The role of ESG-based assets in generating the dynamic optimal portfolio in Indonesia

Asih, Kiki Nindya; Achsani, Noer Azam; Novianti, Tanti; Manurung, Adler Haymans Article The role of ESG-based assets in generating the dynamic optimal portfolio in Indonesia Cogent Business & Management Provided in Cooperation with: Taylor & Francis Group Suggested Citation: Asih, Kiki Nindya; Achsani, Noer Azam; Novianti, Tanti; Manurung, Adler Haymans (2024) : The role of ESG-based assets in generating the dynamic optimal portfolio in Indonesia, Cogent Business & Management, ISSN 2331-1975, Taylor & Francis, Abingdon, Vol. 11, Iss. 1, pp. 1-22, https://doi.org/10.1080/23311975.2024.2382919 This Version is available at: https://hdl.handle.net/10419/326457 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Cogent Business & Management ISSN: 2331-1975 (Online) Journal homepage: www.tandfonline.com/journals/oabm20 The role of ESG-based assets in generating the dynamic optimal portfolio in Indonesia Kiki Nindya Asih, Noer Azam Achsani, Tanti Novianti & Adler Haymans Manurung To cite this article: Kiki Nindya Asih, Noer Azam Achsani, Tanti Novianti & Adler Haymans Manurung (2024) The role of ESG-based assets in generating the dynamic optimal portfolio in Indonesia, Cogent Business & Management, 11:1, 2382919, DOI: 10.1080/23311975.2024.2382919 To link to this article: https://doi.org/10.1080/23311975.2024.2382919 © 2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group View supplementary material Published online: 29 Jul 2024. Submit your article to this journal Article views: 1169 View related articles View Crossmark data Citing articles: 1 View citing articles Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=oabm20 Banking & Finance | ReseaRch aRticle Cogent Business & ManageMent 2024, VoL. 11, no. 1, 2382919 The role of ESG-based assets in generating the dynamic optimal portfolio in Indonesia kiki nindya asiha , noer azam achsania , tanti noviantia and adler haymans Manurungb aschool of Business, iPB university, Bogor, indonesia; bBhayangkara university, Bekasi, indonesia ABSTRACT the aim of this study is to evaluate the role of indonesia’s esg-based equities (represented by etF sRi-kehati) in achieving optimal portfolio diversification with different asset classes, i.e. conventional assets (equities and government bonds), safe-haven assets (gold), commodities (crude oil) and digital assets (Bitcoin). the data used in the study was the daily return of each asset from January 2018 to December 2023, representing data from before, during and the recovery after the cOViD-19 pandemic. the methodology consists of a four-step process using the Pruned exact linear time algorithm, the Dcc-gaRch model and quadratic programming optimization. the study implies that investment managers and investors can reduce investment risk and balance their portfolios by including esg-based assets in their portfolios. the study also urges investment managers to more actively manage their portfolios through rebalancing strategies. the results could also encourage policymakers to continue to support sustainable financing initiatives. 1. Introduction in recent decades, much research has focused on sustainable and esg-based investments. this has been driven by various factors, firstly the increasingly evident impacts of climate change, coupled with an increased focus on environmental ethics such as reducing carbon emissions, green energy and green technology, which has led to increasing attention from various stakeholders towards esg issues (Deloitte, 2020; indriastuti & chariri, 2021). One of the stakeholders is the institutional investors who are taking measures to add sRi and esg investments to their portfolio radar, accelerated by the cOViD-19 pandemic, which encourages investors to do ‘the right thing’ (Morelli & nesta, 2021). the capital market plays an important role in providing a variety of new investment products related to sustainable responsible investing (sRi) and esg (Panagopoulos, 2023). esg-based investments are expected to remain high in the future and become a driving force for market growth. PWc predicts that esg assets under management will reach $34 trillion, or 21.5% of all global assets, by 2026 (almubarak et al., 2023). Meanwhile, Bloomberg estimates that esg asset management could reach more than $53 trillion by 2025, accounting for a third of global asset management, while remaining dominated by developed markets (Rumyantseva & tarutko, 2022). current research in sustainable investing encompasses a variety of focus areas. historically, sharma etal. (2022) showed that research in the 1980s and 1990s was predominantly concerned with themes of © 2024 the author(s). Published by informa uK Limited, trading as taylor & Francis group CONTACT Kiki nindya asih kikinindy[email protected].ac.id school of Business, iPB university, Padjajaran st., Bogor, West Java, 16151, indonesia. https://doi.org/10.1080/23311975.2024.2382919 this is an open access article distributed under the terms of the Creative Commons attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. the terms on which this article has been published allow the posting of the accepted Manuscript in a repository by the author(s) or with their consent. ARTICLE HISTORY Received 26 november 2023 Revised 14 May 2024 accepted 16 July 2024 REVIEWING EDITOR David McMillan, University of stirling, United kingdom of great Britain and northern ireland KEYWORDS cOViD-19; Dcc-gaRch; dynamic portfolio; esg-based equities; indonesia JEL CLASSIFICATION c22; g11; Q59 SUBJECTS investment & securities; environmental economics; econometrics; Finance 2 k. n. asih etal. personal values such as ‘sacrifice’, ‘morality’, and ‘religion’. Meanwhile, in the 2000s the focus shifted to more empirically oriented research, focusing on ‘activism’, ‘sustainability’, ‘stakeholders’, and ‘financial performance’. this study also showed that specific themes of sustainable and esg-based investments account for only 16 percent, while the rest is largely made up of socially responsible/ethical investments and responsible/impact investments. this decade, research on esg-based investing has focused primarily on performance and comparing it to traditional investing. Pedersen etal. (2021) propose a theory that demonstrates the potential costs and benefits of responsible investing to reconcile opposing views on whether esg promotes or detracts from performance. however, numerous gaps remain. the conflicting results and measurement issues regarding performance in empirical articles have been well documented by abhayawansa and Mooneeapen (2022). the general conclusion from the empirical literature is that there is no statistical difference in risk-adjusted performance between esg and conventional investing (erragraguy & Revelli, 2015; Friede etal., 2015; humphrey & lee, 2011). although most of the empirical literature demonstrates that there is no significant difference in returns, there is a small but persistent trend suggesting that actively managed esg portfolios can generate additional returns (sparkes & cowton, 2004). Despite a noticeable global shift towards sustainable practices, research on esg investing in emerging markets, particularly in asia (with the exception of Japan, south korea, taiwan and singapore), remains strikingly sparse, accounting for just 5% of the literature (sharma et al., 2022). there are only very limited studies for indonesia so far; One of them is the study by Robiyanto etal. (2023), which compares esg-based with conventional investments and shows that portfolios that take esg into account can reduce investment risk. the current trend towards sustainable and esg-based investments as well as the limited studies in asia provide the fundamental basis for the importance of this research, namely the diversifying role of esg investments in