Is a portfolio of socially responsible firms profitable for investors?

ABSTRACT

This paper investigates the presence of integration between six socially responsible stocks for the purpose of portfolio composition. Data for our study are based on the daily frequency and range from March 2016 to April 2019. Our results highlight that Gender Diversity funds exhibit a low correlation pattern with Low Carbon, Social Choice and USA ESG funds, whereas the Social Choice equity demonstrates a low correlation pattern with USA ESG and Social ETF funds across all decomposed scales. These results are also supported by the findings of non-linear Granger causality test across all investment horizons, i.e. from D1 to D8. Our results imply the inclusion of different asset classes together with socially responsible funds in a portfolio which may have useful implications for investors.

Research Highlights

• We investigate the co-movement between socially responsible funds returns.

• We decompose the correlation between socially responsible funds using maximal overlap discrete wavelet transformation (MODWT).

• Decomposed correlations are further captured using the multi-scale rolling window wavelet coherence

• Non-linear Granger causality is applied to infer implications for risk spillover under short-, medium- and long-run investment horizons.

• Results suggest that the inclusion of different asset classes together with socially responsible funds can yield optimal returns.

KEYWORDS:

• Socially responsible stocks
• rolling window wavelets
• wavelets transformation

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributors

Mobeen Ur Rehman is a Research Fellow at Institute of Business Research, University of Economics, Ho Chi Minh City, Vietnam.

Xuan-Vinh Vo is a dean at Institute of Business Research, University of Economics, Ho Chi Minh City, Vietnam.

Notes

1 For details, see Muñoz (2019).

2 Also known as intrinsic motivation.

3 Existing literature also sometimes documents the use of the term ‘integration’ as stock returns correlation. Because the word correlation usually refers to conditional or unconditional correlation between stocks, we use the term ‘integration’, hereby focusing on short- and long-run stock returns’ association by employing other novel econometric techniques.

4 Risk spillover refers to the transmission of risk from one stock to another. In the process of risk spillover between two stocks, one stock acts as a transmitter of information, whereas other acts as a receiver of information. The greater the magnitude or risk spillover between two stock markets, the higher the risk associated with a portfolio containing these two stocks.

5 The MODWT decompositions will be made available to the readers upon request.

6 Whitcher, Guttorp, and Percival (1999).

7 This is because after the application MODWT to a sub-window containing 250 data points by avoiding the boundary wavelet coefficients, the number of data points becomes much lesser than 250 for the fifth scale. Therefore, we use calculations as N-W, where N = 1043 and w = 250 which makes N-w = 793 windows, and thus the correlation coefficient.