A Comparative Performance Analysis of Sustainability Themed Indices in India: Markov Regime Switching Approach

Abstract

Sustainability Indices serve as a benchmark for the companies screened for their superior performance on environmental, social and governance (ESG) parameters. This article intends to compare the overall and regime-specific financial performance of socially responsible indices of the National Stock Exchange, Nifty100 ESG and Nifty100 ESG Enhanced with Nifty50 (representing the market) from 1 April 2012 to 31 March 2020. Overall comparative performance analysis of these indices is conducted using risk-adjusted return measures and volatility has been captured through the TGARCH model. Further, time duration has been decomposed into regimes using Markov Regime Switching Model and the comparison of indices has been undertaken in both regimes. Our findings suggest that there is no significant difference between the return performance of sustainability indices and market benchmark index in single time duration and sustainability indices performing marginally better in both the regimes identified. This implies that socially responsible investments in India are providing reasonable returns to investors without comprising non-financial objectives. For corporates, it is a win–win situation to focus on ESG parameters to attract capital from investors and deliver better corporate financial performance and hence increasing the potential of growth of socially responsible investing in India.

Keywords

Markov regime switching Nifty100 ESG Nifty100 ESG Enhanced socially responsible investing sustainability volatility

Introduction

Socially responsible investing (SRI) is a financial investment strategy specifically designed to include social performance along with financial performance. SRI believes in ‘doing well by doing good’ (Azmi et al., 2019). SRI is an investment tool that negatively screens out companies not performing in pursuit of Sustainable Development Goals (SDG) and hence holds relevance in working towards sustainability of limited resources. Environmental, social and governance (ESG)-based investing is a new diversified field of capital markets leading to the development of interest among the academicians and practitioners concerning performance measurement (Schröder, 2007). The origin of SRI can be traced back to the clamorous mid-1960s, where a series of events beginning from the anti-Vietnam war followed by the civil rights movement escalated to SRI (Schueth, 2003). Bullish capital markets in the 1990s and increasing pressure from Institutional Investors in SRI led to the growth of responsible investing in developed markets of the US and Europe (Managi et al., 2012). Recently, the importance and awareness of SRI have been drastically increased and the focus is brought back in developing economies including Asian-Pacific countries owing to environmental awareness, need for social change, need for better corporate governance, equitable pay and standards, and labour welfare.

Sustainable investing includes incorporating ESG principles—environmental, social and governance—in the existing financial investment strategy (de Souza Cunha & Samanez, 2013). The environment includes concerns over climate change, carbon footprints and energy efficiency which can be transformed into opportunities. Social factors include human capital, health and safety which can offer the potential to generate alpha. Lastly, governance includes corporate governance, transparency and reporting practices. The investment risk due to climatic change, poor corporate governance, non-transparent business practices and inadequate corporate social responsibility are some of the drivers behind the rise of SRI. With constant pressure from government, regulatory authorities, the conduct of committees, rounds of deliberations, and rising concerns over environmental resources and human resources, the concept of socially responsible impact investing is on the front line of developed countries including the US, UK, Australia and Europe. Investors have increasingly shifted from self-interest values to ecological values and pro-environmental beliefs (Bhuian & Sharma, 2017).

Although the concept of sustainable investment is nascent in developing countries especially in India and is still growing, it is expected to soar high in the next couple of years. As per the Indian Impact Investors Council, almost 30 impact funds have invested in socially responsible corporates in India. There are few strategic initiatives taken by SEBI including mandated business responsibility report by listed companies, ESG disclosures norms and integration of Global Reporting Initiatives. All these efforts are making corporates more aware and inclined towards focussing on ESG scores and become sustainable companies in the long run. The increase in the CSR disclosures for Indian companies has created pressure to generate value for investors by behaving in a socially responsible manner (Jaisinghani & Sekhon, 2020). Researches have proved that embedding CSR into the company’s core strategy results in a positive impact on competitive performance (Yadav & Sinha, 2021).

SRI is noticeable in capital markets as companies have now realized that business benefits emerge from responsible practices and investors now focus more on ‘value returns’ by sustainability strategy (Jain et al., 2019). Stock Exchanges are increasingly becoming aware of sustainable business practices. Corporates on the other side operating in the global economy are also expected to operate within social constraints necessitating environment accounting and reporting (Acar & Temiz, 2020).

To keep a monitor on the performance of socially responsible companies following ESG principles, secondary markets all over the globe have introduced socially responsible indices or sustainability indices. The leading sustainability index in the world is ‘Dow Jones Sustainability Indices’ with a long history in global sustainability. India is no far behind; S&P ESG India Index, the first of its kind sustainable investable index, was introduced in the year 2008 and with no looking back. BSE launched S&P BSE GREENEX and S&P BSE CARBONEX in the basket of sustainability indices. National Stock Exchange (NSE) also came up with two ESG indices: NIFTY100 ESG and NIFTY100 ESG Enhanced in 2017 aligned with the theme of ESG investing. These ESG themed indices are designed to monitor the performance of companies having ESG scores and forming the part of Nifty100 (National Stock Exchange of India Ltd.). The weight of each stock in the index is derived from its ESG score and free-float market capitalization. Similar to this, Nifty100 Enhanced ESG where the constituents having normalized ESG score of at least 50% are included. As per the official website of NSE, companies engaged in the business of tobacco, alcohol and gambling are excluded.

Now the question arises, do these Indexes perform better than the overall Index? Is the financial performance in a single study period similar to the results in distinct regimes? All these questions are addressed by this particular research. The results of this study would imply the preference of investors towards investment in socially responsible companies along with the corporate implication to focus on ESG parameters and hence raise the bar of corporate governance.Research on the financial performance of socially responsible investment avenues has been developing as investors today want to reflect the broader values in their investments along with targeting reasonable financial returns which is the essence of sustainability (Talan & Sharma, 2019). This research is further developed after the review-related research undertaken in the area of sustainability investing. The particular research paper has been attempted to provide insights into the performance of socially responsible indices concerning the market index of NSE in India. Schröder (2004) in his research highlighted that Index return values eliminate the need for transaction cost of funds and managerial skills of fund managers. It becomes more relevant to study ESG, and ethical indices have resulted in awareness and acceptance of sustainability principles amongst the investors and corporates. Performance evaluation and comparison of ESG indices will guide the investors in their decision-making and hence contribute to the development of sustainable goals.

The rest of the article has six sections in total: Section 2 comprises of literature review on sustainability indices and their financial performance along with research gaps. Section 3 elaborates on data and research methodology followed by Section 4 on analysis of data. Section 5 covers results and Section 6 includes discussion and conclusion. Lastly, Section 7 is added on the implications of the study followed by future research avenues in the SRI performance measurement.

Review of Related Literature and Research Gaps

This section contains reviews of the concept of SRI followed by the related literature in the area of performance evaluation of sustainable investing.

According to Galema et al. (2008), SRI is a strategy of inclusion of non-financial dimensions in the financial decision-making process of investors. This has led to the increased attention and awareness of investors around the world to particularly analyse the performance of responsible investing. Academic studies on the performance of SRI and sustainability investing are broadly based on three research areas—socially responsible indices, socially responsible mutual funds and socially responsible stock portfolios.