dynamic portfolio asset allocation and its optimal performance analysis. the aim of this study is to comprehensively shed light on the contribution of indonesian esg-based equities to optimal portfolio diversification. Broadly speaking, diversification allows investors to reduce their cumulative risk by integrating financial assets with typically weak positive correlations. Possible declines in one asset category may be somewhat offset by upward movements in others (Díaz et al., 2022). this study also contributes to the existing literature by providing a systematic approach to building a dynamic portfolio, which consists of combining indonesian esg-based assets with other different asset classes, while taking into account active management of esg portfolios in order to achieve a achieve optimal portfolio performance. thus, the present study extends the previous literature by raising two general research questions. First, to what extent has esg investing in indonesia contributed to optimal portfolio construction? second, does the performance of different types of asset classes improve when combined with local esg-based assets? to answer our research questions, we rely on daily data from six exchange-traded funds (etFs) tied to various assets from 2018 to 2023. the use of etFs offers several advantages, in line with Dutta et al. (2021) and Broadstock etal. (2021), while the time frame of the study was chosen to capture the instability associated with the outbreak of the cOViD-19 pandemic, we divided the data from 2018 to 2023 into three time windows, namely pre-cOViD-19, during cOViD-19, and recovery. the etFs in this study cover multiple asset classes. First, equities and government bonds are used as proxies for traditional investments in the equity and bond markets, respectively. second, gold is used to represent assets traditionally considered safe havens. thirdly, crude oil is a conventional commodity assets. Fourth, Bitcoin is considered a proxy for new trends in digital assets. Fifth, investments in the indonesian equity market that address environmental concerns, i.e. esg-based etFs, represented by the etF sustainable and Responsible investment-kehati (sRi-kehati). the etF consists of the sRi kehati equity index, which uses the principles of sustainability, finance, good governance and environmental concerns as benchmarks. the sRi-kehati equity index was launched by the indonesian stock exchange (iDX) in collaboration with Yayasan keanekaragaman hayati indonesia (kehati, 2017; kurniatama et al., 2021). to achieve the aim of this study, our methodological approach is to examine the dynamic correlation of optimal portfolios of alternative assets under consideration when esg-based etFs are included. Using appropriate techniques, we examine the ability of this etF asset to diversify and improve the performance of a cross-asset portfolio over time. the dynamic method will help us to grasp the tendency of stock fluctuations and somehow better simulate real situations according to the capital market dynamics cOgent BUsiness & ManageMent 3 (Ogata, 2012; Zinecker et al., 2016). We apply portfolio rebalancing with different frequencies based on the minimum variance optimization method, which allows us to evaluate the portfolios for both risk and return during three cOViD window periods. to model the conditional co-moments between each pair of asset returns, a Dcc-gaRch is used as it is proven to be able to successfully estimate the time-varying covariance matrices (Filis et al., 2011) and is very useful for formulation various dynamic portfolio rebalancing scenarios (Díaz et al., 2022). Various studies on the formation of cross-asset portfolios using a dynamic approach have been carried out before using esg-based assets in the indonesian capital market. these studies were mainly conducted by Robiyanto et al. (2021), which specifically examined the formation of cross-assets portfolio between esg-based equities and gold, and Robiyanto et al. (2023), which expands on previous studies to include cryptocurrencies, gold, and bonds. however, to the best of the author’s knowledge, there is no research on the paired portfolios between esg-based etFs (represented by etF sRi-kehati) and the individual other assets, namely blue-chip domestic equity etFs, gold and crude oil etFs, treasury etFs and Bitcoin etFs. the paired portfolio investments, which consist of only two assets, are often used in certain strategies such as hedging or pair trading. in these strategies, investors use the relationship between two specific assets to mitigate the risk or return of relative price movements. a paired portfolio compared to a broader cross-asset portfolio can provide a better understanding of asset correlation and is therefore more focused and suitable for specific trading or risk management objectives. additionally, this research provides portfolio rebalancing exercises that allow us to gain insight into whether actively managing esg portfolios provides better performance than passive investment in a single asset. the study found that first, from a risk perspective, the time-varying volatilities of the individual asset are much higher than the paired asset volatilities of esg-based etFs (represented by etF sRi-kehati) and any other asset, indicates that adding esg-based asset can diversify and reduce investment risk. second, from a performance perspective, the addition of esg-based assets improves portfolio performance for almost all asset classes used in this study, especially during the cOViD-19 pandemic, and suggests hedging functions of esg-based assets to maintain optimal performance in the turbulent times. third, the dynamic performance of each paired asset is consistent regardless of rebalancing scenarios and window times, demonstrating the robustness of the model. this study will contribute to the literature on cross-asset portfolios in terms of the importance of dynamic portfolio formulation and provide a better understanding of why investors support sustainable financing and why policymakers need to continue to facilitate and promote sustainable financing initiatives. the following sections of this study are organized as follows: the second section provides an overview of the literature. the third section describes the data and methods used. the fourth section presents the results, followed by discussions in the fifth section. the final section presents conclusions from the overall study. 2. Literature review Portfolio theory has undergone significant development. historically, its development can be traced from traditional portfolio theory (tPt) to postmodern portfolio theory (PMPt), with modern portfolio theory (MPt) serving as a crucial milestone (leković, 2021). tPt focuses on analyzing individual securities with a more subjective and non-systematic approach, making it less efficient at optimizing returns and risks. MPt, developed based on the pioneering ideas of individuals such as Fama (1996) and hakansson (1975), offers a more analytical and systematic approach. MPt emphasizes the importance of diversification and interaction between assets within a portfolio and aims to achieve the maximum expected return for any given level of risk (risk-adjusted return). this theory views the portfolio as a holistic entity and not just a collection of individual securities (shefrin & statman, 2000). a portfolio includes a variety of asset types or financial instruments that are used to diversify risk by dividing investments across a number of instruments (sinchukov, 2022). at its core, diversification aims to minimize risk in order to achieve optimal returns (Pourbabaee et al., 2016). asset allocation is about distributing investments across different asset classes, such as equities, bonds, and commodities that make up the portfolio (Bresnan & gelb, 1999). asset allocation, a cornerstone