Does trade size restriction affect trading behavior? Evidence from Indian single stock futures market

• Published date: 01 March 2020
• Author: Ashok Banerjee,Anirban Banerjee
• Date: 01 March 2020
• DOI: http://doi.org/10.1002/fut.22073

J Futures Markets. 2020;40:355–373. wileyonlinelibrary.com/journal/fut © 2019 Wiley Periodicals, Inc.

|

355

Received: 20 May 2018

|

Accepted: 22 October 2019

DOI: 10.1002/fut.22073

RESEARCH ARTICLE

Does trade size restriction affect trading behavior?

Evidence from Indian single stock futures market

Anirban Banerjee

1

|

Ashok Banerjee

2

1

Finance, Accounting and Control Group,

Indian Institute of Management

Kozhikode, Kozhikode, India

2

Finance and Control Group, Indian

Institute of Management Calcutta,

Kolkata, India

Correspondence

Anirban Banerjee, Finance, Accounting

and Control Group, Indian Institute of

Management Kozhikode, Kozhikode,

Kerala 673570, India.

Email: anirban@iimk.ac.in

Abstract

Algorithmic traders use their advantage of speed to execute a large number of

small‐sized trades in a very short time. In the presence of a minimum trading

unit (MTU) restriction, they are forced to trade at the smallest possible sizes,

often restricted by the MTU. Using a novel data set of single stock futures

market obtained from the National Stock Exchange of India, we show that the

MTU restriction acts as a binding constraint for traders while optimizing trade

sizes. Contrary to expectation, we find weak evidence that liquidity is positively

impacted by the contract size revision.

KEYWORDS

algorithmic trading, contract size, liquidity, minimum trading unit, trading volume

1

|

INTRODUCTION

This paper provides empirical evidence, perhaps for the first time, of how a market‐wide upward revision of

minimum trading unit (MTU)

1

impacts various trader groups. We find that high‐frequency traders are not

affected by this revision. Exchanges all over the world often impose a restriction on the minimum size of a trade.

This restriction can be either imposed by explicitly specifying the MTU for a security or by specifying the

minimum size of a traded contract. Specification of MTU o r minimumcontract size is an intriguing question as it

requires a trade‐off between transaction cost and volume (Karagozoglu & Martell, 1999). While lower contract

size can increase the transaction cost, higher contract size may force out certain market participants. In the

absence of any trade size restrictions, traders are faced with a task to determine the optimum trade size that

balances impact cost and transaction cost. Presence of trade size restriction can, however, significantly impact

this optimization exercise. Our objective in this paper is twofold. First, we try to observe how the ordering and

trading behavior of different groups of traders, more specifically algorithmic traders, is impacted by MTU

restrictions. Our second objective is to observe how a market‐wide upward revision of MTU impacts the market

quality parameters. Existing literature on trade size restriction primarily focuses on the reduction of lot size or

MTU (Karagozoglu & Martell, 1999; Karagozoglu, Martell, & Wang, 2003) and its impact on improving liquidity

(primarily in terms of bid‐ask spread), and increase in traded volume. Reduction in MTU (Amihud, Mendelson,

& Uno, 1999; Hauser & Lauterbach, 2003) is also seen to increase the number of individual shareholders and

appreciation in share price, which is consistent with Merton (1987). However, we do not find any existing

literature that studies the response of different trader groups to trade size restriction.

Algorithmic traders gain competitive advantage through their ability to execute a large number of small‐sized trades

in a short time. We base our analysis on the Indian derivatives market for single stock futures (SSF). We utilize a

natural experiment setting of upward revision of MTU as proposed by the Indian capital market regulator Securities and

1

This is often referred to as trading lot size.

Exchange Board of India (SEBI) in 2015, to observe how market participants react to such exogenous shocks. With the

introduction of dematerialized trading in the equity market (1999), most of the trading in the Indian equity market at

present is carried out in the paperless format with no concept of trading lots or MTU. Traders can buy or sell single

units of equity shares. In the derivatives segment, however, the concept of trading in lots is still in vogue where the lot

sizes are specific to the underlying security. Unlike most other markets in the world where the MTU is fixed (usually

100 units), in the Indian market, the minimum size of the derivatives contracts is specified by the exchange based on

the price level of the underlying stock. During 2015, the market regulator revised the minimum size of a derivative

contract upward from INR 0.2 million to 0.5 million. This natural laboratory setup also provides us a unique

opportunity to observe the impact of market‐wide upward revision of MTU on market liquidity and trading volume.