Past research studies concentrating on performance comparison of socially responsible themed indices with their conventional benchmark indices have three alternatives. Studies of the performance of mutual funds in the US by (Albaity & Ahmad, 2011; Bauer et al., 2005; Goldreyer & Diltz, 1999) computed arithmetic measures of risk and return suggest that there is no significant difference between mean returns of socially responsible funds and conventional counterparts. Clark et al. (2014) matched the funds by age, investment and compared socially responsible funds with conventional funds in the UK and found no significant difference between both the funds. Other researches have shown that socially responsible investment underperformed the broad conventional investment (Benson et al., 2010; Hoti et al., 2005).

The underperformance of ethical investments can be attributed to a higher risk of SRI/ESG indices/funds probably due to inadequate ESG disclosures or lighter regulatory framework to monitor SRI performance (Rehman et al., 2016). Entine (2003) criticized the criteria used to include funds in socially responsible investments and found that SRI investments don’t yield superior returns. Taking into consideration the publications of the latest years, Borgers et al. (2015) and Charfeddine et al. (2016) investigated the performance of Islamic and unscreened counterparts and found that Islamic investments underperform the conventional benchmark. This indicates that a lack of norms, disclosures and restrictions in capital allocation can harm the returns of such investments. An empirical study of Śliwiński and Łobza (2017a) which has reported outperformance over benchmark has found it to be statistically insignificant. Bertrand and Lapointe (2015) and Lean et al. (2015) claimed that socially responsible investments have a higher risk-adjusted performance relative to the conventional benchmarks. Collison et al. (2008) compared the indices of the FTSE4Good series with their benchmarks and observed that the indices outperformed their unscreened benchmarks. Similar results of abnormal positive returns were reported by Paul (2017), thus encouraging higher flows of money in responsible investment funds.

Apart from the performance evaluation of SRI indices, there is a great deal of research that has been done to investigate the performance of SRI portfolios. These studies are based on the fact that ESG and its impact on market value may be different for different sectors of the economy. So, the construction of stock portfolios across sectors and their performance comparison with the conventional portfolios may display a better picture. BinMahfouz and Hassan (2013) concluded that Islamic sustainability investment portfolios are more exposed to technology, health care, and oil and gas. For Abramson and Chung (2000), telecommunications and the financial sector in the US were relatively more exposed to SRI due to their inherent volatility and de-regulation processes.

SRI as an investment option is widely used in developing nations including the US, UK and European nations. However, not much of the academic literature is available from developing countries. Talan and Sharma (2019) emphasized systematic literature on SRI and believe that studies from developing countries are lagging, indicating that there is a need for more comprehensive research in the area of performance comparison of socially responsible investment avenues for investors who are willing to align their beliefs and values along with financial returns. Pertaining to India, similar studies on performance analysis of sustainability indices are very few, this has been done by Vasal (2009), Gupta (2011) and Sudha (2014). These studies report relatively higher or similar returns with that of generic investment. Tripathi and Bhandari (2014) in their study compared the performance of Greenex, Carbonex and S&P ESG India Index with that of CNX Nifty using various risk-adjusted return measures and found that sustainability indices perform better than the benchmark CNX Nifty in the said duration. Jain et al. (2019) comparatively analysed the financial returns of ESG indices of developed and emerging markets and found that there is no significant difference between the returns, and hence ESG investing can be a good substitute for conventional indices. Sudha (2014) applied the TGARCH model to study the conditional volatility in the S&P ESG India index. Tripathi and Kaur (2020) in their recent study on performance analysis of BRICS nations found that India is leading in a crisis period in terms of its high risk-adjusted performance of socially responsible investments.

Majority of the studies undertaken to compare the risk and return performance of sustainability indices employed risk-adjusted return measures and Fama–French models (Tripathi & Bhandari, 2014) in a single period. Single-period studies argue that the performance of indices does not remain constant over time (Managi et al., 2012). According to Shank et al. (2005), different findings over different timings lengths have originated the need for regime effect. This notion suggests that the performance of SRI versus non-SRI might get affected by stock market regimes. This has been validated by the studies of Becchetti and Ciciretti (2009) and Cortez et al. (2009). Without considering the stock market regimes, negative and persistent shocks might get diluted with positive returns and become impossible to identify (Cheung et al., 2010). Several research studies have divided their study periods as per changes in the business cycle, but the stock market may not follow similar business cycle patterns (Angelidis et al., 2015). Hence, researchers have tried to use the Markov Switching (MS) model to identify the regimes endogenously (Ortas et al., 2012).

Also, the indices considered in the previous research studies are specifically S&P ESG India Index, for which the historical data is available till 2013 only. After the launch of Nifty100 ESG and Nifty100 Enhanced ESG in 2013, very few significant studies have been undertaken to compare the regime-specific financial performance of sustainability indices. Hence, this study will add new empirical insights into the existing literature as the focus is on comparing risk and return characteristics of sustainability indices of NSE vis-à-vis benchmark market index, CNX Nifty50 in the single period as well as regime specific periods because the performance of indices cannot stay constant over time (Li et al., 2010).

Second, almost all the studies on sustainability indices in India have used the Generalized Auto-Regressive Conditional Heteroscedastic Model (GARCH) model to predict the volatility in the index returns, but the stock market in India is majorly behavioural and sentiment-driven; hence, TGARCH model which can capture the leverage effect can also be used to accurately model the volatility.

From mandatory publishing of annual business responsibility report to BSE guiding on ESG disclosures, regulators and stock exchanges are driving changes in reporting and disclosures which has led to an increase in the scope of sustainable investing. Hence, sustainability indices can act as a catalyst in the development of SRI. Studies based on the performance comparison will help the investors to make financial decisions on selecting the stock portfolio to fulfil their social objective without compromising the returns on the investment. Hence, the main objectives of this study are to find out the risk-adjusted returns and volatility persistence in three stock indices, CNX Nifty50, Nifty100 ESG and Nifty100 Enhanced ESG over constant duration and comparison of statistical difference between mean returns of the indices and to divide the single time duration into regimes and comparison of financial performance of indices in distinct regimes based on high and low volatility.

Data and Research Methodology

Data and Variables Description

To accomplish the desired objectives, daily closing prices of sustainability (Nifty100 ESG and Nifty100 Enhanced ESG) and Nifty50 as a proxy of a market index (Sauer, 1997) have been retrieved from the official website, respectively. Tables 1 and 2 represent the top constituent stocks of sustainability indices of NSE with their relative weights in the given index. ESG indices were introduced by NSE in 2017. However, data is available from the base date of 2012 only. Hence, the time duration has been selected as a period of 1 April 2012 to 31 March 2020. Another reason for selecting this sample duration is that the major ESG and sustainability-related developments in the area of SRI took place from early 2011 followed by the pressure from regulatory bodies (Dalal & Thaker, 2019). Hence, ESG indices came in place to track the performance of the companies leading in ESG principles.

To get risk-free rate, securities with no or least default and with guaranteed returns are considered. Treasury bills (T-bill) are a combination of these two. Several books and research studies have shown that T-bills have practically no risk of default (Basu & Chawla, 2010; Sudha, 2014). Following the convention, the 91-days T-bill Index was maintained by the Clearing Corporation of India Ltd. (CCIL) as the proxy for the risk-free rate of return (Sudha, 2014). CCIL T-bills Index reflects the movement along with the short-term market, tracking the movement of the T-bills by computing the volume-weighted average prices of all traded securities. After matching the dates of the liquidity weighted T-bill index and NSE indices, 1,930 daily data points are obtained. The daily adjusted closing prices of indices are converted into returns by applying the given formula (R_t = (P_t – P_t-1)/P_t). The logarithm returns are generally used to stabilize the variance of a series (Changyong et al., 2014; Lütkepohl & Xu, 2012). In this case, the log transformation of the data variables did not make any change in the stability of variances. Further, the results were analysed using both simple returns and log returns and no significant difference was found. Hence, the author has opted for actual figures in place of the natural logarithm.