of portfolio management, navigates the complex landscape of financial markets using various methods (Manurung, 2016). strategic 4 k. n. asih etal. asset allocation, based on a buy-and-hold philosophy, formulates asset proportions based on long-term return expectations and risk tolerance, often based on fundamental analysis, and periodically rebalances these fixed proportions (Diris et al., 2015). When transitioning to tactical asset allocation, this approach provides flexibility by allowing temporary deviations from strategic allocation to optimize returns by exploiting short-term market opportunities while remaining within investment guidelines (Faber, 2017). at the same time, dynamic asset allocation, often driven by algorithmic models, continuously adapts to market trends and seeks to maximize the value of positive market changes and mitigate negative impacts through precise, timely asset reallocations, with the aim of improving risk-adjusted returns (chang etal., 2017). Portfolios constructed using different asset allocation strategies need to be rebalanced regularly to maintain their desired target performance (Boyante et al., 2022). a portfolio’s expected return is calculated as a weighted average of each stock’s expected return, adjusted for risk or fluctuations. a notable advantage of the MPt approach and its derivatives is their ability to design a portfolio with less risk than the total risk of the individual securities within it, thereby creating an optimal portfolio. Diversifying the assets in a portfolio should at least reduce risk or increase performance. Risk-adjusted return metrics are typically measured using treynor, sharpe, and alpha Jensen metrics (Manurung, 2016). according to the theory of risk-adjusted returns, esg-based investments can theoretically reduce risk by maximizing returns and thus act as diversification investments (giese et al., 2019). integrating esg aspects into portfolio management has attracted considerable attention due to its potential impact on returns and risk reduction. today, there are only a limited number of studies that provide empirical evidence on the diversifying properties of esg-based assets. arif et al. (2021) have uncovered a weak dynamic link between green and conventional stock indices over time, implying diversification opportunities for investors who prioritize long-term investment horizons, particularly using data from developed countries. these results support both Breitung and Breitung and candelon (2006) and sharma et al. (2022) confirmed evidence of bidirectional causality between sustainable and conventional indices in the short run, but did not observe the same in the long run, with the econometric analysis firmly based on data from developed nations. in other words, Ferrer etal. (2021) have shown that strategies combining clean energy stock indices and general stocks offer no diversification or hedging benefits, while clean energy stock and green bond could offer interesting diversification prospects, all underscored by data from advanced economies. Moreover, Yousaf et al. (2022) were unable to identify robust empirical evidence across various quantiles, designed to record extreme downturns in the s&P 500, to validate the safe-haven characteristic of clean energy stocks, even when they are integrated with positions in the s&P 500, underscoring the utilization of data derived exclusively from developed countries. Meanwhile, Díaz etal. (2022) concluded that sRi-based assets play a role in diversifying and improving the financial performance of investment portfolios with dynamic asset allocation strategies and certain rebalancing scenarios, especially when combined with different asset classes such as conventional equities, government bonds, gold, crude oil, and Bitcoin. Overall, the literature review showed that there is a paucity of research examining the contribution of esg investments in generating dynamic, optimal investment portfolio diversification, especially during the cOViD-19 pandemic in asian contexts, particularly in indonesia. 3. Data and methodology 3.1. Data the full data sample for each etF asset consists of time series data with approximately +/- 1300 observations from January 2018 to December 2023. the data was obtained from relevant market authorities, and it is available on the Bloomberg and Refinitiv eikon terminals, with all necessary rights and permissions duly secured for its use. the daily close price data is first transformed into log-returns using the following formula: r X X t t t =      − 252 1 *log (1) cOgent BUsiness & ManageMent 5 where rt represents the annualized log-return for all analyzed variables, Xt is the price index at period t, and X t− 1 denotes the price index from one day before. Local etF data was converted into the same currency unit (us dollar). the detailed data used in the analysis are available in table 1. an etF is a type of mutual fund that trades on exchanges, similar to a stock, and is designed to track an index, commodity, or compilation of assets, thereby providing broad investment exposure, typically with lower fees and better liquidity, thus enabling investors to purchase or sell shares throughout the trading day at market prices. in this study, both local and global etFs are used, first, equities and treasury bonds as a proxy for conventional investments in the equity and fixed-income markets i.e. blue-chip equity represented by etF lQ45 and bond etF (sPDR Bloomberg Barclays international treasury Bonds etF). second, gold (sPDR gold trust etF) is used to represent assets traditionally considered safe havens. third, crude oil (United states crude Oil etF) represents conventional commodity assets. Fourth, Bitcoin (grayscale Bitcoin trust etF) is considered a proxy for new trends in digital assets which is both risky and contemporary (abdullah et al., 2022). Fifth, investment in the local equity market that include environmental concerns, are represented by the sRi-kehati etF. 3.2. Methodology a four-step process is prepared to determine the contribution of esg investments in achieving optimal investment portfolio diversification, especially during the cOViD-19 pandemic in indonesia. First, we estimate three cOViD-19 window periods using the Pruned exact linear time (Pelt) algorithm. second, we calculate the various conditional volatilities, the conditional covariance matrix, and the conditional correlations for the entire sample period using the Dcc-gaRch model. third, we estimate the dynamic portfolio weights that minimize the variance of the respective portfolios from each asset type and the esg asset using a quadratic programming optimization. Fourth, we measure the portfolio returns of three cOViD-19 window periods using the sharpe ratio. to increase robustness, several rebalancing scenarios are considered. 3.2.1. Change points detection this research uses three different time periods: pre-cOViD-19, cOViD-19, and the recovery period of cOViD-19. to distinguish these periods, we use the Pelt algorithm developed by killick etal. (2012). this algorithm efficiently detects change points in the variance of daily returns with a linear computational cost (see appendix a). Based on the detection of these change points, three time windows were determined. the pre-cOViD-19 period extends from January 1, 2018 to February 26, 2020. the cOViD-19 period begins on February 27, 2020 and ends on February 3, 2021. Meanwhile, the recovery period begins on February 4. 2021 to December 31, 2023. see Figure 1 for more details on selecting these time slots. 