Improving momentum strategies using residual returns and option‐implied information

• DOI: http://doi.org/10.1002/fut.21988
• Published date: 01 April 2019
• Author: Ming‐Yu Liu
• Date: 01 April 2019

Received: 6 May 2018

|

Revised: 2 December 2018

|

Accepted: 2 December 2018

DOI: 10.1002/fut.21988

RESEARCH ARTICLE

Improving momentum strategies using residual returns

and option‐implied information

Ming‐Yu Liu

Department of Finance, National Taiwan

University, Taipei, Taiwan

Correspondence

Ming‐Yu Liu, Department of Finance,

National Taiwan University, No. 1,

Section 4, Roosevelt Road, 10617 Taipei,

Taiwan.

Email: d98723005@ntu.edu.tw

Abstract

This paper provides an alternative method for enhancing momentum profits by

combining residual returns and option‐implied information. The results show

that the main benefit of applying residual returns to construct momentum

portfolios is generating stable returns. Additionally, the incorporation of

implied volatility (IV) spread or IV skew into a residual momentum portfolio

is found to significantly raise profits, particularly during bad times and

high‐sentiment periods. This is because IV spread and IV skew can dissociate

winners/losers with a price underreaction from those with a price overreaction,

which suggests that informed traders who perceive price underreactions/

overreactions trade in option markets.

KEYWORDS

momentum, residual returns, underreaction/overreaction, volatility spread/skew

1

|

INTRODUCTION

While the significant profitability of a price momentum strategy is well documented in the literature (Jegadeesh &

Titman, 2011), some studies find that momentum has diminished in recent decades (Bhattacharya, Li, & Sonaer, 2017;

Chordia, Subrahmanyam, & Tong, 2014). Previous research identifies two possible causes for the decreasing

momentum. One is that the financial turbulence of the 2000s (e.g., the dot‐com bubble and the 2008 financial crisis)

resulted in severe momentum losses (Barroso & Santa‐Clara, 2015; Daniel & Moskowitz, 2016).

1

Another is that

increasing market liquidity and trading activity facilitate market efficiency, attenuating price momentum (Chordia

et al., 2014). To address the issue of decreasing momentum due to financial turbulence and increasing market

efficiency, this paper constructs momentum portfolios based on residual returns and option‐implied information and

examines their performance.

The extant literature shows that momentum strategies exhibit dynamic exposure to systematic factors (Blitz, Huij, &

Martens, 2011; Daniel & Moskowitz, 2016; Grundy & Martin, 2001; Martens & van Oord, 2014). This can result in

severe momentum losses, particularly in times of financial turbulence. One way to mitigate the impact of systematic

dynamics is grouping stocks into winners and losers based on residual returns. Blitz et al. (2011) show that a residual

momentum portfolio (where a portfolio is formed based on residual returns) has lower loadings on systematic factors

and more stable returns than a conventional momentum portfolio (where a portfolio is formed based on raw returns).

In addition as market efficiency is increasing over time (Chordia, Roll, & Subrahmanyam, 2008; Chordia, Roll, &

Subrahmanyam, 2011), one possible method to enhance momentum profits is by incorporating option‐implied

information into momentum portfolios. This is because informed traders may trade based on their private information

J Futures Markets. 2019;39:499–521. wileyonlinelibrary.com/journal/fut © 2018 Wiley Periodicals, Inc.

|

499

1

Barroso and Santa‐Clara (2015) find that the cumulative momentum return from March to May 2009 is −73.42%. Daniel and Moskowitz (2016) document that momentum experiences a large loss

of −49.19% and −45.52% in January 2001 and April 2009, respectively.

regarding winners/losers in option markets,

2

and thus option‐implied information could be used to dissociate winners/

losers whose prices are going to move in the same direction as before (price underreaction) from those whose prices are

going to move in the opposite direction as before (price overreaction).

3

A momentum portfolio composed of winners/

losers with a price underreaction is expected to outperform a conventional momentum portfolio that is composed of all

winners/losers.