Table 1.

Top Constituents—Nifty100 ESG

Company’s Name	Weight (%)
Infosys Ltd.	10.22
Reliance Industries Ltd.	10.09
HDFC Bank Ltd.	9.90
Housing Development Finance Corporation	5.97
Tata Consultancy Services Ltd.	5.69
Hindustan Unilever Ltd.	4.21
Axis Bank Ltd.	3.13
Larsen & Toubro Ltd.	2.71
ICICI Bank Ltd.	2.38

Source: NSE Website.

Table 2.

Top Constituents—Nifty100 Enhanced. ESG

Company’s Name	Weight (%)
Infosys Ltd.	10.63
Reliance Industries Ltd.	10.05
HDFC Bank Ltd.	9.86
Housing Development Finance Corporation	5.95
Tata Consultancy Services Ltd.	5.92
Hindustan Unilever Ltd.	4.38
Axis Bank Ltd.	3.12
Larsen & Toubro Ltd.	2.64
Bharti Airtel Ltd.	2.47
ICICI Bank Ltd.	2.37

Source: NSE Website.

Methodology

This section covers methodology practiced calculating risk-adjusted performance measures of said indices and the description of econometric models applied to model the conditional inherent volatility in a single period and dynamics of the MS model to study the performance regime-wise.

Unit Root Diagnostic Testing

Any stochastic process is said to be stationary if its mean and variance are not changing over time, and the value of the covariance between the two time periods depends only on the lag rather than the actual time at which the covariance is ascertained (Gujarati & Sangeetha 2007, p. 816). Regression and other statistical tools applied to non-stationary time series can lead to misleading and spurious results. In this research study, the existence of unit root is tested through the Augmented Dickey–Fuller (ADF) test. The null and alternative hypotheses of the ADF test can be stated as:

H _o: Presence of unit root in the index return series.

H _a: Index return series is stationary.

The regression equation of ADF test is as follows:

Δ Y_{t} = β_{1} + β_{2} t + δ Y_{t - 1} + \sum_{t - 1}^{m} α i Δ Y_{t - 1} + ε_{t}

Where,

Y(t-1) = Lag 1 of time series

ΔY_t_–1 = First differential of the time series at the time (t-1)

where ε_t is a pure white noise error term

ARCH LM Test for Conditional Heteroskedasticity

Financial time series many times exhibits the property of jagged peaks implying large (small) changes in prices tend to cluster together, called volatility clustering or autoregressive conditional heteroskedasticity (Mandelbrot, 1997). The presence of heteroskedasticity in stock returns signifies that the unexpected volatility in the previous period affects the investment decisions in the current time (Niyitegeka & Tewar, 2013).

ARCH Model Specification

The ARCH model was introduced by Engle (1982) to determine the volatility in the time series. Before that, the volatility was determined under the assumption of constant variance (homoscedasticity) using the past observations. However, practically the variance is not constant. ARCH model on the other hand helps in getting better estimators by handling the heteroscedasticity in the errors. In ARCH, the conditional variance of the error is dependent on lagged square error terms. Higher-order Arch may pose computational problems; hence, GARCH is a more widely accepted model for volatility estimation.

GARCH Model Specification

To address the problems of multicollinearity of explanatory variables (Lin, 2018), GARCH was proposed by Bollerslev (1986) which has been proved successful for volatility modelling of index returns where variance is not constant across (Lean & Nguyen, 2014; Sariannidis et al., 2010). GARCH models are particularly applied in the financial time series where volatility clustering is observed. The coefficients of the lag period squared error terms indicate the intensity of shocks in the short term whereas coefficients of past variances(GARCH term) are indicative of persistence of long-run shocks (Yavas & Rezayat, 2016). So, to model the volatility through the GARCH model, the presence of the ARCH effect has been confirmed along with testing of a unit root. The equation of the GARCH model is as follows:

R_{t} = a_{0} + a_{1} R_{t - 1} + ε_{t}

(1)

Where,

R_t = Index returns at time t

a₀ = Constant term

R_t_-1 = Lagged index returns

a₁ = Coefficient of lagged index returns

ε_t = Error term

ε_t~N(0, σ²)

σ_{t}^{2} = ω + α_{1} ε_{t - 1}^{2} + β_{1} σ_{t - 1}^{2}

(2)

Where,

ω = Constant

α₁ = Volatility of past residual squared returns

β₁ = Volatility of past variances of index return

ε_2t−1 = Lagged squared past error terms

σ_2t−1 = Past variances of mean returns

Most existing empirical studies suggest the first order (p = 1, q = 1) GARCH model (Awartani & Corradi, 2005; Emenike, 2010; Hansen & Lunde, 2005; Sabiruzzaman et al., 2010). Hence, we have also considered the GARCH (1,1) in our study. Several empirical studies have also applied TGARCH over GARCH models to predict the persistent volatility (Coffie, 2015; Hafner & Herwartz, 2000).

σ_{t}^{2} = ω + \sum_{i = 1}^{q} α_{i} ε_{\dot{t} - i}^{2} + \sum_{i = 1}^{γ} γ_{i} ε_{t - i}^{2} d_{t - i} + \sum_{j = 1}^{q} β_{j} σ_{t - i}^{2}

Where,

$σ_{t}^{2}$ = Conditional variance

ω = Constant

$ε_{\dot{t} - 1}^{2}$ = Previous period squared error term from the return equations

$σ_{t - 1}^{2}$ = Previous period conditional variance

d_t_–1 = Threshold value

Where d_t_–1 = 1 if $ε_{\dot{t} - 1}^{2}$ <0 and 0 otherwise.

This study has also made a comparison between standard GARCH(1,1) and TGARCH models. For lag length and selecting the best fit model, minimum of Akaike Information Criterion (AIC) and the Schwarz Criterion (SC) values (Alberg et al., 2008).

Model Diagnostics

Diagnostic testing of the residuals in the fitted model is conducted to determine the model fit of GARCH (1,1) and TGARCH. The residuals in the model are checked through Ljung Box Q-statistics for autocorrelation, ARCH LM test for heteroscedasticity and Jarque–Bera test for normality (Bhatia & Gupta, 2020). Generally, the absence of the ARCH effect and autocorrelation in the residuals determines the good fit of the model applied for volatility persistence.

Risk-adjusted Return Performance Measures

Risk-adjusted return measures provide tools to compare different investments of a similar asset class. These measures prove to be useful to compare the performance of the portfolio as well as mutual funds. Sharpe, Treynor, Sortino, Information and Jensen’s Alpha is also applied to the valuation by Śliwiński and Łobza (2017), Cortez et al. (2009) and Statman (2006).

Markov Switching Model

We have applied the MS model to decompose the study period duration into a finite sequence of regimes. MS model has an edge to identify the market regimes precisely (Cheung et al., 2010). It has the advantage of allowing the time-dependent parameters to switch endogenously. It also identifies the market regimes using time series only (Krolzig, 1997). A Markov Chain is a process where switching among the states to the next state is dependent on the current state having the probability of P_ij.

