Overnight Trading and Price Discovery over the Course of a Trading Day: Evidence from Stock Index Futures in Korea

• Date: 01 June 2016
• Author: Sang Lyong Joo,Junghoon Seon,Ji Soo Lee
• DOI: http://doi.org/10.1111/ajfs.12136
• Published date: 01 June 2016

Overnight Trading and Price Discovery over

the Course of a Trading Day: Evidence from

Stock Index Futures in Korea*

Sang Lyong Joo

College of Business Management, Hong Ik University

Junghoon Seon**

School of Business, Konkuk University

Ji Soo Lee

Department of Business Administration, Konkuk University

Received 17 November 2015; Accepted 25 March 2016

Abstract

This paper investigates the effect of overnight trading on the price discovery process over the

course of a trading day in the Korean stock index futures market. The paper uses Trade and

Quote data on nearest-to-maturity KOSPI 200 futures contracts from 2 January 2009 to 31

March 2011 and finds the following results. We find evidence that overnight trading con-

tributes significantly to price discovery and, as a result, accelerates the process, and improves

the efficiency, of price discovery. We also find evidence that 26.27% of the entire day’s price

discovery occurs during overnight trading by impounding private information as well as pub-

lic information. Altogether, our results provide insights that the futures market serves as a

price discovery vehicle for the cash index where the cash market closes, and that futures

trades during overnight trading play an important role in price discovery, as they aggregate

investors’ private information about the fundamental values of the cash index.

Keywords Price discovery; Market efficiency; Stock index futures; After-hours trading; Asym-

metric information

JEL Classification: G10, G14, G15

1. Introduction

Price discovery is one of the most important functions of futures markets. Futures

markets provide venues for investors to trade an underlying asset that will be

*All authors have contributed significantly, and all authors are in agreement with the content

of the manuscript.

**Corresponding author: Junghoon Seon, School of Business, Konkuk University, 120 Neung-

dong-ro, Gwangjin-gu, Seoul 05029, Korea. Tel: 82-2-450-4099, Fax: 82-2-3436-6610, email:

jhseon@konkuk.ac.kr.

Asia-Pacific Journal of Financial Studies (2016) 45, 463–491 doi:10.1111/ajfs.12136

©2016 Korean Securities Association 463

delivered in the future and, thus, contribute to the discovery of the fundamental

values of the underlying asset (Grossman, 1977). The price discovery process is the

dynamic process by which new information conveyed by the trade is incorporated

into prices (O’Hara, 1995). Prior empirical studies on the lead–lag relationship

between futures and cash prices provide evidence that futures lead cash prices. The

case of stock index futures is analyzed in many previous studies (e.g. Chan, 1992;

Hasbrouck, 2003). Chan (1992) and Hasbrouck (2003) demonstrate that new infor-

mation is first aggregated in the futures market when a stock index futures and its

underlying assets are traded simultaneously.

1

Fleming et al. (1996) show that lower

trading costs in the stock index futures market result in price leadership in the

stock index futures market.

2

To date, however, few studies have considered the price discovery role of a stock

index futures market in the absence of cash trading. This is because there are lim-

ited numbers of proper samples with which to compare price and trading behavior

between a futures market with and without cash trading. Investigation on the issue

needs to be conducted, since trading of stock index futures in the absence of cash

trading might change price and trading behavior in a futures market in a range of

ways. On the one hand, as futures prices reflect news released outside of trading

hours for the cash, the price discovery process might become more efficient and

speedier than previously. On the other hand, as the cash market closes, index arbi-

trage between a futures contract and its underlying asset

3

is limited and, thus, the

function of price discovery in a futures market may be weakened.

This paper is the first to address the issue by exploiting the event of starting

overnight trading in the Korea Stock Price (KOSPI) 200 futures market.

4

The start

1

Chan (1992) investigates the intraday lead–lag relation between the returns of the Major

Market cash index and returns of the Major Market Index futures and S&P 500 futures and

finds that futures lead the cash index to a greater degree. Hasbrouck (2003) studies the extent

to which stock index futures exhibit price leadership for S&P 500 and NASDAQ 100 indexes

and finds that most price discovery takes place through futures.

2

When futures contracts have a liquidity advantage, the transaction costs are smaller for trad-

ing an index futures contract than for trading individual stocks. The leveraged nature of

futures contracts implies that the costs of trading futures are a small fraction of the costs of

trading a similar position in underlying equities. Fleming et al. (1996) document that the

total cost of trading one S&P 500 index futures contract is one-thirteenth the cost of trading

an equivalent portfolio of equities.

3

If interest and dividends are paid continuously at a constant annual rate of rand qrespec-

tively, the stock index futures, the underlying asset of which is a stock index, are priced by

the following non-arbitrage condition, F0¼S0erqðÞT, where F

0

is the futures price and S

0

is

the stock index. If futures prices deviate from the condition, trades by index arbitrageurs

occur until the condition is met. See Hull (2000, pp. 64–65) for details.

4

The Korea Stock Price (KOSPI) 200 futures contract is one of the most actively traded stock

index futures contracts in the world. Its underlying asset is the KOSPI 200 index, which

includes 200 of the largest stocks listed on the Stock Market Division of the Korea Exchange.

S. L. Joo et al.

464 ©2016 Korean Securities Association