With the introduction of algorithmic trading in various exchanges, presently a significant proportion of trades are initiated

automatically from computer terminals without any real‐time manual intervention. This paradigm shift in trading mechanism

has led traders to adopt appropriate trading strategies to minimize impact costs. Over the last decade and a half, the average

tradesizeinexchangesovertheglobehassignificantlyreduced

2

(Angel, Harris, & Spatt, 2011; O’Hara,Yao,&Ye,2014),owing

much of it to the increase in algorithmic trading activity. Aitken, Cumming, and Zhan (2017) show that the introduction of

algorithmic trading and high‐frequency trading (HFT), proxied by colocation services, significantly impacts trade sizes. Traders

often face a challenge of choosing optimum trade sizes to reduce overall impact cost and transaction cost (Bertsimas & Lo,

1998), especially when confronted with the problem of buying or selling a predefined quantity. Algorithmic traders use their

advantage of speed to split a larger order into smaller lots so that the price impact is minimal. They are more likely to carry out

a large number of small trades throughout the day rather than a few bulk trades. Algorithmic traders are also mostly intraday

traders who rarely carry over their positions to the next trading day.

3

In this context, it may not be wise to assume that MTU

restrictions will affect all trader groups (algorithmic vs. nonalgorithmic) uniformly.

Our work tries to draw from the existing strands of literature on algorithmic trading and MTUs. We look at how

MTU specification affects trading strategies of algorithmic traders vis‐à‐vis nonalgorithmic traders. A novel intraday

data set obtained from the National Stock Exchange (NSE) of India enables us to decompose the order‐level and trade‐

level data for different trader groups, namely proprietary, custodians, and nonproprietary noncustodians (NCNP)

traders. The data set also allows us to identify orders that were automatically generated from algorithmic trading

terminals. We observe that algorithmic traders try to trade at the smallest possible trade sizes, limited by the MTU.

Institutional investors, who have been known to trade on information, are observed to trade at relatively larger sizes

while not using algorithms to execute their trades. However, when using algorithms, they trade at much smaller sizes,

possibly to reduce the chances of being front‐run.

The announcement by SEBI to increase minimum contract size created much outcry in the Indian market, fueling

speculation that this move could force out retail traders from the market.

4,5

Considering that retail traders contribute a

significant proportion of the traded volume in the Indian SSF market, this regulatory change could translate to a

significant reduction in overall traded volume.

The contribution of this paper to the existing literature is threefold. First, we empirically show that algorithmic traders

prefer smaller order and trade sizes and are more likely to influence the size of trades they participate in. Second, we show that

MTU restriction refrains market participants, especially algorithmic traders, from optimizing their trade sizes and force them to

trade at the minimum specified trade size. Finally, we observe that unlike the impact of MTU reduction that most certainly

improves liquidity, the converse is not always true. Contrary to expectations, the upward revision of the minimum size of

derivative contracts in NSE SSF market had a weak positive impact on liquidity and traded volume. We also find a transitory

negative impact on traded volume immediately after the announcement by SEBI.

2

|

INSTITUTIONAL SETTING

2.1

|

Derivatives trading in India

Derivatives trading was first introduced in India by NSE and Bombay Stock Exchange (BSE) in 2000. First products to

be introduced were index futures, followed by index options, options in individual stocks, and futures in single stocks.

2

Securities and Exchange Commission (SEC) release 34‐61358, 2010

3

This feature is more pronounced in case of HFT.

4

http://economictimes.indiatimes.com/markets/stocks/news/increase‐in‐futures‐lot‐size‐may‐shut‐out‐retail‐investors/articleshow/48430033.cms

5

http://economictimes.indiatimes.com/markets/stocks/news/lot‐size‐revision‐in‐fo‐puts‐small‐traders‐in‐a‐spot/articleshow/49105221.cms

356

|

BANERJEE AND BANERJEE