Social trading is a novel phenomenon that integrates social media into online trading, forming a social trading platform (STP) that allows participants to communicate and explicitly copy each other's trades in realtime. STPs are governed by a scopic environment, which is characterized by high information transparency and constant reciprocal scrutiny by participants. We categorize participants into two main groups: trade leaders, who execute original trades and refrain from explicit copying, and investors, who solely or partially copy trades. This dissertation focuses on the former. First, we investigate herding behavior using popular metrics developed by Lakonishok, Shleifer, and Vishny (1992) and Frey, Herbst, andWalter (2014). We nd levels of, and persistence in herding behavior that exceed those found in traditional nancial settings. We argue that this is due to the scopic environment governing STPs. Second, we examine the disposition effect of trade leaders. Building on the learning hypothesis discussed by Dhar and Zhu (2006), we propose that traders learn not only from their own trades, but also from the trades of others to adjust for this behavioral bias. We find ample evidence of a weaker disposition for trade leaders on a STP compared to traders on a traditional platform. This suggests that high information transparency erodes this bias. Third, we investigate the predictive ability of trade leaders in 16 currency pairs and three commodities. Using methods similar to Henriksson and Merton (1981) and Fishe and Smith (2012) we find that, although around 50% of traders trade profitably more than half of the time, very few possess the skill to do so in all market conditions. Nevertheless, our findings suggest that the scopic environment yields pro table short-term information that is contained in the order flow. The concluding chapter reviews the main findings of this thesis and discusses potential future work.
展开▼
