed into an indicator for detecting suspected ramping manipulations.
