WHAT DRIVES INVESTMENT FLOWS INTO SOCIAL TRADING PORTFOLIOS?

• Author: Florian Röder,Andreas Walter
• DOI: http://doi.org/10.1111/jfir.12174
• Date: 01 July 2019
• Published date: 01 July 2019

The Journal of Financial Research Vol. XLII, No. 2 Pages 383–411 Summer 2019

DOI: 10.1111/jfir.12174

WHAT DRIVES INVESTMENT FLOWS INTO SOCIAL TRADING

PORTFOLIOS?

Florian Röder and Andreas Walter

University of Giessen

Abstract

We investigate investment flows into more than 5,300 socialtrading portfolios that are

issued as structured products and are tradable at a regular exchange. We find that

investment flows chase past performance. However, in contrast to mutual fund flows,

the flow–performance relation exists nearly exclusively for the best performing social

trading portfolios. Flows follow raw returns rather than factor model alphas.

Additionally, flows are highly persistent. Finally, social trading portfolios with higher

visibility on the web page of the social trading platform as well as traders

communicating actively to investors via public comments attract higher inflows.

JEL Classification: D14, G11, G23

I. Introduction

More than 2 billion people actively use social media services (Kemp 2016). However,

social media affects not only our everyday life but also financial markets (Mao,

Counts, and Bollen 2015) as hundreds of thousands of investors use specific social

media platforms to discuss investment decisions and introduce their investment

strategies to other investors (Chen et al. 2014).

We investigate a social trading

1

platform that not only allows traders to

communicate their investment ideas but also provides investors the opportunity to

participate in these strategies by investing in exchange‐traded structured products. This

growing platform offers access to both institutional portfolio managers and amateur

traders without any barriers to entry. The basic idea of this form of financial technology

(FinTech) is comparable to mutual fund investment. Investors commit their money to a

trader, expecting her to generate positive (adjusted) returns. However, although mutual

fund managers are always professionals, most social traders on our platform are not.

Additionally, mutual fund managers differ from social traders with regard to

regulations, interactions between participants, and sales networks.

In this study, we explore the determinants ofinvestment flows into social trading

portfolios and analyze whether these determinants differ from those of mutual fund

flows. We define a “social trading portfolio”as a virtual portfolio managed by a virtual

1

Note that social trading is not a synonym for social (responsible) investing. Social trading is characterized

by a combination of a social network and a trading platform, allowing users to follow the investment ideas of

other traders.

383

© 2019 The Southern Finance Association and the Southwestern Finance Association

portfolio manager (“trader”) and a corresp onding structure d product (“certificate”). The

financial services provider that collaborates with the social trading platform can issue a

certificate that replicates the performance of the virtualportfolio. This allows investors to

indirectly invest real money in the virtual portfolio.

Although Hirshleifer (2015) postulates that it is time to move from behavioral

finance to social finance, it is not clear to what extent the novel possibilities of social

interaction change investor behavior. Investigations of the mutual fund industry show

that mutual fund flows are mainly determined by past fund performance (Ferreira et al.

2012; Huang, Wei, and Yan 2007; Sirri and Tufano 1998) and that investment flows

are most sensitive to capital asset pricing model (CAPM) alphas (Barber, Huang, and

Odean 2016). We investigate whether the social network component of social trading

induces different investor decisions and whether social‐trading‐specific variables play

a decisive role for investment flows.

We analyze data from a European social trading platform from June 2012 to

June 2016. The sample contains data on more than 5,300 investable social trading

portfolios. Theportfolios were issued as structured products that are tradable on a regular

European exchange. Consequently, investors not registered on the social trading platform

also have the opportunity to invest in these certificates. During our study period,

investment flows into and out of these certificates total more than €955 million.

Hirshleifer (2015) states there is a need for research into social finance topics. We

contribute to this gap in several ways. First, we examine whether past portfolio performance

determines social trading investment flows and analyze to which performance measure

investment flows are most sensitive. Second, we investigate whether communication on the

social trading platform influences investment flows by analyzing the relation between

investment flows and comments written by traders. Third, we explore whether the ranking of

social trading portfolios on the web page of the social trading platform significantly affects

investment decisions. Finally, we examine whether institutionally managed social trading

portfolios receive more inflows than those managed by amateurs.

We explore the determinants of social trading investment flows by using a

piecewise linear regression framework following Fama and MacBeth (1973). We show

that investment flows are momentum driven as investors chase past performance. This

result is in line with the mutual fund flow literature (e.g., Ferreira et al. 2012; Huang,

Wei, and Yan 2007; Sirri and Tufano 1998). However, this positive flow–performance

relation is valid only for the top‐performing social trading portfolios. For mutual funds,

on the contrary, most authors find a positive and mostly significant flow–performance

relation among medium performers as well (e.g., Ferreira et al. 2012; Huang, Wei, and

Yan 2007; Sirri & Tufano 1998).

We find that social trading investors rely on raw returns rather than on the

Jensen (1968) one‐factor alpha, Fama–French (1993) three‐factor alpha, or Carhart

(1997) four‐factor alpha when evaluating past returns. In contrast, mutual fund flows

are most sensitive to one‐factor alphas (Barber, Huang, and Odean 2016). We argue

that social trading investors may follow raw returns because social trading portfolios

lack concrete benchmarks. Nevertheless, the difference between the determinants of

384 The Journal of Financial Research