Short‐Term Trading Skills of Individuals, Institutions, and Foreigners: A New Approach Based on Relative Performance

• Author: Jong‐Bom Chay,Wonse Kim
• Date: 01 October 2018
• Published date: 01 October 2018
• DOI: http://doi.org/10.1111/ajfs.12234

Short-Term Trading Skills of Individuals,

Institutions, and Foreigners: A New

Approach Based on Relative Performance*

Jong-Bom Chay

SKKU Business School, Sungkyunkwan University, Republic of Korea

Wonse Kim**

Research Institute of Mathematics, Seoul National University, Republic of Korea

Received 26 December 2017; Accepted 24 June 2018

Abstract

This paper evaluates the short-term trading skills of three types of investor: individuals,

institutions, and foreigners. Based on a dataset that identifies the types of sellers and buyers

for all trades, we examine (i) how stock prices move contemporaneously with trading activi-

ties between investor types, and (ii) whether, given such price movements, one type of inves-

tor has superior skills in timing purchases (sales) of stocks over another type of investor.

Using tests based on daily and weekly horizons, we find evidence that individuals have poor

trading skills and provide liquidity to both institutions and foreigners. When domestic insti-

tutions trade with foreigners, we find that foreigners generally show superior trading skills.

Keywords Individual investor; Institutional investor; Foreigner; Short-term trading skills; Liq-

uidity; Market microstructure

JEL Classification: G10, G14

1. Introduction

Recent decades have witnessed phenomenal growth in foreign investors’ participa-

tion in emerging-economy stock markets as these markets have opened their doors

*We would like to thank Hee-Joon Ahn, Heungju Park, an anonymous referee, and Sung

Won Seo (discussant) who provided valuable insights and comments. We are also grateful to

Eun-Hee Choi, Alaudeen Hameed, Joongho Han, Andrew Karolyi, Andy Young Han Kim,

Dahea Kim, Dong-Ha Kim, Heuijung Kim, Young-Sung Kim, Dong-Hoon Lee, Hoo Rock

Lee, Youngjoo Lee, Kwangwoo Park (editor), Kyoung Gook Park, Young Kyu Park, Hyeuk-

Sun Ryu, Youngsuk Shin, Giseok Song, Kyojik Song, Jungwon Suh, and participants at the

2017 Annual Conference on Asia-Pacific Financial Markets for their helpful comments.

**Corresponding author: Department of Mathematics, Research Institute of Mathematics,

Seoul National University, Seoul 08817, Korea. Tel: +82-2-887-7851, email: aquinasws@s-

nu.ac.kr.

Asia-Pacific Journal of Financial Studies (2018) 47, 660–694 doi:10.1111/ajfs.12234

660 ©2018 Korean Securities Association

to foreigners. Considering foreign institutions’ investment expertise and research

support, many studies classify foreign investors as a separate type of investor and

investigate their performance relative to that of domestic investors. It should be

noted, however, that foreign investors are not always at an advantage; they also face

barriers related to geography and linguistic and cultural differences. Empirical find-

ings from prior studies vary across types of sample data and empirical methods.

Some evidence suggests that foreign investors perform better than domestic

investors. Grinblatt and Keloharju (2000) find that foreign investors in Finland

attain superior performance by pursuing momentum strategies. They show that the

momentum-adjusted performance of foreigners is still highly significant. Employing

daily trading data from the Taiwan Stock Exchange, Seasholes (2000) finds that for-

eigners earn abnormal profits by predicting future price movements. Froot and

Ramadorai (2001) exploit data on cross-border portfolio flows from the United

States to 25 countries to reveal that foreign purchases forecast equity prices, even

after controlling for the price impact caused by their purchasing activities.

In contrast, several papers present evidence of foreign investors’ poor performance

compared to that of domestic investors. Brennan and Cao (1997) use data on portfo-

lio flows between the United States and Canada, Germany, Japan, and the United

Kingdom to investigate the correlation between equity flows and returns and conclude

that domestic investors have information advantages over foreign investors in their

domestic markets. They show that investors are likely to buy foreign assets when

returns on foreign assets are high and to sell when the returns are low. Choe et al.