P_{i j} = P (S_{t - 1} = i | S = j), \sum^{​} P_{i j} = 1 \forall i, j \in {1,,, M)

The P_ij indicates that the probability for a variable t S in the state i is followed by state j.

Each regime will be originated from the first-order Markov chain where S_t will be considered as unobserved. Following the studies of Guidolin and Hyde (2009) and Cheung et al. (2010), the regression model is estimated as:

Y_{t} = v_{s_{t}} + \sum_{k = 1}^{p} β_{k} \cdot X_{t - k} + \sum^{​} α z_{t} + u_{s_{t}}

Where,

Y_t = R_i – R_f (Excess return of said index)

$v_{s_{t}} =$ A vector of intercepts in state t

β_k = Market risk premium at lag k

X_t–k = Market risk premium

This regression model has been applied taking returns of SRI indices as dependent variables. To determine the efficient MS Model, a diagnostic check of stationarity and SIC test to decide optimal lag and regimes is conducted.

Analysis

Descriptive Statistics

Table 3 represents the descriptive properties of sustainability indices, Nifty100 ESG and Nifty100 Enhanced ESG, and market index, Nifty50. Nifty100 Enhanced ESG has the highest mean returns of 3.38% as compared to Nifty100 ESG having 3.3% while Nifty50 has the lowest mean returns of 2.8%. Standard deviation or total risk is not significantly different, implying that sustainability indices are performing better than a market index in terms of return with a similar level of risk. Negative values of skewness suggest that there are higher chances of a large decrease in the returns than the large increase. Meanwhile, the Kurtosis value greater than 3 suggests the leptokurtic distribution of returns of the said indices. Lastly, Jarque–Bera test probability values suggest that the returns follow non-normal distribution as the null hypothesis of returns following normal distribution is strongly rejected.

Results of Welch t-test

The results of the t-test are summarized in Table 4 along with test statistics and p value. The p value for all the pairs of mean returns of sustainability indices and benchmark index is >.05, that is, .871, .8642 and .9932. This gives enough evidence to accept the null hypothesis that there is no significant difference between mean returns of Nifty50, Nifty100 ESG and Nifty100 Enhanced ESG as the sustainability indices are the subset of the general benchmark index and hence is expected to exhibit superior or par performance.

Table 3.

Descriptive Statistics of Nifty100 ESG and Nifty100 Enhanced ESG and Nifty50

Variable	Nifty50	Nifty100 ESG	Nifty100 Enhanced ESG
Mean	0.0280	0.0335	0.0338
Standard Deviation	1.0420	1.0504	1.049
Skewness	−1.4661	−1.3091	−1.3170
Kurtosis	19.042	16.293	16.521
Jarque–Bera	29893	21930	22539
Probability	0.0000	0.0000	0.0000

Source: The authors.

Results of Unit Root Test

To test the unit root, the ADF test has been applied. The results are summarized in Table 5. The ADF value of the indices has a probability value of 0.01 which is less than a 5% level of significance. Henceforth, the null hypothesis of having unit root in the series is rejected. Therefore, the value of indices return is stationary at the level and there is no requirement of taking the first difference of the return series.

Table 4.

Results of Welch t-test

Index Variable	t-statistics	p Value
Nifty50 and Nifty100 ESG	−0.1623	.871
Nifty50 and Nifty100 Enhanced ESG	−0.1710	.864
Nifty100 ESG and Nifty100 Enhanced ESG	−0.0085	.993

Source: The authors.

Note: Insignificant at 5% level of significance.

Results of Overall Risk-Adjusted Measures of Return Performance

Assortment of risk-adjusted measures of return performance is used in the research study to explore how the sustainability indices are performing when concerning the market index, Nifty50. The results of all the measures used are summarized in Table 6. First, unsystematic risk is calculated as the difference between total risk in relation to market risk and systematic risk negligible in the market index (Nifty50) and Nifty100 ESG followed by a little lower unsystematic risk as compared to 0.0159 of Nifty100 ESG Enhanced. Systematic risk is inherent in the market index as a beta value stands at 1. Sustainability indices of NSE have the beta value 0.9937 and 0.9919, respectively, signifying they are marginally less sensitive to the market movements. Further, the Sharpe ratio is highest for Nifty100 ESG Enhanced (0.0134) followed by Nifty100 ESG (0.0130) and Nifty50 (0.0079), implying that sustainability indices are performing better in terms of excess return per unit of risk. Treynor ratio is consistently higher for Nifty100 ESG Enhanced (0.0141), followed by 0.0138 of Nifty100 ESG and lowest for benchmark market Index Nifty50. Comparing similar types of investments, investors usually prefer a higher Sortino ratio as it gives a return of an investment per unit of bad risk, Sortino ratio is highest for Nifty100 ESG Enhanced being 0.0163 and least for Nifty50 with a magnitude of 0.0095, favouring SRI. Further, the magnitude of Nifty100 ESG Enhanced index information is relatively higher than Nifty100 ESG by 0.0005 points implying the SRI index delivering better returns than the benchmark index.

Table 5.

Results of Unit Root Test

Index Return	ADF Value	p Value
Nifty50	−11.558	.01**
Nifty100 ESG	−11.804	.01**
Nifty100 Enhanced ESG	−11.772	.01**

Source: The authors.

Note: **Significant at 5% level of significance.

Table 6.

Results of Risk-Adjusted Measures of Return

Measures	Nifty50	Nifty100 ESG	Nifty100 Enhanced ESG
Unsystematic Risk	0.0000	0.01422	0.0159
Systematic Risk	1.0000	0.9937	0.9919
Sharpe Ratio	0.0079	0.01307	0.0134
Treynor Ratio	0.0083	0.0138	0.0141
Sortino Ratio	0.0095	0.0159	0.0163
Information Ratio	NA	0.0308	0.0313
Omega Ratio	1.0000	1.0000	1.0050
Jenson’s Alpha	0.0000	0.0055	0.0058
M2 Model	0.0000	0.0054	0.0057
FAMA Net Selectivity	0.0000	0.0054	0.0057

Source: The authors.

Omega ratio is higher for Nifty100 ESG Enhanced being 1.005, followed by Nifty100 ESG and Nifty50 being at 1.000. This implies considering standard deviation, kurtosis and skewness, the SRI index is delivering relatively higher returns. Analysing the results of the study concerning Jensen’s Alpha, it is 0.000 for market index and relatively higher for Nifty100 ESG Enhanced with a value being 0.0058 signifying that this index is ‘beating the market’ with manager’s stock selecting skills. The M-Squared measure is highest consistently for Nifty100 ESG Enhanced with 0.57% implying the better performance of Nifty100 ESG Enhanced in terms of mimicking the market portfolio. Lastly, Fama’s Net Selectivity measure is also in favour of sustainability indices, being 0.0057, highest for Nifty100 ESG Enhanced followed by Nifty100 ESG index implying SRI indices outperforming market portfolio.

Visualization of Volatility Clustering

Visualizing the Figures 1, 2 and 3, it is observed that there is a presence of the ARCH effect which is considered to be a pre-requisite of applying the GARCH models.

Results of ARCH LM Test

Results of the ARCH LM Test are summarized in Table 7. Index variables are having a probability value of 0.0000 which is significantly less than 0.05, level of significance. Hence, it can be concluded that the null hypothesis of no ARCH effect is rejected, confirming the presence of the ARCH effect empirically.

