Elsevier

Pacific-Basin Finance Journal

Volume 52, December 2018, Pages 123-133
Pacific-Basin Finance Journal

Islamic spot and index futures markets: Where is the price discovery?

https://doi.org/10.1016/j.pacfin.2017.04.003Get rights and content

Highlights

  • We test for price discovery (PD) in Islamic spot and futures markets.

  • Unique dataset: 900 stocks from 19 countries over 1982 to 2015

  • Spot market dominates PD in 63% of countries.

  • Using evidence of PD leads to annualized profits of 4.91%.

Abstract

This paper examines the source of price discovery for Islamic stocks. We pair a large number of Islamic stocks to country-specific index futures and estimate price discovery using a vector error correction model. The results obtained using data for 19 countries suggest that for most countries (63% of the sample) price discovery is dominated by the spot market. We show that for these countries, a mean-variance investor makes annualized average profit of 4.91% compared to an average buy-and-hold profit of 2.97% per annum.

Introduction

In this paper we test whether price discovery is dominated by the Islamic stock market or the index futures market to which these Islamic stocks belong to. The inspiration for this idea has roots in the large empirical literature which has evolved over the last decade-and-a-half, promoted in large part by the influential Presidential Addresses by O’Hara (2003). On Islamic stock markets in particular, Narayan et al. (2016b) test whether price discovery predicts asset prices using a sample of 188 Asia-Pacific Islamic stocks. They show that price discovery matters in pricing stocks. One imminent gap in this Islamic finance literature is the lack of understanding of where exactly the price discovery takes place. In fact, nothing is known about this process at all. Our research question therefore is: Is it the Islamic stock market or the index futures market that dominates the price discovery process? This question is not trivial because understanding the source of price discovery can guide investment decisions.

Our approach to testing for price discovery follows three steps. In the first step, we compile stock-level data comprising of 900 Islamic stocks listed across 19 different markets for the 1982 to 2015 period. We then match this stock-level price data with the corresponding market index futures price data. This gives us a unique spot-futures price dataset. In the second step, we utilize a panel vector error correction model (VECM) recently proposed by Karabiyik et al. (2016) to get panel estimates of price discovery. The main advantage of using this panel approach when compared to the more conventional time series approach is its ability to exploit the information contained in the cross-sectional dimension of our panel, leading to relatively accurate estimates of price discovery. Another advantage of the panel approach to price discovery is its ability to summarize the information contained in a large cross-section of stocks. The country-level panels that we consider contain up to 167 stocks. Imagine trying to report and summarize the results from 167 stock-specific VECMs – a daunting task, to say the least. In the final step, we design a framework that links the evidence of price discovery to investment decision making. We, in particular, show that one can utilize the information on price discovery to forecast returns and devise economically meaningful trading strategies.

We make three main findings. Our first finding is that the price discovery process is dominated by the spot market. This is the case for 12 out of the 19 markets. Our second finding relates to the robustness of the evidence of price discovery. We use both monthly and weekly data, and conclude with the same results, suggesting that the evidence of price discovery we document is insensitive to data frequency. Our third finding relates to the economic significance of price discovery. For the 12 countries for which spot market dominates price discovery, we show that the annualized average profit turns out to be 4.91%, whereas the average buy-and-hold profit is only 2.97% per annum. This represents a 62.35% increase in average profits when using the evidence on price discovery.

Our findings make three contributions to the literature. The first literature our work connects to is that on Islamic finance. Islamic finance research offers an avenue for understanding the behaviour of financial markets from a different perspective. The distinction between Islamic and non-Islamic (or conventional) financial markets have been well-documented and we refrain from repeating what is well-known; see, for instance, Ibrahim (2015) . There has been a surge in research on Islamic finance, attempting to understand this relatively new investment class. In this regard, recent studies have examined the profitability of Islamic markets (Narayan and Phan, 2016;Narayan and Bannigidadmath, 2015, Ashraf and Mohammad, 2014, Al-Khazali et al., 2014), the efficiency of Islamic banks (see Johnes et al., 2014, Belanes et al., 2015), interactions among Islamic equity markets (Yilmaz et al., 2015) and Islamic bond market (Azmat et al., 2014a,Azmat et al., 2014b, Azmat et al., 2015), the connection between Islamic markets and global crises (Kenourgios et al., 2016), corporate social responsibility (Mallin et al., 2014Abdelsalam et al., 2014), and the predictability of Islamic stock prices (Narayan et al., 2016a).1 We contribute to this literature by exploring the source of price discovery, which is an important component of the functioning of financial markets. While from previous studies we understand that Islamic stock markets are profitable, Islamic banks are more efficient compared to conventional banks and that Islamic stocks, like conventional stocks, are predictable, from our study we now understand that when it comes to price discovery in Islamic stock markets, in most countries it is the spot market that dominates the price discovery process. In discovering this empirical evidence, our findings are consistent with the trading volume hypothesis, which associates high trading volumes in the spot market (compared to the futures market) to a greater role of the spot market in the price discovery process.

