Intraday Liquidity and Return Dependency Structure Modeling of a Portfolio in Tehran Stock Exchange with ACP-GARCH-High Dimension Vine Copula

Modeling the joint distribution of liquidity and return to determine the dependency structure of a 15-stock portfolio using intraday data in 2019 provides a suitable model for the commonalities. Based on the complexities of higher dimensions multivariate modeling (combined distribution of liquidity and return), after univariate modeling of the stocks’ liquidity based on Autoregressive Conditional Poisson model (ACP) and the returns with the Generalized Autoregressive Conditional Heteroscedastic (GARCH), the marginal distributions were incorporated into a Copula Vine to model the dependency. The findings of this study, based on high-dimensional high-frequency data, indicate that there is an extreme nonlinear correlation between the liquidity of stocks and also between liquidity and returns across the portfolio, which is necessary to take into account in risk assessments to prevent inaccurate assessment of risk indicators such as value at risk. In addition, the results have shown that modeling the joint distribution of liquidity and stock returns in high dimensions relying on the Copula Vine model due to the flexibility performs well.