3.2.2. Conditional volatilities, conditional covariance matrix, and conditional correlations the next step is to calculate the optimal asset allocation to put together the portfolio. One way to maximize returns while minimizing risk is through optimal asset allocation. this becomes increasingly relevant as we consider sustainable investing trends, such as portfolios focused on esg-based equities. to address this complexity, this research utilizes the Dcc-gaRch (Dynamic conditional correlation Table 1. Data. no etF ticker asset class Proxy exposure 1Premier etF sRi-KeHati XisR.JK equity esg equity Local 2Premier etF LQ-45 R-LQ45X.JK equity Bluechip equity Local 3sPDR gold trust gLD Commodity gold Global 4united states Crude oil etF uso Commodity Crude oil Global 5sPDR Bloomberg Barclays international treasury Bond etF BWX Bond government Bond Global 6grayscale Bitcoin trust gBtC Cryptocurrency Bitcoin Global this table reports etF data across various asset classes, including equities, commodities, bonds, and bitcoin. it details each etF’s name, ticker symbol, asset class, proxy, and geographic exposure. Data source: Bloomberg and Refinitiv eikon. 6 k. n. asih etal. generalized autoregressive conditional heteroscedasticity) model and quadratic programming to create a dynamic optimal portfolio. Dcc-gaRch, as an extension of the gaRch model, enables multivariate analysis and is extremely useful in capturing correlated movements between different types of assets, such as esg, with other assets like gold, crude oil, or even Bitcoin. this model allows us to calculate the conditional covariance matrix, which is a crucial input in quadratic programming to obtain optimal asset allocation. this model allows us to calculate the conditional covariance matrix, which is a crucial input in quadratic programming to achieve optimal asset allocation. this is crucial because the ideal asset allocation continually changes according to market dynamics captured by the Dcc-gaRch model (Orskaug, 2009). the basis for the formation of the Dcc-gaRch model is gaRch itself, and the gJR-gaRch model was selected in this study. the gJR-gaRch model was developed by glosten et al. (1993). the gJR-gaRch model is able to take into account asymmetric effects or leverage effects (situations in which volatility tends to be higher when asset prices are falling than when they are rising). the following is the conditional variance model of gJR-gaRch: σ ω αε βσ γ ε t j q j tj j p j tj j q tj tj S 2 1 2 1 2 1 2 =++ + = − = − = − − − ∑∑∑ (2) where Stj − − is a dummy variable that takes the value of 1 when ԑt−j is negative and 0 when ԑt−j is positive. the model states that positive volatility (ԑt > 0) and negative volatility (ԑt < 0) have different effects on conditional variance. the leverage effect is detected in the model when γ is positive (ngunyi etal., 2019). the specification of gJR-gaRch in this study is the standard gJR-gaRch (1,1), which provides computational efficiency and ease of interpretation while capturing the key features of asset return dynamics (Díaz et al., 2022). the conditional covariance matrix is calculated from the Dcc-gaRch model. engle (2002) and Orskaug (2009) state that the Dcc-gaRch model is derived from the following equations: rt tt = + µε (3) Figure 1. Changepoints on all assets. source: author calculations. this figure presents the changepoints of variance on all etF assets based on the PeLt algorithm. the determination of the research time window is subjectively based on the esg and LQ45 assets due to their exhibiting similar patterns of change. cOgent BUsiness & ManageMent 7 where ε t tt Hz =1 2/ rt = n × 1 matrix of returns from n variables used µ = n × 1 matrix of expected values from r ε = n × 1 matrix of time-varying residuals H = n × n matrix of conditional covariance of ε z = n × 1 matrix of residuals distributed normally with a mean of zero and constant variance the equation rt tt = + µε represents the mean equation of the Dcc-gaRch model. the residual consists of conditional covariance (Ht) and residuals with a zero mean and constant variance ( z t). conditional covariance is also illustrated in the following equation. this equation includes the conditional standard deviation (volatilities) and the conditional correlation of standardized residuals that change over time. H DRD where D t tt t t = = {} =      diag it t t it σ σ σ σ , , , , 1 2 00 0 0 00 ⋯ ⋱⋮ ⋮ ⋱⋱ ⋯        (4) note: ht = n × n matrix of conditional covariance at time t Dt = n × n matrix of conditional standard deviation at time t Rt = n × n matrix of conditional correlation at time t. 3.2.3. The dynamic optimal weight of assets to obtain the optimal asset allocation (dynamic optimal weights) with optimal returns at a tolerable level of risk for the best combination of the esg portfolio with other assets, quadratic programming can be used. this is achieved by minimizing portfolio variance (Díaz et al., 2022; Manurung et al., 2021; sahamkhadam etal., 2018). min H constraint: wPt t t t i n it t ww w σ 2 1 1 () =′ = = ∑ (5) where Ht represents the conditional covariance matrix of dimensions n × n and w t is the vector of dynamic optimal portfolio weights over time. the weights are subject to non-negativity constraints ( w t1, wt 2 … wit ≥ 0). the optimal weights obtained are dynamic, that is, they are optimal at time t due to the presence of a conditional covariance up to t. 3.2.4. Portfolio returns and performance measure the performance of each portfolio combination formed is then evaluated using the sharpe Ratio, a metric that allows us to evaluate risk-adjusted returns (sharpe, 1994): SR E rf r r r p p p () () () = − σ (6) where E rp () and σ rp () are the mean and standard deviation of the portfolio returns in the period under consideration. rf is the risk-free interest rate, which is obtained from the Bi 7-day reverse repo rate retrieved from Bank indonesia (central bank) website. this study also includes additional transaction scenarios to test the robustness of the research model. the scenarios are created whereby an investor changes the proportion or weight of each asset in the 14 k. n. asih etal. returns. When we combine esg-based equities with gold, we see an increase in risk-adjusted return 3.1 percentage points higher than just holding gold, and when we combine esg-based equities with government bonds, the risk-adjusted return increases by 5.7 Percentage points higher than simply holding gold government bonds. this points to the potential role of esg-based equities in improving the portfolio performance of the less risky assets. investing in Bitcoin generates the highest risk-adjusted return among individual asset classes, but its performance deteriorates slightly when assembled as a portfolio of esg-based equities. in this portfolio, esg-based equities still have a dominant weight (more than 80%), which means that esg-based equities not only help to hedge and mitigate the risks of investing in Bitcoin, but also somewhat preserves its performance value. in the case of oil, the individual investment produces a negative result. however, when combined with esg-based equities, the portfolio has a positive risk-adjusted return, indicating the ability of esg-based equities to absorb risks arising from the increased uncertainty in the commodity market during the pandemic. throughout the recovery period, only gold and crude oil investments generate a positive risk-adjusted return. however, when esg-based equities are included in both gold and crude oil, there are only minor changes in the risk-adjusted return. the combination of esg-based equities with lQ45 and esg-based Figure 7. Portfolio weights 22-day rebalance scenario. source: author calculations. this figure presents the weight evolution for minimum variance portfolios constructed by combining the etFs of each asset class and the etF of esg-based equities over the study periods for the 22-day rebalance under constrained portfolios. cOgent BUsiness & ManageMent 15 equities with government bonds performs in negative territory. On the contrary, combining esg-based equities with Bitcoin improved the risk-adjusted return from negative to