Are option traders more informed than Twitter users? A PVAR analysis

• Published date: 01 September 2022
• Author: Alex Frino,Caihong Xu,Z. Ivy Zhou
• Date: 01 September 2022
• DOI: http://doi.org/10.1002/fut.22303

Received: 29 November 2021

|

Accepted: 11 December 2021

DOI: 10.1002/fut.22303

RESEARCH ARTICLE

Are option traders more informed than Twitter users? A

PVAR analysis

Alex Frino

1

|Caihong Xu

2

|Z. Ivy Zhou

1

1

School of Business, University of

Wollongong, Northfields Ave,

Wollongong, New South Wales, Australia

2

Stockholm Business School, Stockholm

University, Stockholm, Sweden

Correspondence

Z. Ivy Zhou, Faculty of Business and

Law, University of Wollongong, Bldg 40,

Northfields Ave, Wollongong, NSW 2522,

Australia.

Email: izhou@uow.edu.au

Funding information

Jan Wallanders och Tom Hedelius

Stiftelse samt Tore Browaldhs Stiftelse,

Grant/Award Number: P19‐0114

Abstract

Prior research has examined whether Twitter information predicts stock returns

and volatility. We study the causality between Twitter information, stock‐realized

volatility, and option‐implied volatility using a panel vector autoregressive model.

Using panel data on S&P/ASX 200 stocks, we reveal a bidirectional causality

between realized volatility and Twitter activity and divergence of opinion. We also

find strong evidence of causality from implied idiosyncratic volatility to Twitter

activity, sentiment, and divergence of opinion. Our results highlight the role of the

options market in predicting Twitter information and monitoring social media

flows to prevent the spread of fake news.

KEYWORDS

implied volatility, realized volatility, social media, Twitter

1|INTRODUCTION

Traditionally, the main sources of financial information are company reports, broker reports, and newspaper releases.

The advent and fast propagation of social media microblogs and networks, particularly Twitter, have created new

platforms for an almost real‐time dissemination of financial and other information to many users. Guidance published

by the United States (US) Securities and Exchange Commission (SEC) on April 2, 2013 permits companies to use social

media, including Twitter, to communicate corporate announcements,

1

with this reinforcing the credibility of Twitter as

a platform for disseminating corporate news. In traditional information sources, most news articles need to be verified

by publishers or specialists before they are made available to the public; however, with the new social technologies,

both factual and inaccurate information can be spread rapidly to many market participants. False rumors and mis-

information have been known to affect stock prices and large‐scale investments. For example, a false tweet claiming

that US President Barack Obama was injured in an explosion wiped out US$130 billion in US stock value.

2

The literature suggests that the derivatives market leads stock market prices, therefore playing an information

assimilation role (Anthony, 1988; Easley et al., 1998; Frino and West, 1999). Researchers have previously looked at

Twitter as an information dissemination platform: the current study extends their work by examining whether option

traders are more informed than social media users about firm‐specific information. Specifically, we examine the

J Futures Markets. 2022;42:1755–1771. wileyonlinelibrary.com/journal/fut © 2022 Wiley Periodicals LLC

|

1755

1

Press release on April 2, 2013: “The Securities and Exchange Commission today issued a report that makes clear that companies can use social

media outlets like Facebook and Twitter to announce key information in compliance with Regulation Fair Disclosure (Regulation FD) provided

investors have been alerted about which social media will be used to disseminate such information.”https://www.sec.gov/news/press-release/2013-

2013-51htm.

2

K. Rapoza, “Can ‘fake news’impact the stock market?”Forbes, February 26, 2017. https://www.forbes.com/sites/kenrapoza/2017/02/26/can-fake-

news-impact-the-stock-market/#37eb20502fac.

lead–lag structure between social media, stock, and option market variables to determine whether social media

platforms produce new information or only disseminate existing information already incorporated into stock and option

markets. Findings from this study provide a framework from which stock exchanges and policy makers can develop

innovative and effective procedures and regulations to curb “rumourtrage”on social media platforms.

