Data for Five Factor Asset Pricing Model of Shariah compliant firms in the US
Description
This data was utilised in answering the following research hypothesis: The Debt ratio within the contemporary Shariah Stock Screening procedure significantly impact the corporate financial behaviour of Shariah compliant firms, so much so that their asset pricing behaviour will be different compared to conventional firms. The data (and subsequent regressions) will show that samples of Shariah compliant firms will share similar asset pricing behaviour vis-a-vis the conventional sample, however, some clear differences will also manifest. The most striking is that the Shariah compliant samples will tend to have significant intercepts, which imply that the five-factor model fails to completely explain the variation of average excess returns within Shariah compliant samples. In short, there exists more room to add additional variables, alongside the five-factor model, when explaining the asset pricing behaviour of Shariah compliant samples in the US. The data comprises of monthly risk factor premiums of four samples (defined in the Steps-to-reproduce section). All data are sourced from Thompson Reuters Datastream. Please note that the data are in STATA .dta format, therefore, use the STATA program to open them. The data is ready to use as-is for regression purposes.
Files
Steps to reproduce
The data presented here are the risk factor premiums for four samples of public firms in the US used as the explanator for a regression on average excess returns of sorted firms. The sorting process and the subsequent construction of risk factor premiums exactly follows the methodology of Fama & French (2015). The four samples begin with all stocks listed in the NYSE, NASDAQ and IEX from January 2000 to December 2019, this sample is dubbed the All-Stocks sample (also known as the AS sample), it represents the "conventional" sample with which the "Shariah-compliant" samples explained next are compared with. Next, using the firms in the AS sample, we filter them through the Qualitative screening found in contemporary Shariah stock screening methodologies (the Non-compliant codes are based on S&PDJ Islamic Index definition) to arrive at the Qualitative sample. Then using the firms in the Qualitative sample, we run two parallel quantitative screening, the first with total market capitalisation as the denominator, whilst the second using total assets as the denominator, to arrive at the Market value Sample (also known as the MV sample) and the Total Assets sample (also known as the TA sample). Using the firms within these four samples, we then downloaded monthly total return index data, as well as annual firm-specific data including size, book-to-market ratio, profitability and investment from January 2000 to December 2019 via Thompson Reuters Datastream. In this data, a 2x2 sort is applied. To begin, first, sort the firms into small and big portfolios using their total market capitalisation as the proxy for size. For the AS and Qualitative sample, this is done by using NYSE size medians. For the MV and TA samples, this is done using size medians found within their own sample. The reasons for this practice is given in the linked article. Next, using each small and big portfolios, further sort the firms independently by book-to-market ratio, profitability and investments, all using the respective variable's medians. One should end up with 12 portfolios within each sample, then, using these 12 portfolios, one may then construct the risk factor premiums exactly in the manner of Fama & French (2015), or by following the steps in the linked article (both are the same). The sorting process, as well as the regression codes are carried out via STATA 14.