Based on the grounds above, I combine residual returns and option‐implied information to construct momentum

portfolios. I first group stocks into winners and losers according to residual returns and then use implied volatility (IV)

spread (Cremers & Weinbaum, 2010) and IV skew (Xing et al., 2010) to identify winners/losers with a price

underreaction to prior good/bad information. Winners with high IV spreads (or low IV skews) and losers with low IV

spreads (or high IV skews) are expected to exhibit price underreaction. The empirical results support the findings of

Blitz et al. (2011) that sorting stocks by residual returns rather than by raw returns can mitigate the dynamic exposure

to systematic factors and market‐wide influences, and thus generates stable returns. In addition incorporating IV spread

or IV skew into a residual momentum portfolio can significantly raise profits, particularly during bad times and high‐

sentiment periods. Therefore, a residual momentum portfolio composed of winners with high IV spreads (low IV

skews) and losers with low IV spreads (high IV skews) is superior to a momentum portfolio constructed simply based

on residual returns or raw returns.

Furthermore, I find that winners with high IV spreads (low IV skews) and losers with low IV spreads (high IV

skews) exhibit price continuation, which suggests that the prices of these winners/losers have underreacted to prior

information. In contrast, winners with low IV spreads (high IV skews) and losers with high IV spreads (low IV skews)

exhibit price reversals, which suggests that the prices of these winners/losers have overreacted to prior information.

These results show that IV spread and IV skew have the ability to dissociate winners/losers with a price underreaction

from those with a price overreaction.

This paper contributes to the literature in several ways. First, this paper is related to the research on managing

momentum risk. Barroso and Santa‐Clara (2015) and Daniel and Moskowitz (2016) hedge momentum risk by

dynamically adjusting the weight in a momentum portfolio. Martens and van Oord (2014) hedge the dynamic exposure

to the Fama‐French factors that momentum traders suffer from by taking opposite positions in the factors. Differing

from the prior studies that focus on technical hedging methods, I select winners and losers based on information about

a firm per se. Residual returns are believed to be a proxy for firm‐specific information (Blitz et al., 2011) since they are

unrelated to systematic factors, and IV spread/skew captures informed traders’perspective on a firm. Taken together,

this paper provides an alternative way to enhance momentum profits in the face of financial turbulence and increasing

market efficiency.

Second, this paper makes a contribution to the literature on investor underreaction and overreaction.

The extant studies present the evidence that short‐term price momentum results from investor underreaction

(L. K. C. Chan, Jegadeesh, & Lakonishok, 1996; Z. Chen & Lu, 2017; Y. Chen & Zhao, 2012; Da, Gurun, &

Warachka, 2014; Hong, Lim, & Stein, 2000; Jegadeesh & Titman, 1993), while others attribute long‐term price

reversals to investor overreaction (Chopra, Lakonishok, & Ritter, 1992; De Bondt, & Thaler, 1985, 1987).

However, these studies do not simultaneously incorporate the two effects of momentum and reversals, which the

recent paper by Conrad and Yavuz (2017) argues separate phenomena. To the best of my knowledge, only Lee

and Swaminathan (2000) reconcile underreaction and overreaction well. By using option‐implied information,

which may contain informed traders’private information about winners and losers, I present evidence in support

of investors’initial underreaction and delayed overreaction (Barberis, Shleifer, & Vishny, 1998; Hong &

Stein, 1999; Lee & Swaminathan, 2000) and suggest that momentum and reversals are two sides of a single

phenomenon.

4

Third, this paper adds to the residual momentum literature by showing that investors indeed overreact to firm‐

specific information. Blitz, Hanauer, and Vidojevic (2017) and Chang, Ko, Nakano, and Ghon Rhee (2018) find that

residual momentum returns do not exhibit a long‐term reversal while conventional momentum returns do. This

indicates that investors underreact to firm‐specific information but overreact to market‐wide information. However, it

2

Easley, O’Hara, and Srinivas (1998) argue that option markets are an alternative venue for informed traders. This is supported by the empirical evidence that option markets lead stock markets (see,

e.g, An, Ang, Bali, & Cakici, 2014; Cremers & Weinbaum, 2010; Ge, Lin, & Pearson, 2016; Johnson & So, 2012; Xing, Zhang, & Zhao, 2010).

3

The link between price underreaction/overreaction and option‐implied information is discussed in Section 2.1.

4

It is worth noting that the purpose of this paper differs from that of Z. Chen and Lu (2017), who use call IVgrowth to identify stocks’information diffusion speed and test the hypothesis of Hong and

Stein (1999) that price momentum results from gradual information diffusion.

500

|

LIU