Previous studies report a significant contemporaneous relationship between social media activity and the stock market, as

well as the role of social media in predicting stock returns (e.g., Antweiler & Frank, 2004;Bartovetal.,2017;Chenetal.,2014;

Duz Tan & Tas, 2021; Giannini et al., 2018;Renault,2017;Sprengeretal.,2014) and volatility (Glasserman & Mamaysky, 2019;

Jiao et al., 2020). Nevertheless, there is a lack of research on these associations in the Australian market, with most of the

existing literature centered on the US context. Although the SEC requires companies to promptly disclose nonpublic in-

formation, the acceptable methods for public disclosures do not have statutory backing and may still leave space for selective

disclosures and information asymmetry (Bushee & Miller, 2012;Miller,2006). To circumvent this constraint, we collect data

from the Australian market, which, unlike the US market, is characterized by a statutory‐backed continuous disclosure

framework.

3

ThedifferencebetweentheUSandAustralianmarketscan be explained by the materiality requirement of the

continuous disclosure rules. The stringent continuous disclosure regime in Australia provides a unique setting in which to

study the effectiveness of social media on stock volatility and option‐implied volatility (OIV). This is because in the Australian

context, all price‐sensitive information is required to be released to the public through official channels without delay.

Several studies suggest that nonpublic information is reflected in the options market, providing an ideal environment for

investigating the effects of social media activity on financial instrument prices (Chakravarty et al., 2004; Easley et al., 1998;

Manaster & Rendleman, 1982; Pan & Poteshman, 2006). Twitter also provides a platform for nonscheduled or nonperiodic

news to be made available to the public. Understanding how this information is incorporated into market prices, in

particular, how the information flows between social media platforms, option traders, and stock market participants, is

important to financial market transparency and stability. This study aims to extend past research by investigating the

relationship between stock‐related social media messages and trading activity in the options market. More specifically, this

study investigates whether changes in the volume and sentiment of stock‐related tweets affect OIV. OIV has been shown to

be a superior predictor of future stock volatility (Beckers, 1981;Chiras&Manaster,1978; Christensen & Prabhala, 1998;

Latané & Rendleman 1976;Xingetal.,2010), while following completely different patterns from the realized volatility

examined in the existing social media literature (Glasserman & Mamaysky, 2019;Jiaoetal.,2020).

Antweiler and Frank (2004) present a significant contemporaneous relationship between activity in social microblogs

and in the US stock market, and address the role of social microblogs in predicting stock returns and volatility. We extend

the work of Antweiler and Frank (2004) by providing the following contributions on social media and financial markets.

First, we examine the contemporaneous relationship between Twitter information and stock and option market variables,

such as realized stock volatility, option‐implied idiosyncratic volatility, and OIV. Consistent with results presented in extant

studies using US market data, we find that realized stock volatility measures are positively correlated with the volume of

Twitter messages and the dispersion of Twitter sentiment, but negatively correlated with average Twitter sentiment.

Furthermore, we find that the contemporaneous relationship between Twitter information and option market variables is

consistent with results using stock market data. Second, we conduct a panel vector autoregressive (PVAR) model to

investigate the Granger causality between Twitter, stock, and option market variables. We find a bidirectional Granger

causality between the Twitter disagreement index and stock volatility, and between Twitter positing activity and stock

volatility. Most importantly, we show that option‐implied idiosyncratic volatility presents strong causality to all three Twitter

variables: messages, sentiment, and disagreement, indicating that option traders are more informed than Twitter users.

2|DATA

2.1 |Stock and option data

Our analysis is based on a data set of ASX 200 stocks in the period between March 7, 2016 and February 28, 2020,

sourced from Bloomberg. We filter stock options that are inactively traded; that is, they are traded for less than 60% of

our sample period, and we end up with 66 stocks with active option series.

3

Continuous disclosure in Australia refers to a listed entity's legal obligation to immediately inform the Australian Securities Exchange (ASX) of

information likely to have a material impact on the price of its securities.

1756

|

FRINO ET AL.