Intraday order placement and execution in a limit order market: Evidence from the Indonesia stock market

• Published date: 01 June 2021
• Author: Irwan A. Ekaputra,Chunlin Liu,S. Ghon Rhee,Hongchao Zeng
• Date: 01 June 2021
• DOI: http://doi.org/10.1111/irfi.12278

ORIGINAL ARTICLE

Intraday order placement and execution in a limit

order market: Evidence from the Indonesia

stock market

Irwan A. Ekaputra

1

| Chunlin Liu

2

| S. Ghon Rhee

3

| Hongchao Zeng

2

1

Department of Management, Universitas

Indonesia, Depok, Indonesia

2

College of Business, University of Nevada,

Reno, Reno, Nevada

3

Shidler College of Business, University of

Hawaii, Honolulu, Hawaii

Correspondence

Chunlin Liu, College of Business, University of

Nevada, Reno, Reno, Nevada 89557.

Email: liuc@unr.edu

Abstract

Using order and trade information from the Jakarta Stock

Exchange (JSX), we examine the intraday patterns of order

placement and execution. We find that the probability of an

order being executed is the lowest in the first trading hour

even though more than 50% of orders are submitted during

the period. The execution probability increases steadily

along the trading hours and reaches the highest close to the

end of trading. Controlling for order aggressiveness and

other factors that might impact investors' order placement

behavior, we find that orders placed in the morning take

longer to be executed compared to those submitted in the

afternoon. However, the execution cost is relatively lower

for orders of early trading. Our results suggest that choos-

ing the right time to place an order is important in designing

an optimal trading strategy.

KEYWORDS

Indonesian stock market, intraday pattern, order execution, order

submission, trading behavior

JEL CLASSIFICATION

G14; G15

1|INTRODUCTION

Limit orders are essential to the functioning of an order-driven market. In recent years, there is growing interest in

investors' order placement and execution behavior in order-driven markets. Much of the literature has examined

Received: 29 May 2017 Revised: 1 May 2019 Accepted: 29 June 2019

DOI: 10.1111/irfi.12278

© 2019 International Review of Finance Ltd. 2019

404 International Review of Finance. 2021;21:404–429.wileyonlinelibrary.com/journal/irfi

order submission (e.g., Anand, Chakravarty, & Martell, 2005; Bae, Jang, & Park, 2003; Kaniel & Liu, 2006; Menkhoff,

Osler, & Schmeling, 2010; Ranaldo, 2004), order revision and cancelation (e.g., Fong & Liu, 2010; Liu, 2009), and

order execution (e.g., Cho & Nelling, 2000; Garvey & Wu, 2009). Despite the understanding we have gained of these

order activities, we know little about the intraday patterns in investors' order submission and execution behavior in a

limit order market largely due to data limitations. The predominant intuition in the theoretical literature that

informed investors select order type based on the time horizon of the private information they possess, coupled with

empirical research suggesting that investors' order submissions interact with limit order book conditions, naturally

raises questions of whether investors' order decisions are time dependent. Does an investor deliberately submit dif-

ferent types of orders in the morning (or at the market open) versus afternoon in a trading day? Does order execu-

tion exhibit intraday time-dependent patterns? In this study, we fill the gap in the literature by investigating intraday

variations in investors' order placement and execution in a limit order market.

A better understanding of the intraday order placement and execution patterns can help investors reduce limit

order submission risks and enhance the execution quality through designing more effective order submission strate-

gies. Limit order traders face two types of submission risks. Nonexecution risk arises when the prevailing market ask

(bid) price is higher (lower) than the limit bid (ask) price, resulting in the possibility that traders' limit orders will not

be executed. Free trading option risk arises when public information drives up (down) the value of the asset, causing

limit sell (buy) orders to be picked off or traded at a price that incurs losses for limit order traders. Though traders

can reduce these risks by monitoring public information and either revising or canceling their quotes, such monitor-

ing is costly (Liu, 2009). To the extent that intraday patterns in limit order submission and execution may exist,

traders with different preferences for execution probability, execution time, and execution cost can design optimal

trading strategies to meet their objectives. Traders who are more concerned about nonexecution risk can submit

orders during trading hours when the execution probability is the highest, whereas traders who are more concerned

about free trading option risk can submit orders during trading hours when the execution cost is the lowest. This

approach allows traders to trade off order execution probability or cost for reduced limit order submission risks.

Recent empirical studies suggest that traders have varying preferences for when to submit, revise or cancel their

orders over a trading day. Garvey and Wu (2009) find that both order execution time and execution cost exhibit

intraday time-dependent patterns for a sample of Nasdaq stocks. Orders submitted around the open and close take

less time to be filled but midday orders have lower execution costs. They suggest that these patterns are induced by

variations in informed trading levels. Fong and Liu (2010) document that frequencies of limit order revisions and can-

cellations have a U-shaped pattern in the Australian Securities Exchange. Order revision and cancelation activities

are less intense during the lunch hours but increase in the last trading hour. The intraday patterns are closely related

to limit order submission risks. The above research suggests that intraday time-dependent patterns could exist in the

order placement and execution in limit order markets. However, theoretical and empirical literature to date has not

fully investigated the relationship between intraday time and investor's order submission and execution behavior.

This study attempts to extend the literature by focusing on this issue.

Our empirical analysis is based on order and transaction data from the Jakarta Stock Exchange (JSX). We find a

distinct intraday pattern in investors' order submission behavior. We observe that investors' intraday order submis-

sion activity follows a U-shaped pattern even though more than half of the limit orders are submitted in the first

trading hour. We also investigate intraday order execution by computing three representative metrics that highlight:

(a) order execution rate, (b) order execution time, and (c) order execution cost. Our investigation reveals that order

execution also exhibits a particular intraday pattern. We find that orders submitted in the morning hours have the

lowest probability of being filled. The execution probability increases steadily along the trading hours and reaches

the highest level toward the end of the trading session. We find that the time length of order execution steadily

declines from the morning open to the afternoon close, which suggests that orders submitted in the afternoon are

executed faster than those submitted in the morning. We also find that the order execution cost steadily increases

from the morning trading to the close: The lowest (highest) execution cost is observed for the orders submitted in

the first (last) half hour of trading.

EKAPUTRA ET AL.405