Therefore, we have all pre-conditions of volatility clustering and the presence of ARCH effect satisfied, and it is justified to further apply GARCH and TGARCH models. Table 8 comprises parameter estimates of the GARCH model. The coefficients of the GARCH model of sustainability indices and market are statistically significant (p value < .05) suggesting the strong validity of the model. The sum of the coefficient (α + β) is close to unity, that is, 0.98, 0.97 and 0.97 for Nifty50, Nifty100 ESG and Nifty100 ESG Enhanced, respectively, indicating volatility clustering. This leads to the conclusion that large differences in the return will be followed by large changes in return whereas small changes in return will be followed by small differences. The residual diagnostics show that the standardized residuals are not autocorrelated (Q-stat; p values > .05) for Nifty50, Nifty100 ESG and Nifty100 Enhanced ESG.

Table 7.

Arch LM Test

Index Variables	Chi-Square Value	p Value
Nifty50	821.75	.0000**
Nifty100 ESG	767.76	.0000**
Nifty100 ESG Enhanced	770.5	.0000**

Source: The authors.

Note: **Significant at 5% level of significance.

Also, for a model to be deemed as fit, there should not be an ARCH effect in the residuals. The residuals in the GARCH (1,1) are tested for the ARCH effect. It is seen that ARCH p values are insignificant; hence, there is a piece of strong evidence to accept the null hypothesis of no ARCH effect in the residuals. The Jarque–Bera test statistics also reflect the non-normal distribution of residuals, hence indicating the presence of asymmetric effects. The diagnostics reveal that GARCH (1,1) model may be a good fit to capture the volatility during the time duration. However, to check the leverage effect, TGARCH is also applied.

Table 8.

GARCH Model Results

Parameters	Nifty50		Nifty100 ESG		Nifty100 Enhanced ESG
Parameters	Estimates	p Value	Estimates	p Value	Estimates	p Value
Μ	0.0673	.00034**	0.0711	.00023**	0.0723	.00017**
Ω	0.0220	.00203**	0.0246	.00252**	0.0236	.00280**
α₁	0.0904	.00000***	0.0918	.00000***	0.0917	.00000***
β₁	0.8908	.00000**	0.8876	.00000***	0.8888	.00000***
AIC	2.6228		2.6734		2.6671
SC	2.6343		2.6849		2.6787
Model Diagnostics
Q-stat	0.9560	.3285	0.5728	.4499	1.1783	.2786
ARCH-LM	0.9500	.3289	0.5712	.4498	1.1755	.2784
Jarque-Bera	1,881.7788	.0000**	1,382.1877	.0000**	1,242.3567	.0000**

Source: The authors.

Note: **Significant at 5% level of significance; ***Significant at 1% level of significance.

Since vanilla GARCH(1,1) is unable to capture the leverage effect, the TGARCH model is applied to test for ‘leverage effect’. Results are presented in Table 9. This model expects gamma coefficient γ >0 to validate leverage effect. Positive and significant values of γ, Gamma coefficient in TGARCH model estimates confirm the leverage effect in sustainability indices, Nifty100 ESG (0.1623), Nifty100 Enhanced ESG (0.1609) and market index Nifty50 (0.165), implying that volatility responds more to bad news than changes in the returns due to good news in the market. The residual diagnostics demonstrate that residuals from the TGARCH model are not autocorrelated (Q-stat; p values > .05) and heteroscedastic (ARCH-LM; p values > .05). The diagnostics confirm that TGARCH is a good fit model to capture leverage effect for the given time duration.

Table 9.

TGARCH Model Results

Parameters	Nifty50		Nifty100 ESG		Nifty100 Enhanced ESG
Parameters	Estimates	p Value	Estimates	p Value	Estimates	p Value
Μ	0.03825	.03788**	0.0421	.02725**	0.0437	.02111**
Ω	0.02755	.00002**	0.0301	.00005**	0.0292	.00005**
α₁	0.0000	.99999	0.0000	.99999	0.0000	.99999
β₁	0.8905	.00000***	0.8897	.00000***	0.8913	.00000***
Γ	0.1652	.00000***	0.1623	.00000***	0.1609	.00000***
AIC	2.5791		2.6330		2.6271
SC	2.5936		2.6475		2.6415
Model Diagnostics
Q-stat	0.3968	.529	1.1546	.283	1.1185	.290
ARCH-LM	0.3956	.5294	1.1517	.2833	1.1157	.2910

Source: The authors.

Notes: The findings are in contrast with the findings of Onwukwe et al. (2011) and Goudarzi and Ramanarayanan (2011).

**Significant at 5% level of significance; ***Significant at 1% level of significance.

Results of Markov Model Selection Process

SIC test has been used to decide on the appropriate MS Model based on the number of regimes and appropriate lag length. SIC test is selected over AIC and log-likelihood due to its property of stronger penalty on parameters, hence making it more efficient (Sardy, 2009). Regression has been run by adjusting the number of regimes and lag length to find the best model based on the lowest SIC criterion. Table 10 comprises of Markov Model selected for the sustainability indices. Both Nifty100 ESG and Nifty100 Enhanced ESG have a Markov model of two regimes with two lag lengths. Regime Switching Models can be extended to k regimes, but the two regimes of variance may identify the periods of low and high volatility (Bazzi et al., 2017). Hence, the two-state model has been chosen for the identification of states.

Table 10.

MS Models for Each Index

Index	Benchmark	MSM	SIC
Nifty100 ESG	Nifty50	MSM(2)-(2)	−0.682
Nifty100 Enhanced ESG	Nifty50	MSM(2)-(2)	−0.595

Source: The authors.

Note: The first digit in the bracket signifies the number of regimes and the second digit represents the number of lags.

Return and Risk Characteristics of Sustainability Indices in Different Regimes

It can be seen from Table 11 that sustainability indices, Nifty100 ESG and Nifty100 Enhanced ESG outperform the benchmark index, Nifty50 in both the market regimes with the mean returns of 0.028 and 0.027 as against 0.0244. Regime 1 is characterized with the higher mean returns as compared to mean returns 0.0199, 0.0247 and 0.021 of Nifty50, Nifty100 ESG and Nifty100 ESG Enhanced. This implies that sustainability indices are performing marginally better relative to their benchmark, hence, can be considered as an investment option by potential investors.

Table 11.

Risk and Return Characteristics across Different Market Regimes

		Mean Returns	Standard Error	p Value	Beta Coefficient	Standard Error	p Value
Nifty50	Regime 1	0.0244	0.0006	.0000**	0.9987	0.0008	.0000**
	Regime 2	0.0199	0.0066	.0002**	1.000	0.0033	.0000**
Nifty100 ESG	Regime 1	0.02827	0.0145	.0521	0.9810	0.0106	.0000**
	Regime 2	0.02475	0.0040	.0000**	1.0035	0.0067	.0000**
Nifty100 Enhanced ESG	Regime 1	0.02712	0.0042	.0000**	0.9837	0.0069	.0000**
Nifty100 Enhanced ESG	Regime 2	0.0218	0.0144	.1302	0.9971	0.0101	.0000**

Source: The authors.

Note: **Significant at 5% level of significance.