We also connect with the recent study of Narayan et al. (2016a), which is to the best of our knowledge the only study on price discovery in Islamic stocks. This is in fact the study which comes closest to our work although it is completely different in its aims. This other study searches for time-varying price discovery in a wide range of stock price-based portfolios (made up of 188 Asia-Pacific stocks) and then uses this as a predictor of Islamic stock returns. The hypothesis is that time-varying price discovery is able to predict Islamic stock returns, and the authors find strong support for their hypothesis. Our goal is to search for price discovery in Islamic spot and corresponding country futures market price index for 900 stocks belonging to 19 countries. Our study can therefore be seen as complementing the work of Narayan et al. (2016a). In interpreting the connection between our work and that of Narayan et al. (2016a) it is important to note that while we take the classical definition of price discovery Narayan et al. (2016a) depart from this definition. Their point is that as long as stocks belong to the same universe the sum of stocks contribution to price discovery should be 100%.

Our finding of mixed evidence on price discovery documenting a role in price discovery for both spot and futures markets joins studies in other strands of the literature where similar sort of evidence has been found. In particular, the evidence from the commodity market suggests that the process of price discovery is not clear-cut. Dolatabadi et al. (2015) test whether price discovery in the commodity markets is dominated by the spot or the futures market. They find mixed evidence. Moreover, analysing the stock index cash and futures markets, Pizzi et al. (1998) also document mixed evidence on the source of price discovery.2 Our results seem consistent with this literature.

Our final contribution relates to the economic significance of price discovery. Identifying price discovery is one thing but understanding its economic relevance is another. We design a framework within which one can utilize the information on price discovery to specify return-forecasting models. The relevance of such models is that return forecasts can be utilised to devise trading strategies. We demonstrate that these strategies are meaningful, and are robust to different risk aversion parameters and to the allowance for both short-selling and borrowing. The trading strategy framework we design offers a fresh perspective on understanding the importance of price discovery. In this regard, our paper makes both a statistical and an economic contribution to the literature on price discovery.

We organise the balance of the paper as follows. In Section 2, we discuss the econometric methods used to examine price discovery. Section 3 discusses the dataset and presents the results. The final section provides concluding remarks.

Section snippets

Econometric method

The majority of the empirical literature on price discovery analysis has been focusing on time series estimation of price discovery parameters. However, recently there has been some attempts to use a panel data approach. In this paper we follow this fashion and adopt a panel data approach to price discovery. The theory behind this approach is developed by Karabiyik et al. (2016). They consider the two most popular time series approaches for price discovery and adapt them to a panel data setup.

Dataset

Our dataset is unique and has two components. The first is stock-level data, which we obtain from Datastream. The second is the stock price index futures data, which we obtain from Bloomberg. There is no specific database that contains Islamic stock price data. The data therefore need to be manually compiled. This we do; hence, the data are unique. In compiling the Islamic stock price data, we implement a criteria suggested in recent work by Narayan et al., 2016a, Narayan et al., 2016b. We only

Concluding remarks

There is no knowledge on whether the price discovery process with respect to Islamic stocks is spot market or futures market driven. This is the subject of our investigation. We compile a unique stock-level dataset covering 900 Islamic stocks belonging to 19 countries. We then form country-wise panels of stock prices and the stock market index futures price. Following this, we measure price discovery within a panel data setting following the proposal of Karabiyik et al. (2016). The evidence we

References (41)

  • NarayanP.K. et al.

    An analysis of commodity markets: what gain for investors?

    J. Bank. Financ.

    (2013)
  • NarayanP.K. et al.

    Are Islamic stock returns predictable? A global perspective

    Pac. Basin Financ. J.

    (2016)
  • NarayanP.K. et al.

    Price discovery and asset pricing

    Pac. Basin Financ. J.

    (2016)
  • RahimA.M. et al.

    Portfolio diversification benefits of Islamic investors with their major trading partners: evidence from Malaysia based on MGARCH-DCC and wavelet approaches

    Econ. Model.

    (2016)
  • ShabanM. et al.

    SME's lending and Islamic finance: is it a win-win situation?

    Econ. Model.

    (2016)
  • ShamsuddinA.