positive, thereby confirming the hedging function of esg-based equities. tables 7 and 8, respectively, show the results of scenarios for 5-day and 22-day rebalancing. the performance of risk-adjusted returns for individual assets and in combination with esg-based equities produced relatively similar results compared to the 1-day rebalancing scenario. the consistency of these results confirms the robustness of our analysis. it is important to note that we do not take transaction costs into account when comparing investment risks in 1-, 5and 22-day rebalancing scenarios. Under normal circumstances, more frequent trading may result in higher transaction costs. Table 6. summary of the 1-day portfolio rebalancing scenario performance. Pre-CoViD-19 CoViD −19 Recovery LQ45 esg + LQ45 LQ45 esg + LQ45 LQ45 esg + LQ45 Mean 0.080 0.074 0.034 −0.127 −0.003 0.013 std.Dev 4.100 4.025 7.593 7.516 3.072 2.973 sharpe 0.016 0.058 0.002 −0.002 −0.006 −0.008 gold esg + gold gold esg + gold gold esg + gold Mean 0.109 0.068 0.141 0.194 0.045 0.056 std.Dev 1.845 1.635 3.530 3.212 2.455 2.237 sharpe 0.051 0.029 0.035 0.066 0.012 0.012 oil esg + oil oil esg + oil oil esg + oil Mean −0.089 −0.036 −0.995 −0.293 0.241 0.114 std.Dev 5.110 2.988 13.094 6.788 6.342 3.559 sharpe −0.021 −0.018 −0.077 0.011 0.035 0.028 Bond esg + Bond Bond esg + Bond Bond esg + Bond Mean −0.001 −0.007 0.087 0.061 −0.115 −0.098 std.Dev 0.886 0.875 2.072 2.180 1.796 1.719 sharpe −0.019 −0.017 0.034 0.091 −0.073 −0.098 Bitcoin esg + Bitcoin Bitcoin esg + Bitcoin Bitcoin esg + Bitcoin Mean −0.461 −0.102 1.664 0.128 −0.040 0.090 std.Dev 15.284 3.805 15.670 6.832 13.337 4.084 sharpe −0.031 −0.030 0.105 0.063 −0.004 0.019 source: author calculations. this table reports on the summary statistics of the observed returns (Mean), risk (std.Dev), and performance measure (sharpe) for the 1-day rebalancing scenario assessment on investment for single asset: LQ45, gold, crude oil, government bonds, and bitcoin as compared to paired assets: esg-based equities and LQ45, esg-based equities and gold, esg-based equities and crude oil, esg-based equities and government bonds, and esg-based equities and bitcoin. Table 7. summary of the 5-day portfolio rebalancing scenario performance. Pre-CoViD-19 CoViD −19 Recovery LQ45 esg + LQ45 LQ45 esg + LQ45 LQ45 esg + LQ45 Mean 0.080 0.076 0.034 −0.006 −0.003 0.001 std.Dev 4.100 4.027 7.593 7.350 3.072 2.960 sharpe 0.016 0.065 0.002 0.018 −0.006 −0.009 gold esg + gold gold esg + gold gold esg + gold Mean 0.109 0.078 0.141 0.203 0.045 0.053 std.Dev 1.845 1.646 3.530 3.264 2.455 2.240 sharpe 0.051 0.034 0.035 0.071 0.012 0.011 oil esg + oil oil esg + oil oil esg + oil Mean −0.089 −0.019 −0.995 −0.164 0.241 0.131 std.Dev 5.110 3.020 13.094 6.922 6.342 3.646 sharpe −0.021 −0.014 −0.077 0.033 0.035 0.036 Bond esg + Bond Bond esg + Bond Bond esg + Bond Mean −0.001 −0.004 0.087 0.096 −0.115 −0.103 std.Dev 0.886 0.874 2.072 2.219 1.796 1.725 sharpe −0.019 −0.014 0.034 0.097 −0.073 −0.100 Bitcoin esg + Bitcoin Bitcoin esg + Bitcoin Bitcoin esg + Bitcoin Mean −0.461 −0.084 1.664 0.153 −0.040 0.090 std.Dev 15.284 3.807 15.670 6.992 13.337 4.072 sharpe −0.031 −0.026 0.105 0.067 −0.004 0.020 source: author calculations. this table reports on the summary statistics of the observed returns (Mean), risk (std.Dev), and performance measure (sharpe) for the 5-day rebalancing scenario assessment on investment for single asset: LQ45, gold, crude oil, government bonds, and bitcoin as compared to paired assets: esg-based equities and LQ45, esg-based equities and gold, esg-based equities and crude oil, esg-based equities and government bonds, and esg-based equities and bitcoin. 16 k. n. asih etal. 5. Discussion this study showed how esg investing contributes to optimal portfolio construction. First, from a risk perspective, the time-varying volatilities of any individual asset are always higher than those of paired assets of esg-based equities and any other asset. this result suggests that the addition of esg-based equities can diversify and reduce investment risk for conventional and safe havens, commodities and digital assets, particularly crude oil and Bitcoin. second, from a performance perspective, the addition of esg-based equities improves portfolio performance for most asset classes, i.e. gold, government bonds, and oil, especially during the cOViD-19 pandemic. this demonstrates the diversifying role of esg-based equities and the hedging functions of esg-based equities to maintain optimal performance in turbulent times. these results were consistent with Dai (2021) and Jin (2022), which stated that esg investing can reduce risk and improve the risk-adjusted return of the overall portfolio. the result of the study also agrees with hoepner (2010) that optimal portfolio diversification, particularly through the use of esg-based assets, can be achieved across a variety of asset classes. a notable finding of this study highlights that a high weighting of esg-based equities in the crude oil and Bitcoin portfolio has the potential to reduce risk and is consistent with liu and hamori (2020). conversely, the relatively low weighting of esg-based equities with assets such as bonds and gold has the potential to improve portfolio performance, although without associated risk reduction. Overall, the empirical results of this study support the fact that, beyond diversification, the inclusion of esg-based equities also provides insurance and improves portfolio performance. During the cOViD-19 period, most asset classes used in this study performed positively and even better when each individual asset was paired with esg-based equities, particularly gold, treasuries and oil. Moreover, a portfolio of esg-based equities and Bitcoin generates positive risk-adjusted returns during the pandemic, although slightly lower than holding just Bitcoin. the portfolio combining esg-based stocks with gold not only achieved a higher mean return (+5.3 percentage points) compared to a gold-only strategy, but also showed an improvement in the associated risk (-9.0 percent). in fact, a portfolio of esg-based equities and gold achieved a 3.1 percentage point improvement in risk-adjusted returns. this is in marked contrast to the findings of akhtaruzzaman et al. (2021) and Ji et al. (2020), which highlight gold’s unwavering role as a safe haven during periods of financial instability caused by the cOViD-19 pandemic, and that of salisu et al. (2021), which underline its conditional hedging properties. Table 8. summary of the 22-day portfolio rebalancing scenario performance. Pre-CoViD-19 CoViD −19 Recovery LQ45 esg + LQ45 LQ45 esg + LQ45 LQ45 esg + LQ45 Mean 0.080 0.083 0.034 0.045 −0.003 0.035 std.Dev 4.100 4.088 7.593 7.446 3.072 3.265 sharpe 0.016 0.066 0.002 0.017 −0.006 0.005 gold esg + gold gold esg + gold gold esg + gold Mean 0.109 0.074 0.141 0.130 0.045 0.052 std.Dev 1.845 1.645 3.530 3.192 2.455 2.246 sharpe 0.051 0.031 0.035 0.049 0.012 0.012 oil esg + oil oil esg + oil oil esg + oil Mean −0.089 −0.020 −0.995 −0.267 0.241 0.114 std.Dev 5.110 3.071 13.094 7.284 6.342 3.815 sharpe −0.021 −0.015 −0.077 0.031 0.035 0.037 Bond esg + Bond Bond esg + Bond Bond esg + Bond Mean −0.001 −0.006 0.087 0.110 −0.115 −0.102 std.Dev 0.886 0.872 2.072 2.096 1.796 1.756 sharpe −0.019 −0.016 0.034 0.100 −0.073 −0.097 Bitcoin esg + Bitcoin Bitcoin esg + Bitcoin Bitcoin esg + Bitcoin Mean −0.461 −0.109 1.664 0.218 −0.040 0.074 std.Dev 15.284 3.795 15.670 6.977 13.337 4.141 sharpe −0.031 −0.031 0.105 0.076 −0.004 0.021 source: author calculations. this table reports on the summary statistics of the observed returns (Mean), risk (std.Dev), and performance measure (sharpe) for the 22-day rebalancing scenario assessment on investment for single asset: LQ45, gold, crude oil, government bonds, and bitcoin as compared to paired assets: esg-based equities and