Moving further to beta coefficients as a measure of systematic risk, for sustainability indices Nifty100 ESG and Nifty100 Enhanced ESG, systematic risks increasing from regime 1 are 0.981 and 0.983, respectively. Whereas, for market benchmark Nifty50, systematic risk is highest in regime 2 with the beta coefficient being 0.9987. This implies that sustainability indices perform marginally better in regime 1. Since regime 2 is characterized by low return and high risk, the sustainability as well as market index systematic risk increase in regime 2.

Results of Transition Probabilities across Market Regimes

From Table 12 given, it can be seen that Regime 1 is the most persistent regime which is featured with high returns and low volatility with probabilities being 0.989, 0.985 and 0.98, respectively, for market benchmark index (Nifty50) and sustainability indices, Nifty100 ESG and Nifty100 Enhanced ESG. Whereas the transition probability from Regime 1 to Regime 2 is very low specifically 0.12, 0.05 and 0.06, respectively, for market and sustainability indices. This indicates that for sustainability indices, higher returns and low volatility are persistent for a long time (Figure 4); hence, a benefit to investors to invest in sustainable companies aligned with the ESG principles to reap reasonable returns.

Table 12.

Transition Probabilities (in%) of Market Regimes

		Regime 1	Regime 2	Constant Expected Duration
Nifty50	Regime 1	0.9897	0.0102	97.1095
	Regime 2	0.1218	0.8781	08.2056
Nifty100 ESG	Regime 1	0.9853	0.0146	68.3646
	Regime 2	0.0514	0.9485	19.4456
Nifty100 Enhanced ESG	Regime 1	0.9802	0.0197	50.5492
	Regime 2	0.06350	0.9364	15.7475

Source: The authors.

Discussion and Conclusion

In recent years, the relevance of establishing sustainability principles, ESG disclosures and awareness towards contributing to SDGs has increased manifold. Extant literature is being contributed to analysing the performance of SRI avenues in the context of the Indian stock market. Indian stock markets have made a further contribution to this trend by introducing sustainability-themed indices as the investors become more aware of SRI. These indices have a proven track record to monitor the performance of companies based on ESG scores. Lately, two more such types of indices were introduced by NSE in the year 2017; Nifty100 ESG and Nifty100 Enhanced ESG. With new indices being added to the basket of sustainability indices, there is a need to examine its performance in the Indian stock market. A study on the comparative analysis of the performance will help investors become aware of new options of investing in ESG and take better decisions.

This research study is undertaken to investigate two main issues: First, to explore the overall risk and return performance of sustainability indices of NSE, that is, Nifty100 ESG and Nifty100 Enhanced ESG concerning market benchmark index, Nifty50. Due to SRI still being at its nascent stage in India, investors are a little sceptical about the volatility in the returns of ESG indices. Past studies have also reported that there exists a cost of responsible investing in terms of restricted investment universe (Adler & Kritzman, 2008). So, there was a need to examine the variations in the returns in the form of volatility from ESG companies. Hence, the second objective was to model the inherent conditional volatility in the SRI indices and see if there exists any leveraging effect through the TGARCH model. Lastly, the aim was to decompose the single period of the study into distinct regimes by applying the Markov Switching regression model. It further finds out the regimes of high and low volatility, respectively.

The findings of empirical results suggest that the returns of sustainability indices (0.033) are marginally higher than the market index (0.028). Nevertheless, there is no significant difference between the daily mean returns of sustainability and conventional indices. The beta values lower than one also suggest Nifty100 ESG (0.993) and Nifty100 ESG Enhanced (0.991) are less sensitive to the market counterpart Nifty50. All the measures of risk-adjusted return report higher returns from sustainability indices. This hints that SRI avenues in India are providing reasonable returns to investors. Hence, it is not risky to invest in sustainability indices. The results are consistent with the findings of Vasal (2009) and Śliwiński and Łobza (2017). The findings of these researches were indicative of a marginally higher risk-adjusted performance of ESG indices. Notably, both the SRI indices and the market index were reported to be stationary at a level, hence ruling out any spurious relationship. This paves the way for SRI investment thus achieving long-term sustainability.

On the volatility front, the results of the TGARCH model estimates that there is long-run persistent volatility in Nifty100 ESG and Nifty100 Enhanced ESG with the sum of α and β close to 1. As shown by γ for all three indices, the leverage effect is also demonstrated stating that negative news leads to higher volatility than the impact of positive information. Sudha (2014) and Tripathi and Kaur (2020) also came up with an idea of volatility persistence in SRI indices using GARCH (1,1) models and have concluded that there exists persistence volatility in the SRI indices. Model diagnostics was performed using Ljung Box Q statistics and ARCH LM test for testing auto correlation and ARCH effect in the residuals. Diagnostic tests also confirmed the absence of autocorrelation and ARCH effect in the residuals. Hence, it can be concluded that TGARCH was a good fit model to measure the leverage effect in the returns.

Our study has extended the previous studies of Albaity and Ahmad (2011) and Ti et al. (2019) through the application of the Markov Switching regression model. This model is prevalent in assessing the existence of two distinct regimes in equity indices and the market portfolio. Markov Switching regression has been applied to determine two distinct regimes of high or low volatility, expected duration and probability of each regime. It is observed that there are two distinct regimes, with the first regime of high returns and low volatility and the second regime of low returns and high volatility. Sustainability indices performed better in terms of return and systematic risk in both regimes relative to the market benchmark index. It was also observed that regime 1 indicating high returns and low volatility is the most persistent one as it has the highest probability (0.98) of remaining in regime 1 only. Findings are consistent with similar research works by Reboredo (2010) and He et al. (2018). This implies that investors can generate reasonable returns through investment in sustainable companies for the long term. These results can help investors and portfolio managers to make informed decisions while including ESG and SRI as criteria in their decision-making process.

Implications

The current study evaluates the performance of SRI indices vis-a-vis the conventional benchmark. This research puts forth implications for several stakeholders of the society. First, this work is aimed to benefit the investors who are keen to align their non-financial objectives and social values without compromising the financial returns. Results reveal that Indian investors are not losing by investing in socially responsible companies. They are reaping reasonable returns with a marginally less amount of risk as compared to the market index along with satisfying their environmental, social and ethical concerns. Second, this study has practical implications for asset managers. Managers can contribute to ethical investing by suggesting several sustainability-themed portfolios to the clients who are willing to achieve social goals along with financial goals. This will lead to the inclusion of SRI as a criterion for small investors in India. Third, the results can benefit the corporates as it is implied that companies with high ESG scores can take an advantage over peers by attracting capital at a lower cost from domestic and foreign investors. This will lead to raising the bar of good corporate governance in India. Lastly, adhering to CSR regulations, sustainability reporting and following ESG disclosures will help investors to make better-informed decisions. Our findings corroborate the fact that SRI in India should not be considered a niche market and should be used as mainstream investing by investors and asset managers. Positive results of the study also suggest that regulatory bodies, stock exchanges along the government can encourage disclosure practices by designing a sustainability reporting framework to facilitate the growth of SRI in India. Aligning the business models with the SDGs will help in making India a sustainable economy.

