Connectedness between Currency Risk Hedging and Firm Value: A Deep Neural Network-based Evaluation

Despite a massive amount of research on the association between currency risk hedging and firms value, the previous literature provides no clear-cut findings on whether currency risk hedging adds value or not. However, the present study investigates the interaction between currency risk hedging and automobile firms' value of an emerging economy by deploying the most powerful an artificial neural network (ANN) based technique using a quarterly dataset over the period 2007–2021. The magnitude of an empirical data sample demonstrates that 71.4% automobile firms of Pakistan are currently using foreign currency derivatives to hedge their currency risk. We propose a deep neural network-based multivariate regression model (DNN-MRM) to examine the relationship between endogenous, exogenous and control variables of the study. For robustness estimation, the present study includes the structural-based Bayesian network with PC algorithm and generalized least square-random effect (GLS-RE) regression model to explore the join probability distribution of random variables, the individual and combined value-enhancing effect of currency risk hedging respectively. The output shows the automobile firms value are dominantly affected by currency risk hedging and firm characteristics. Moreover, the individual effect estimation depicts the currency risk hedging positively influences 9.27% Tobin’s Q (TQ) and 13.60% Adjusted Tobin’s Q (ATQ). The findings concluded that the value of automobile firms is significantly enhanced by hedging their currency risk. The results are shown to be consistent through a series of robustness tests.