LQ45, esg-based equities and gold, esg-based equities and crude oil, esg-based equities and government bonds, and esg-based equities and bitcoin. cOgent BUsiness & ManageMent 17 a similar observation applies to government bonds, which can be classified as less risky assets. During the pandemic, a portfolio of esg-based equities and government bonds has similar characteristics to a portfolio of esg-based equities and gold. combining esg-based equities and government bonds can increase the portfolio’s risk-adjusted return by 5.7 percentage points. Despite the prevailing evidence that gold and government bonds are a relatively stable and risk-mitigating asset in turbulent times, our results show that including esg-based equities in the investment portfolio can actually mitigate risk even further. in times of market turmoil and the health crisis, the diversification and hedging function of esg-based equities takes on great importance, especially with Bitcoin and crude oil. By combining Bitcoin with esg-based equities during the cOViD-19 period, the mean return decreases by 1536 percentage points compared to what we would achieve if we invested exclusively in Bitcoin, but the associated risk is reduced by 56.4 percent. across the sample period, portfolios of esg-based stocks and Bitcoin reduced associated risk by approximately 66.9 percent. given the collapse in crude oil prices triggered by the coronavirus outbreak, a portfolio of esg-based equities and crude oil was successful in mitigating the losses before and during the cOViD-19 periods (37.8 percentage points) compared to a portfolio composed exclusively of crude oil in the same periods, and the associated risk was significantly reduced (50.1 percent). During the recovery period, combining crude oil with esg-based equities continues to reduce the associated risk (43.9 percent) compared to investing in crude oil alone. the rebalancing scenarios aimed at improving performance were supported by existing research such as horn and Oehler (2020). the central idea behind such rebalancing is the ability to recalibrate a portfolio’s asset allocation, aligning it with an investor’s goals and risk tolerance while managing the dynamic nature of market conditions (hong, 2021). this study provides a noteworthy finding on rebalancing scenarios. the performance of the risk-adjusted returns for each portfolio shows a similar result regardless of the rebalancing scenarios (daily, weekly or monthly), although a slightly better or lower value in some portfolios. Both esg-based equities with government bonds and esg-based equities with Bitcoin portfolios consistently experienced optimal risk-adjusted returns across all rebalancing scenarios. While this study contributes to the cross-asset portfolio literature in terms of the importance of a dynamic portfolio and understanding rebalancing scenarios, it has limitations that must be acknowledged. First, constructing a portfolio using asset classes from different market areas may be limited by the existence of different rules and regulations. second, differences in market areas can also impact the portfolio rebalancing process, particularly with regard to transaction costs. third, portfolio performance is measured using only the sharpe ratio. therefore, providing room for further research. 6. Conclusion and remarks the aim of this study is to examine the role of indonesia’s esg-based equities in achieving optimal portfolio diversification and contribute to the current state of knowledge by building a dynamic portfolio consisting of combining the esg-based equities with different asset classes. the study found that the addition of esg-based equities can diversify and reduce investment risk. additionally, esg-based equities have hedging features to maintain optimal performance during turbulent times. this study expands our understanding of the importance of dynamic portfolios and actively managed portfolios. During the cOViD-19 pandemic, most asset classes used in this analysis performed positively and even performed better when each individual investment was paired with esg-based equities, particularly gold, bonds and oil, with only Bitcoin underperforming. the results of this study reaffirmed the evidence that government bonds and gold are comparatively stable and risk-reducing investments in uncertain times. they also highlighted the fact that including esg-based equities in investment portfolios can help reduce risk even further. the study also highlights the importance of esg-based equities in portfolios containing riskier assets such as Bitcoin and crude oil. esg-based equities have proven to be a reliable source of diversification and hedging in both crude oil and Bitcoin portfolios. During the observation period, esg-based equities always had a dominant mean dynamic weight, accounting for more than 60 percent (crude oil) and 80 percent (bitcoin). this suggests that esg-based equities 18 k. n. asih etal. not only help reduce the risks associated with investing in Bitcoin and crude oil, but also help preserve performance. this means that indonesia’s esg-based equities, although still at an early stage of development, are playing quite an impressive role, similar to their more mature esg-based equities counterparts in global markets. the study implies that investment managers and investors can reduce investment risk and balance their portfolio by including esg-based equities in their portfolio. the study also provides a better understanding of why investors support the emergence of esg-based assets, but challenges investment managers to firstly create more diverse esg-based investment products and, secondly, to operate more actively managed portfolios by rebalancing their strategy needs and risk tolerance of the customer. the results could encourage policymakers to continue supporting sustainable financing initiatives. Finally, this study opens several avenues for further research, such as the use of various esg-based assets, adding more diverse assets, incorporating transaction costs into the rebalancing strategy, and incorporating other methods of calculating performance metrics such as treynor and kappa, or Omega. Acknowledgments We extend our appreciation to school of Business, iPB University for providing access to facilities and resources that greatly support our research effort. Authors’ contributions kiki nindya asih and noer azam achsani were involved in the conception and design. kiki nindya asih and tanti novianti were involved in the analysis and interpretation of the data, including drafting the paper. noer azam achsani and adler haymans Manurung were involved in the critical revision of the intellectual content. all authors have given final approval to the version to be published and agree to be accountable for all aspects of the work. Consent for publication not applicable. Ethics approval and consent to participate not applicable Geolocation information this study was conducted at school of Business, iPB University, Bogor, indonesia. -6.586648122212661, 106.80609973030377. Disclosure statement no potential conflict of interest was reported by the author(s). Funding We do not receive any grants or any other forms of funding for this research. About the authors Kiki Nindya Asih is a doctoral candidate at the school of Business at iPB University and a central banker. she was known for her work on monetary operations and instruments, financial markets, and the external sector. Over the past five years, she has been the lead author of numerous internal publications for the bank. cOgent BUsiness & ManageMent 19 Noer Azam Achsani is a full professor of economics in the Department of economics and Dean of the school of Business, both at iPB University. he mainly teaches economics, statistics, and econometrics as well as banking and finance. he served as an expert at the World Food