Scope for Further Research

The current research study has included the market index, Nifty50 as the only benchmark index. However, other market indices like Nifty100 can also be used. Besides overall indices, there are varied socially responsible investment avenues available in India. Further areas of future research can include the performance comparison of SRI stock portfolios with the conventional portfolio of general companies. Few researches have revealed that expected returns based on ESG can be different for different sectors and industries. So, the performance of sector-specific SRI portfolios can be investigated to get more robust results based on sector-specific characteristics. Several new ESG (Quantum India ESG mutual fund and SBI Magnum Equity ESG fund and many more) and ethical mutual funds are introduced in India having the companies following ESG criteria. Hence, a comparative performance analysis of ESG mutual funds can be undertaken with conventional funds. Along with India, other emerging economies are also including ESG as a criterion in mainstream investing. Thus, a cross-country comparison of ESG performance can be conducted to get a better understanding of country differences in sustainable investment options.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

References

Abramson

, & Chung

(2000). Socially responsible investing: Viable for value investors? The Journal of Investing, 9(3), 73–80.

Acar

, & Temiz

(2020). Empirical analysis on corporate environmental performance and environmental disclosure in an emerging market context. International Journal of Emerging Markets. https://doi.org/10.1108/IJOEM-04-2019-0255

Adler

, & Kritzman

(2008). The cost of socially responsible investing. The Journal of Portfolio Management, 35(1), 52–56.

Albaity

, & Ahmad

(2011). Return performance and leverage effect in Islamic and socially responsible stock indices evidence from Dow Jones (DJ) and Financial Times Stock Exchange (FTSE). African Journal of Business Management, 5(16), 6927–6939.

Alberg

, Shalit

, & Yosef

(2008). Estimating stock market volatility using asymmetric GARCH models. Applied Financial Economics, 18(15), 1201–1208.

Angelidis

, Sakkas

, & Tessaromatis

(2015). Stock market dispersion, the business cycle and expected factor returns. Journal of Banking & Finance, 59, 265–279.

Awartani

B. M.

, & Corradi

(2005). Predicting the volatility of the S&P-500 stock index via GARCH models: The role of asymmetries. International Journal of Forecasting, 21(1), 167–183.

Azmi

, Ng

, Dewandaru

, & Nagayev

(2019). Doing well while doing good: The case of Islamic and sustainability equity investing. Borsa Istanbul Review, 19(3), 207–218.

Basu

, & Chawla

(2010). An empirical test of CAPM—The case of Indian stock market. Global Business Review, 11(2), 209–220.

10.

Bauer

, Koedijk

, & Otten

(2005). International evidence on ethical mutual fund performance and investment style. Journal of Banking & Finance, 29(7), 1751–1767.

11.

Bazzi

, Blasques

, Koopman

S. J.

, & Lucas

(2017). Time-varying transition probabilities for Markov regime-switching models. Journal of Time Series Analysis, 38(3), 458–478.

12.

Becchetti

, & Ciciretti

(2009). Corporate social responsibility and stock market performance. Applied Financial Economics, 19(16), 1283–1293.

13.

Benson

C. C.

, Gupta

, & Mateti

(2010). Does risk reduction mitigate the costs of going green?—An empirical study of sustainable investing. An Empirical Study of Sustainable Investing, 2, 7–25.

14.

Bertrand

, & Lapointe

(2015). How performance of risk-based strategies is modified by socially responsible investment universe? International Review of Financial Analysis, 38, 175–190.

15.

Bhatia

, & Gupta

(2020). Sub-prime crisis or COVID-19: A comparative analysis of volatility in Indian banking sectoral indices. FIIB Business Review, 9(4), 286–299.

16.

Bhuian

, & Sharma

S. K.

(2017). Predicting consumer pro-environmental behavioral intention: The moderating role of religiosity. Review of International Business and Strategy, 27(3), 352–368.

17.

BinMahfouz

, & Hassan

M. K.

(2013). Sustainable and socially responsible investing. Humanomics, 29(3), 164–186.

18.

Bollerslev

(1986). Generalized autoregressive conditional heteroskedasticity. Journal of Econometrics, 31(3), 307–327.

19.

Borgers

, Derwall

, Koedijk

, & Ter Horst

(2015). Do social factors influence investment behavior and performance? Evidence from mutual fund holdings. Journal of Banking & Finance, 60, 112–126.

20.

Changyong

F. E. N. G.

, Hongyue

W. A. N. G.

, Naiji

L. U.

, Tian

C. H. E. N.

, Hua

H. E.

, & Ying

L. U.

(2014). Log-transformation and its implications for data analysis. Shanghai Archives of Psychiatry, 26(2), 105.

21.

Charfeddine

, Najah

, & Teulon

(2016). Socially responsible investing and Islamic funds: New perspectives for portfolio allocation. Research in International Business and Finance, 36, 351–361.

22.

Cheung

W. A.

, Li

, & Roca

(2010). The performance of socially responsible investments across different market regimes (No. finance: 201018). Griffith University, Department of Accounting, Finance and Economics.

23.

Clark

, Deshmukh

, & Belghitar

(2014, June). Does active ethical investing pay–evidence from the UK. In 21st Annual Conference of the Multinational Finance Society.

24.

Coffie

(2015). Modelling and forecasting the conditional heteroscedasticity of stock returns using asymmetric models: Empirical evidence from Ghana and Nigeria. Journal of Accounting & Finance, 15(5), 2158–3625.

25.

Collison

D. J.

, Cobb

, Power

D. M.

, & Stevenson

L. A.

(2008). The financial performance of the FTSE4Good indices. Corporate Social Responsibility and Environmental Management, 15(1), 14–28.

26.

Cortez

M. C.

, Silva

, & Areal

(2009). The performance of European socially responsible funds. Journal of Business Ethics, 87(4), 573–588.

27.

Dalal

K. K.

, & Thaker

(2019). ESG and corporate financial performance: A panel study of Indian companies. IUP Journal of Corporate Governance, 18(1), 44–59.

28.

de Souza Cunha

F. A. F.

, & Samanez

C. P.

(2013). Performance analysis of sustainable investments in the Brazilian stock market: A study about the corporate sustainability index (ISE). Journal of Business Ethics, 117(1), 19–36.

29.

Emenike

K. O.

(2010, February 10). Modelling stock returns volatility in Nigeria using GARCH models. Proceeding of International Conference on Management and Enterprice Development, Ebitimi Banigo Auditorium, University of Port Harcourt - Nigeria, 1(4), 5–11.

30.

Entine

(2003). The myth of social investing: A critique of its practice and consequences for corporate social performance research. Organization & Environment, 16(3), 352–368.

31.

Engle

R. F.

(1982). Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation. Econometrica: Journal of the Econometric Society, 50(4), 987–1007.

32.

Galema

, Plantinga

, & Scholtens

(2008). The stocks at stake: Return and risk in socially responsible investment. Journal of Banking & Finance, 32(12), 2646–2654.

33.

Goldreyer

E. F.

, & Diltz

J. D.

(1999). The performance of socially responsible mutual funds: Incorporating socio-political information in portfolio selection. Managerial Finance, 25(1), 23–36.

34.

Goudarzi

, & Ramanarayanan

C. S.

(2011). Modeling asymmetric volatility in the Indian stock market. International Journal of Business and Management, 6(3), 221–231.

35.

Guidolin

, & Hyde

(2009). What tames the Celtic tiger? Portfolio implications from a multivariate Markov Switching model. Applied Financial Economics, 19(6), 463–488.

36.

Gujarati

D. N.

, & Sangeetha. (2007). Basic econometrics. Tata McGraw Hill.

37.

