بررسی پویایی رابطه بین بازده رمزارزهای منتخب و بازار سهام ایران در چارچوب رویکرد کاپولای متغیر طی زمان

Purpose: The parallel expansion of the global cryptocurrency market and the Tehran Stock Exchange (TSE) presents critical considerations for Iranian investors and policymakers. With an estimated 12 million Iranians participating in the cryptocurrency market by 2023, it is crucial to understand the interplay between this globally-traded decentralized asset class and the traditional Iranian equity market. The present research addresses two key concerns including a) the potential impact of decentralized digital assets, operating outside the direct oversight of the central bank, on the efficacy of domestic monetary policies as investment flows shift, and b) the need for clear insights for investors and portfolio managers regarding the role of cryptocurrencies in diversification, risk management, and potential hedging strategies within portfolios dominated by Iranian assets. Therefore, the main objective of this study is to empirically investigate the dynamic dependence structure between the returns of Bitcoin (BTC), Ethereum (ETH), Litecoin (LTC), Dogecoin (DOGE), and Cardano (ADA), as five major cryptocurrencies, and the overall Iranian stock market performance, represented by the TSE All-Share Index. We aim to characterize the nature, strength, and evolution of this relationship over a period from January 2018 to December 2023, a timeframe marked by significant volatility and growth in both markets. Utilizing advanced econometric models capable of capturing complex, non-linear, and time-varying dependencies, particularly in the tails of return distributions, the study seeks to provide robust evidence on the potential benefits or risks of integrating cryptocurrencies into Iranian investment portfolios. This offers insights for financial regulation and policymaking concerning digital asset integration.
Methodology: To capture the potentially complex and evolving dependency between cryptocurrency and Iranian stock market returns, this research employs a sophisticated two-stage econometric approach centered on time-varying copulas. This framework is advantageous as it allows for separate modeling of marginal return distributions and their joint dependence structure, while it avoids restrictive assumptions like multivariate normality and accommodates the stylized facts of financial data.
In the first stage, we model the marginal distributions of the weekly returns for the TSE Index and each of the five selected cryptocurrencies. Recognizing characteristics like volatility clustering, leverage effects, and potential long memory, we utilize the Autoregressive Fractionally Integrated Moving Average (ARFIMA) - Fractionally Integrated Asymmetric Power Autoregressive Conditional Heteroskedasticity (FIAPARCH) model. The ARFIMA component addresses long-range dependency in the conditional mean, while the FIAPARCH component models conditional variance, capturing long memory in volatility and asymmetric responses to shocks. Preliminary diagnostic tests confirmed the non-normality, conditional heteroskedasticity, and potential long memory in the return series, justifying the ARFIMA-FIAPARCH choice. In the second stage, the standardized residuals obtained from the ARFIMA-FIAPARCH models are used to model the dependence structure via a time-varying t-student’s copula. The t-student’s copula is selected for its ability to capture tail dependency, a crucial feature often missed by simpler correlation measures. Critically, we adopt the dynamic conditional copula framework (Patton, 2006), allowing the copula’s dependence parameter and degrees of freedom to evolve over time. The data consist of 205 weekly observations from January 2018 to December 2023 for the TSE share index and the five cryptocurrencies, converted to logarithmic returns.
Findings and Discussion: An empirical analysis yielded significant insights. First, the ARFIMA-FIAPARCH modeling confirmed the presence of stylized facts in both the TSE and cryptocurrency returns, including evidence of long memory and asymmetric volatility clustering (leverage effects). This validated the chosen marginal modeling approach.
Second, the core finding from the time-varying Student’s t-copula analysis is that the dynamic dependence between the Iranian stock market (TSE) returns and the returns of the selected major cryptocurrencies was generally weak throughout the 2018-2023 sample period. The dependence parameter (ρt) did exhibit statistically significant time variation, confirming that the relationship is not static; however, the overall level of correlation remained low across all the five TSE-cryptocurrency pairs studied.
The dynamic correlation plots revealed fluctuations, with temporary increases in dependency often observed during the periods of heightened global market uncertainty or specific economic events. This indicates some sensitivity to common external shocks. However, even during these episodes, the peak correlation levels rarely exceeded thresholds typically considered low-to-moderate for financial assets, particularly when compared to the correlations often seen between traditional asset classes within more integrated markets. The estimated degrees of the freedom parameter of the t-copula also suggested the existence of statistically significant tail dependence, implying a slightly elevated probability of joint extreme movements compared to what a normal distribution assumption would suggest. However, this effect of the tail dependence appears secondary to the dominant characteristic of the overall weak dependence in the central part of the distribution.
Interpreting these findings through the lens of the Modern Portfolio Theory (MPT), the persistent weak dependence structure has important implications for Iranian investors. The low correlation strongly suggests that these major cryptocurrencies offer substantial diversification benefits when added to portfolios heavily weighted towards Iranian equities. Because cryptocurrency returns largely move independently of the TSE, their inclusion can potentially reduce the overall portfolio risk (volatility) for a given level of expected return, or enhance the portfolio’s risk-adjusted performance profile. This is particularly relevant given the relative isolation of the Iranian market and its distinct drivers compared to global cryptocurrency markets. The results position these cryptocurrencies more as diversifiers rather than consistent hedges or safe havens against TSE downturns. This is because the weak correlation, while generally low, is not consistently negative and can fluctuate.
Conclusions and Policy Implications: This study concludes that, from January 2018 to December 2023, the relationship between the Iranian stock market and the key cryptocurrencies was characterized by a weak, though dynamic and time-varying, dependence structure. Limited evidence was found for strong, systematic co-movement, suggesting minimal direct transmission of return dynamics between these markets under typical conditions.
The primary practical implication for Iranian investors and portfolio managers is the significant potential of portfolio diversification through these cryptocurrencies. The demonstrated low correlation with the domestic stock market makes them valuable tools for risk management and potentially enhances risk-adjusted returns within the specific context of the Iranian financial landscape.
From a policy standpoint, these findings suggest a need for a nuanced approach by Iranian regulatory authorities. A combination of a large existing domestic user base, the empirical evidence of diversification benefits, and the relative weakness of the dependence linkage implies that developing a clear and enabling regulatory framework for cryptocurrency investment could be more beneficial than prohibitive measures. Such a framework could provide Iranian investors with regulated access to potentially valuable diversification tools. It can channel investment interest productively while allowing participation in a global asset class. This approach must inherently address the associated risks (volatility, investor protection, illicit finance) through robust oversight, licensing, and clear operational guidelines. Furthermore, as noted in the original paper, exploring regulated channels for cryptocurrency use in international transactions might offer avenues to mitigate the impacts of economic sanctions. Creating a structured environment appears preferable to leaving this significant market segment unregulated. Future research could expand upon this issue by including a wider range of digital assets, examining specific event impacts, exploring volatility spillovers, and comparing cryptocurrencies with other relevant Iranian assets like gold or foreign exchange.