Program and was a consultant on various projects at the World Bank and the asian Development Bank. he has published more than hundreds of research papers in reputed international journals. Tanti Novianti is a lecturer and vice dean of the school of Business, both at iPB University. she primarily teaches economics, international trade, agricultural trade, and international economics and finance. she has published more than 50 research papers and 7 books focusing on trade economics, labor economics, industry, and development. Adler Haymans Manurung is Professor of capital Markets and Banking at Bhayangkara University Jakarta Raya, indonesia. he teaches finance courses and is also a consultant on corporate finance activities. he has published more than 200 scopus articles. he specialized in portfolio management, iPOs, risk management and capital structure. he wrote books with 62 titles in the areas of finance, investments, risk management, quantity, and research methods. ORCID kiki nindya asih http://orcid.org/0009-0001-3898-977X noer azam achsani http://orcid.org/0000-0002-1478-8586 tanti novianti http://orcid.org/0000-0002-7917-3435 adler haymans Manurung http://orcid.org/0000-0002-6212-0214 Data availability statement the data that support the findings of this study are openly available in Bloomberg and Refinitiv eikon terminals (https://www.bloomberg.com and https://eikon.refinitiv.com). References abdullah, a. M., abdul Wahab, h., ghazali, M. F., Yaacob, M. h., & Masih, a. M. M. (2022). the dynamics of australian stock indices and commodities based on MgaRch-Dcc and wavelet techniques. Cogent Business & Management, 9(1), 2152154. https://doi.org/10.1080/23311975.2022.2152154 abhayawansa, s., & Mooneeapen, O. (2022). Directions for future research to steer environmental, social and governance (esg) investing to support sustainability: a systematic literature review. in Handbook of accounting and sustainability (pp. 318–341). edward elgar Publishing. akhtaruzzaman, M., Boubaker, s., lucey, B. M., & sensoy, a. (2021). is gold a hedge or a safe-haven asset in the cOViD–19 crisis? Economic Modelling, 102, 105588. https://doi.org/10.1016/j.econmod.2021.105588 almubarak, W. i., chebbi, k., & ammer, M. a. (2023). Unveiling the connection among esg, earnings management, and financial distress: insights from an emerging market. Sustainability, 15(16), 12348. https://doi.org/10.3390/ su151612348 arif, M., hasan, M., alawi, s. M., & naeem, M. a. (2021). cOViD-19 and time-frequency connectedness between green and conventional financial markets. Global Finance Journal, 49, 100650. https://doi.org/10.1016/j.gfj.2021.100650 Boyante, R., Muturi, W., & gekara, M. (2022). Moderating influence of portfolio rebalancing on the relationship between asset allocation and financial performance of pension funds in kenya. International Journal of Finance and Accounting, 7(3), 56–67. https://doi.org/10.47604/ijfa.1644 Breitung, J., & candelon, B. (2006). testing for shortand long-run causality: a frequency-domain approach. Journal of Econometrics, 132(2), 363–378. https://doi.org/10.1016/j.jeconom.2005.02.004 Bresnan, B., & gelb, e. (1999). Getting started in asset allocation. Wiley. Broadstock, D. c., chan, k., cheng, l. t. W., & Wang, X. (2021). the role of esg performance during times of financial crisis: evidence from cOViD-19 in china. Finance Research Letters, 38, 101716. https://doi.org/10.1016/j.frl.2020.101716 chang, k. c., tian, Z., & Yu, J. (2017 Dynamic asset allocation—Chasing a moving target [Paper presentation]. in 2017 20th international conference on information Fusion (Fusion). ieee, pp. 1–8. https://doi.org/10.23919/ iciF.2017.8009870 Dai, Y. (2021). can esg investing beat the market and improve portfolio diversification? evidence from china. The Chinese Economy, 54(4), 272–285. https://doi.org/10.1080/10971475.2020.1857063 Deloitte. (2020). ESG platforms – Stakeholder communications and the role of technology. Retrieved 2023, Mar 25, from https://www.processmap.com/ehs-software/esg-stakeholder-communications-and-the-role-of-technology/. Díaz, a., esparcia, c., & lópez, R. (2022). the diversifying role of socially responsible investments during the cOViD-19 crisis: a risk management and portfolio performance analysis. Economic Analysis and Policy, 75, 39–60. https://doi. org/10.1016/j.eap.2022.05.001 Diris, B., Palm, F., & schotman, P. (2015). long-term strategic asset allocation: an Out-of-sample evaluation. Management Science, 61(9), 2185–2202. https://doi.org/10.1287/mnsc.2014.1924 20 k. n. asih etal. Dutta, a., Bouri, e., Dutta, P., & saeed, t. (2021). commodity market risks and green investments: evidence from india. Journal of Cleaner Production, 318, 128523. https://doi.org/10.1016/j.jclepro.2021.128523 engle, R. (2002). Dynamic conditional correlation. Journal of Business & Economic Statistics, 20(3), 339–350. https://doi. org/10.1198/073500102288618487 erragraguy, e., & Revelli, c. (2015). should islamic investors consider sRi criteria in their investment strategies? Finance Research Letters, 14, 11–19. https://doi.org/10.1016/j.frl.2015.07.003 Faber, M. (2017). a quantitative approach to tactical asset allocation revisited 10 years later. The Journal of Portfolio Management, 44(2), 156–167. https://doi.org/10.3905/jpm.2018.44.2.156 Fama, e. F. (1996). Multifactor portfolio efficiency and multifactor asset pricing. The Journal of Financial and Quantitative Analysis, 31(4), 441. https://doi.org/10.2307/2331355 Ferrer, R., Benítez, R., & Bolós, V. J. (2021). interdependence between green financial instruments and major conventional assets: a wavelet-based network analysis. Mathematics, 9(8), 900. https://doi.org/10.3390/math9080900 Filis, g., Degiannakis, s., & Floros, c. (2011). Dynamic correlation between stock market and oil prices: the case of oil-importing and oil-exporting countries. International Review of Financial Analysis, 20(3), 152–164. https://doi. org/10.1016/j.irfa.2011.02.014 Friede, g., Busch, t., & Bassen, a. (2015). esg and financial performance: aggregated evidence from more than 2000 empirical studies. Journal of Sustainable Finance & Investment, 5(4), 210–233. https://doi.org/10.1080/20430795.201 5.1118917 giese, g., lee, l.-e., Melas, D., nagy, Z., & nishikawa, l. (2019). Performance and risk analysis of index-based esg portfolios. The Journal of Index Investing, 9(4), 46–57. https://doi.org/10.3905/jii.2019.9.4.046 glosten, l. R., Jagannathan, R., & Runkle, D. e. (1993). On the relation between the expected value and the volatility of the nominal excess return on stocks. The Journal of Finance, 48(5), 1779–1801. https://doi. org/10.1111/j.1540-6261.1993.tb05128.x hakansson, n. h. (1975). Optimal investment and consumption strategies under risk for a class of utility functions. in stochastic Optimization Models in Finance (2006 edition) (pp. 525–545). academic Press. https://doi.org/10.1016/ B978-0-12-780850-5.50045-9 hoepner, a. g. F. (2010). corporate social Responsibility and investment Portfolio Diversification. SSRN Electronic Journal, 10, 1–16. https://doi.org/10.2139/ssrn.1599334 hong, X. (2021). Portfolio rebalancing: tradeoffs and decisions. SSRN Electronic Journal, 10, 1–21. https://doi. org/10.2139/ssrn.3858951 horn, M., & Oehler, a. (2020). automated portfolio rebalancing: automatic erosion of investment performance? Journal of Asset Management, 21(6), 489–505. https://doi.org/10.1057/s41260-020-00183-0 humphrey, J. e., & lee, D. D. (2011). australian socially responsible funds: Performance, risk and screening intensity. Journal of Business Ethics, 102(4), 519–535. https://doi.org/10.1007/s10551-011-0836-7 indriastuti, M., & chariri, a. (2021). the role of green investment and corporate social responsibility investment on sustainable performance. Cogent Business & Management, 8(1), 1960120. https://doi.org/10.1080/23311975.2021.19 