Gupta

(2011). Socially responsible investing in India: Myth or reality (Working paper). Oikos PRI Young Scholars Academy 2011: The future of responsible investment.

38.

Hafner

C. M.

, & Herwartz

(2000). Testing for linear autoregressive dynamics under heteroskedasticity. The Econometrics Journal, 3(2), 177–197.

39.

Hansen

P. R.

, & Lunde

(2005). A forecast comparison of volatility models: Does anything beat a GARCH (1, 1)? Journal of Applied Econometrics, 20(7), 873–889.

40.

, O’Connor

, & Thijssen

(2018). Is gold a sometime safe haven or an always hedge for equity investors? A Markov-switching CAPM approach for US and UK stock indices. International Review of Financial Analysis, 60, 30–37.

41.

Hoti

, McAleer

, & Pauwels

L. L.

(2005). Modelling environmental risk. Environmental Modelling & Software, 20(10), 1289–1298.

42.

Jain

, Sharma

G. D.

, & Srivastava

(2019). Can sustainable investment yield better financial returns: A comparative study of ESG indices and MSCI indices. Risks, 7(1), 15.

43.

Jaisinghani

, & Sekhon

A. K.

(2020). CSR disclosures and profit persistence: Evidence from India. International Journal of Emerging Markets. https://doi.org/10.1108/IJOEM-03-2020-0246

44.

Krolzig

H. M.

(1997). The Markov-switching vector autoregressive model. In Markov-Switching Vector Autoregressions (pp. 6–28). Springer.

45.

Lean

H. H.

, Ang

W. R.

, & Smyth

(2015). Performance and performance persistence of socially responsible investment funds in Europe and North America. The North American Journal of Economics and Finance, 34, 254–266.

46.

Lean

H. H.

, & Nguyen

D. K.

(2014). Policy uncertainty and performance characteristics of sustainable investments across regions around the global financial crisis. Applied Financial Economics, 24(21), 1367–1373.

47.

, Cheung

, & Roca

(2010, August). Socially responsible investment, good and bad times. In 23rd Australasian Finance and Banking Conference.

48.

Lin

(2018). Modelling and forecasting the stock market volatility of SSE Composite Index using GARCH models. Future Generation Computer Systems, 79, 960–972.

49.

Lütkepohl

, & Xu

(2012). The role of the log transformation in forecasting economic variables. Empirical Economics, 42(3), 619–638.

50.

Managi

, Okimoto

, & Matsuda

(2012). Do socially responsible investment indexes outperform conventional indexes? Applied Financial Economics, 22(18), 1511–1527.

51.

Mandelbrot

B. B.

(1997). The variation of certain speculative prices. In Fractals and scaling in finance (pp. 371–418). Springer.

52.

Niyitegeka

, & Tewar

D. D.

(2013). Volatility clustering at the Johannesburg stock exchange: Investigation and analysis. Mediterranean Journal of Social Sciences, 4(14), 621.

53.

Onwukwe

C. E.

, Bassey

B. E. E.

, & Isaac

I. O.

(2011). On modeling the volatility of Nigerian stock returns using GARCH models. Journal of Mathematics Research, 3(4), 31–43.

54.

Ortas

, Moneva

J. M.

, & Salvador

(2012). Does socially responsible investment equity indexes in emerging markets pay off? Evidence from Brazil. Emerging Markets Review, 13(4), 581–597.

55.

Paul

(2017). The effect of business cycle, market return and momentum on financial performance of socially responsible investing mutual funds. Social Responsibility Journal, 13(3), 513–528.

56.

Reboredo

J. C.

(2010). Nonlinear effects of oil shocks on stock returns: A Markov-switching approach. Applied Economics, 42(29), 3735–3744.

57.

Rehman

R. U.

, Zhang

, Uppal

, Cullinan

, & Naseem

M. A.

(2016). Are environmental social governance equity indices a better choice for investors? An Asian perspective. Business Ethics: A European Review, 4(25), 440–459.

58.

Sabiruzzaman

, Huq

M. M.

, Beg

R. A.

, & Anwar

(2010). Modeling and forecasting trading volume index: GARCH versus TGARCH approach. The Quarterly Review of Economics and Finance, 50(2), 141–145.

59.

Sardy

(2009). Adaptive posterior mode estimation of a sparse sequence for model selection. Scandinavian Journal of Statistics, 36(4), 577–601.

60.

Sariannidis

, Giannarakis

, Litinas

, & Konteos

(2010). Α GARCH examination of macroeconomic effects on US stock market: A distinction between the total market index and the sustainability index. European Research Studies Journal, 13(1), 129–142.

61.

Sauer

D. A.

(1997). The impact of social-responsibility screens on investment performance: Evidence from the Domini 400 Social Index and Domini Equity Mutual Fund. Review of Financial Economics, 6(2), 137–149.

62.

Schröder

(2004). The performance of socially responsible investments: Investment funds and indices. Financial Markets and Portfolio Management, 18(2), 122–142.

63.

Schröder

(2007). Is there a difference? The performance characteristics of SRI equity indices. Journal of Business Finance & Accounting, 34(1–2), 331–348.

64.

Schueth

(2003). Socially responsible investing in the United States. Journal of Business Ethics, 43(3), 189–194.

65.

Shank

, Manullang

, & Hill

(2005). “Doing well while doing good” revisited: A study of socially responsible firms’ short-term versus long-term performance. Managerial Finance, 31(8), 33–46.

66.

Śliwiński

, & Łobza

(2017a). Financial performance of socially responsible indices. International Journal of Management and Economics, 53(1), 25–46.

67.

Śliwiński

, & Łobza

(2017b). The impact of global risk on the performance of socially responsible and conventional stock indices. Equilibrium. Quarterly Journal of Economics and Economic Policy, 12(4), 657–674.

68.

Statman

(2006). Socially responsible indexes. The Journal of Portfolio Management, 32(3), 100–109.

69.

Sudha

(2014). Risk-return and volatility analysis of sustainability index in India. Environment, Development and Sustainability, 17(6), 1329–1342. https://doi.org/10.1007/s10668-014-9608-8

70.

Talan

, & Sharma

G. D.

(2019). Doing well by doing good: A systematic review and research agenda for sustainable investment. Sustainability, 11(2), 353.

71.

, Du

, & Zhang

(2019, December). Analysis on the volatility of sustainable stock index and traditional stock index based on GARCH model. In 2019 International Conference on Economic Management and Model, Engineering (ICEMME) (pp. 47–50). IEEE.

72.

Tripathi

, & Bhandari

(2014). Socially responsible investing—An emerging concept in investment management. FIIB Business Review, 3(4), 16–30.

73.

Tripathi

, & Kaur

(2020). Socially responsible investing: Performance evaluation of BRICS nations. Journal of Advances in Management Research, 17(4), 525–547.

74.

Vasal

V. K.

(2009). Corporate social responsibility & shareholder returns—Evidence from the Indian Capital Market. Indian Journal of Industrial Relations, 44(3), 376–385.

75.

Yadav

M. P.

, & Sinha

(2021). Investigating the impact of corporate social responsibility on competitive performance: An empirical study based on panel data analysis. FIIB Business Review. https://doi.org/10.1177/23197145211015443

76.

Yavas

B. F.

, & Rezayat

(2016). Country ETF returns and volatility spillovers in emerging stock markets, Europe and USA. International Journal of Emerging Markets, 11(3), 419–437.