(2005) argue that domestic investors in Korea have an edge over foreign investors,

presenting evidence that foreign investors trade at worse prices than domestic inves-

tors and that prices are likely to move against foreign investors immediately before

they trade intensively. Using transaction data from Indonesia, Dvo

r

ak (2005) finds

that domestic investors enjoy higher profits than foreign investors and that global bro-

kerages’ domestic clients earn higher profits than either local or Asian brokerages.

Agarwal et al. (2009) exploit complete order and transaction data in Indonesia to clas-

sify executed orders into initiated and non-initiated orders and find that foreign inves-

tors trade at worse prices only when they initiate orders. The authors infer that

foreigners’ poor performance is due to their aggressiveness, such as their tendency to

submit orders to initiate trades. It should be noted however that, between the two

opposing sets of empirical results, Kang and Stulz (1997) report no significant differ-

ence between the performance of domestic and foreign investors in Japan.

When we focus on domestic investors only, there is a growing body of literature

that compares the investment performance of domestic individuals and domestic

institutions. Empirical findings generally indicate that individual investors are less

informed than institutions and that individuals provide institutions with liquidity.

1

1

See, for example, Odean (1999), Barber and Odean (2000, 2001, 2002), Dorn et al. (2008),

Hvidkjær (2008), Kaniel et al. (2008), Barber et al. (2009a), Campbell et al. (2009), Linnain-

maa (2010), Kelley and Tetlock (2013), and Stoffman (2014).

Short-Term Trading Skills of Types of Investors

©2018 Korean Securities Association 661

In this line of research, many studies empirically investigate the investment perfor-

mance of individuals and institutional investors in a single country, finding com-

pelling evidence that individual investors perform worse than institutional investors.

For example, Barber and Odean (2000) show that households earn poor net returns

compared to a value-weighted market index after adjusting for trading costs. Barber

et al. (2009b) report that individual investors in Taiw an incur substantial losses

from trades while institutions enjoy considerable gains. Stoffman (2014) uses trans-

action data from Finland to show that prices consistently move in the direction of

institutional trading when institutions and individuals trade with each other, sug-

gesting that individual investors provide liquidity to institutions. However, some

studies report contradicting results.

2

In summary, prior research on investors’ short-term trading performance gener-

ally classifies investors into three types or groups: domestic individuals (i.e. house-

holds), domestic institutions, and foreign institutions.

3

The consensus of extant

empirical studies is that individuals perform worse than domestic institutions.

However, there is mixed evidence regarding the performance of foreign investors

compared to that of domestic individual or institutional investors. In an attempt to

resolve this disparity, we propose a relative performance evaluation approach using

transaction tick data to compare the performance of pairs of investor groups in a

controlled setting on an equal footing. Most studies in this line of research estimate

each group’s performance separately by aggregating all trades investors in each

group made but without considering the type of investor in the counterparty. Such

an approach is likely to be inaccurate in estimating short-term trading skills because

it does not allow researchers to compare performance between two different types

of investors by holding the stocks being traded and the timing of the trades equal.

One way to address this concern is to compare trading skills between two inves -

tor groups by requiring the stocks being traded and the timing of trades to be the

same. An ideal sample, then, is data for the transactions between two groups trad-

ing as counterparties. This paper employs such an ideal sample to estimate relative

performance more accurately in a controlled setting by using a more refined

methodology than that of extant studies. Essentially, we investigate which type of

investor performs better when two types of investors trade with each other as coun-

terparties. Our results will shed light on why we have mixed evidence on the perfor-

mance of foreign investors relative to domestic individual or institutional investors.

In evaluating the trading skills of these three types of investors, our approach

has a major advantage over the typical approach of previous research, which

2

Several studies suggest that some individual investors perform systematically better than

others. See, for example, Sias and Starks (1997), Ivkovic et al. (2008), and Coval et al.

(2002).

3

We use “investor type” and “investor group” interchangeably. We also use “individuals”

interchangeably with “households.” We use “foreigners” to refer to foreign institutions and

use the two expressions interchangeably.

J.-B. Chay and W. Kim

662 ©2018 Korean Securities Association