60120 Ji, Q., Zhang, D., & Zhao, Y. (2020). searching for safe-haven assets during the cOViD-19 pandemic. International Review of Financial Analysis, 71, 101526. https://doi.org/10.1016/j.irfa.2020.101526 Jin, i. (2022). systematic esg Risk and decision criteria for optimal portfolio selection. The Journal of Portfolio Management, 48(10), 206–225. https://doi.org/10.3905/jpm.2022.1.418 kehati. (2017). laporan tahunan 2017. killick, R., Fearnhead, P., & eckley, i. a. (2012). Optimal detection of changepoints with a linear computational cost. Journal of the American Statistical Association, 107(500), 1590–1598. https://doi.org/10.1080/01621459.2012.737745 kurniatama, g. a., Robiyanto, R., sasongko, g., & huruta, a. D. (2021). Determinants of corporate social responsibility: empirical evidence from sustainable and responsible investment index. Calitatea, 22(183), 55–61. leković, M. (2021). historical development of portfolio theory. Tehnika, 76(2), 220–227. https://doi.org/10.5937/tehnika2102220l liu, g., & hamori, s. (2020). can one reinforce investments in renewable energy stock indices with the esg index? Energies, 13(5), 1179. https://doi.org/10.3390/en13051179 Manurung, a. h. (2016). Konstruksi portofolio efek di Indonesia. adler Manurung Press. Manurung, a. h., kartono, R., tjahjana, D. D., tjiptadi, D., & saputra, n. (2021). Manajemen: Teori dan Perkembangannya. Pt adler Manurung Press. Morelli, a., & nesta, a. (2021). Reader survey, 18+ U.S. Investors. ngunyi, a., Mundia, s., & Omari, c. (2019). Modelling volatility dynamics of cryptocurrencies using gaRch models. Journal of Mathematical Finance, 09(04), 591–615. https://doi.org/10.4236/jmf.2019.94030 Ogata, h. (2012). Optimal portfolio estimation for dependent financial returns with generalized empirical likelihood. Advances in Decision Sciences, 2012, 1–8. https://doi.org/10.1155/2012/973173 Orskaug, e. (2009). Multivariate Dcc-gaRch model: -With various error distributions. Panagopoulos, a. (2023). the use of sustainable financial instruments in relation to the social impact investment. esg policies, capital markets’ approach and investors’ protection. an innovative perspective. SSRN Electronic Journal, 10, 1–66. https://doi.org/10.2139/ssrn.4131735 cOgent BUsiness & ManageMent 21 Pedersen, l. h., Fitzgibbons, s., & Pomorski, l. (2021). Responsible investing: the esg-efficient frontier. Journal of Financial Economics, 142(2), 572–597. https://doi.org/10.1016/j.jfineco.2020.11.001 Pourbabaee, F., kwak, M., & Pirvu, t. a. (2016). Risk minimization and portfolio diversification. Quantitative Finance, 16(9), 1325–1332. https://doi.org/10.1080/14697688.2015.1115891 Robiyanto, R., huruta, a. D., Frensidy, B., & Yuliana, a. F. (2023). sustainable and responsible investment dynamic cross-asset portfolio. Cogent Business & Management, 10(1), 2174478. https://doi.org/10.1080/23311975.2023.2174478 Robiyanto, R., nugroho, B. a., huruta, a. D., Frensidy, B., & suyanto, s. (2021). identifying the role of gold on sustainable investment in indonesia: the Dcc-gaRch approach. Economies, 9(3), 119. https://doi.org/10.3390/economies9030119 Rumyantseva, a., & tarutko, O. (2022). (2022). impact of the esg principles on the corporate financial strategy. in Challenges and Solutions in the Digital Economy and Finance: Proceedings of the 5th International Scientific Conference on Digital Economy and Finances (DEFIN 2022) (pp. 309–318). springer. sahamkhadam, M., stephan, a., & Östermark, R. (2018). Portfolio optimization based on gaRch-eVt-copula forecasting models. International Journal of Forecasting, 34(3), 497–506. https://doi.org/10.1016/j.ijforecast.2018.02.004 salisu, a. a., Raheem, i. D., & Vo, X. V. (2021). assessing the safe haven property of the gold market during cOViD-19 pandemic. International Review of Financial Analysis, 74, 101666. https://doi.org/10.1016/j.irfa.2021.101666 sharma, g. D., sarker, t., Rao, a., talan, g., & Jain, M. (2022). Revisiting conventional and green finance spillover in post-cOViD world: evidence from robust econometric models. Global Finance Journal, 51, 100691. https://doi. org/10.1016/j.gfj.2021.100691 sharpe, W. F. (1994). the sharpe ratio. The Journal of Portfolio Management, 21(1), 49–58. https://doi.org/10.3905/ jpm.1994.409501 shefrin, h., & statman, M. (2000). Behavioral portfolio theory. The Journal of Financial and Quantitative Analysis, 35(2), 127. https://doi.org/10.2307/2676187 sinchukov, a. (2022). general issues of designing and managing a portfolio of financial instruments. Scientific Research and Development. Economics, 10(1), 36–43. https://doi.org/10.12737/2587-9111-2022-10-1-36-43 sparkes, R., & cowton, c. J. (2004). the maturing of socially responsible investment: a review of the developing link with corporate social responsibility. Journal of Business Ethics, 52(1), 45–57. https://doi.org/10.1023/B:BUsi.0000033106.43260.99 Yousaf, i., suleman, M. t., & Demirer, R. (2022). green investments: a luxury good or a financial necessity? Energy Economics, 105, 105745. https://doi.org/10.1016/j.eneco.2021.105745 Zinecker, M., Balcerzak, a. P., Faldzinski, M., Pietrzak, M. B., & Meluzin, t. (2016). Application of DCC-GARCH model for analysis of interrelations among capital markets of Poland, Czech Republic and Germany. institute of economic Research Working Papers. Appendix A. Pruned exact linear time algorithm (PELT) for time series change point detection Pelt algorithm developed by killick etal. (2012), a method devised for efficiently detecting change points within a time series dataset. such change points may indicate significant shifts in means or volatility, making this analysis vital for econometric studies and time series forecasting. time series data often contain points where the underlying statistical properties alter, termed as ‘change points’. Recognizing these points aids in understanding structural breaks, regime changes, or shifts in volatility. the Pelt algorithm offers a precise and computationally efficient approach to this detection problem. Methodology Dynamic programming Pelt employs a dynamic programming approach to minimize a designated cost function. this function measures how well a given model, inclusive of certain change points, explains the observed data. Cost function the cost function evaluates the divergence between the actual data and the data predicted by the model that incorporates the change points. this cost can be predicated upon shifts in means or variances (volatility). Pruning mechanism in determining the optimal model, rather than evaluating a vast range of potential models (each denoting different change points), Pelt uses a pruning mechanism. By disregarding evidently suboptimal models early on, this mechanism accelerates the overall computation process. 22 k. n. asih etal. Penalty component to circumvent overfitting—a scenario where the model aligns exceptionally well with the sample data but performs inadequately on new data—a penalty is imposed for each added change point. Steps of the PELT Algorithm: 1. initialization: Begin with the assumption that there’s only a single change point at the start of the data. 2. evaluation loop: For each data point, t: a. compute the cost of potential change points up to t. b. sum up this cost with the penalty for adding an extra change point. 3. Pruning strategy: at each step, discard change point candidates that have a higher cost than other candidates, ensuring the algorithm remains efficient by focusing only on probable change points. 4. Optimal change Points identification: after evaluating all data points, the change points corresponding to the minimal cost function are deemed the optimal change points. 5. Result consolidation: collate and present the identified change points in an ordered manner, denoting where significant shifts in the statistical properties of the time series occur.