    Are Dow Jones Islamic equity indices exposed to interest rate risk?

    Econ. Model.

    (2014)
  • YilmazM. et al.

    Cross-sectoral interactions in Islamic equity markets

    Pac. Basin Financ. J.

    (2015)
  • AzmatS. et al.

    The Shariah compliance challenge in Islamic bond markets

    Pac. Basin Financ. J.

    (2014)
  • AzmatS. et al.

    The (little) difference that makes all the difference between Islamic and conventional bonds

    Pac. Basin Financ. J.

    (2015)
  • BasakS. et al.

    An equilibrium model with restricted stock market participation

    Rev. Financ. Stud.

    (1998)
  • Cited by (16)

    • Price discovery in bitcoin futures

      2020, Research in International Business and Finance
      Citation Excerpt :

      Our paper contributes to two main strands of literature; first, our study contributes to the investigation of price discovery in futures markets, which is an issue that continues to receive considerable attention in the related empirical literature. The relevant evidence regarding the lead–lag relationship between future and spot prices generally supports the hypothesis that price discovery is dominated by futures trading in a range of asset classes, e.g., in US equity markets (Kawaller et al., 1987; Stoll and Whaley, 1990; Chan, 1992; Wahab and Lashgari, 1993; Koutmos and Tucker, 1996; Pizzi et al., 1998; Tse, 1999; Hasbrouck, 2003; Chou and Chung, 2006), in international equity markets (Iihara et al., 1996; Booth et al., 1999; Covrig et al., 2004; Brooks et al., 2001; So and Tse, 2004; Zhong et al., 2004; Bohl et al., 2011; Theissen, 2012; Yang et al., 2012; Karabiyik et al., 2018; Ahn et al., 2019; Fassas and Siriopoulos, 2019), in commodities (Kuiper et al., 2002; Peri et al., 2013; Shrestha, 2014; Dolatabadi et al., 2015; Adämmer et al., 2016; Hauptfleisch et al., 2016; Dimpfl et al., 2017), in foreign exchange markets (Chen and Gau, 2010) and in other markets, such as the market for carbon emissions (Rittler, 2012; Mizrach and Otsubo, 2014) and volatility (Shu and Zhang, 2012; Frijns et al., 2016; Chen and Tsai, 2017). The particular study also contributes to the expanding literature on bitcoin.

    • A fractional cointegration VAR analysis of Islamic stocks: A global perspective

      2020, North American Journal of Economics and Finance
      Citation Excerpt :

      Studies on the dynamics of Islamic stocks have increased in recent time due to increasing global acceptance of Islamic investment products hitherto exclusively seen as faith-based investment options for Muslims only (see Narayan & Phan, 2019 for an extensive review and direction of previous research in Islamic finance). The vast majority of previous studies explored the profitability and efficiency of Islamic financial markets, their predictability and offer some comparative analyses between Islamic financial markets and their conventional counterparts (see for example, Al-Khazali et al., 2014; Abu-Alkheil et al., 2017; Anwar & Ali, 2018; Charles et al., 2017; Daher, Masih, & Ibrahim, 2015; Firdaus & Hosen, 2013; Hidayati, Siregar, & Pasaribu, 2017; Hkiri et al., 2017; Karabiyik et al., 2018; Kenourgios et al., 2016; Narayan et al., 2016; Narayan & Bannigidadmath, 2017; Narayan & Phan, 2017; Narayan, 2018; Narayan, Sharma, Thuraisamy, & Westerlund, 2018; Solihin, Achsani, & Saptono, 2016; Shahzad et al., 2017; Sukmana & Ibrahim, 2017). As an improvement on these studies and in line with the direction of future research in Islamic finance put forth in Narayan and Phan (2019), our study makes valuable contributions to the Islamic finance literature as a distinct entity but considers conventional stocks for robustness.

    • Applied energy and finance

      2019, Economic Modelling
    • Evidence of price discovery on the Indonesian stock exchange

      2019, Economic Modelling
      Citation Excerpt :

      In this paper, we join recent studies investigating price discovery in emerging markets by considering Indonesia—a market about which very little is known from a price discovery perspective. In this regard, as we review below, Hande et al. (2018) is an exception. We investigate price discovery and start with an acknowledgement that our work is preliminary and will, therefore, set the foundation for additional work on Indonesia’s stock exchanges.

    View all citing articles on Scopus

    Westerlund thanks the Jan Wallander and Tom Hedelius Foundation for financial support under research grant number P2014–0112:1. Westerlund and Karabiyik thank the Knut and Alice Wallenberg Foundation for financial support through a Wallenberg Academy Fellowship.

    View full text