Financial instability and its destructive effects on the economy can lead to financial crises due to its contagion or spillover effects to other parts of the economy. Having an accurate measure of systemic risk gives central banks and policymakers the ability to take proper policies in order to stabilize financial markets. Much work is currently being undertaken on the feasibility of identifying and measuring systemic risk. In principle, there are two main schemes to measure interlinkages between financial institutions. One might wish to construct a mathematical model of financial market participant relations as a network/graph by using a combination of information extracted from financial statements like the market value of liabilities of counterparties, or an econometric model to estimate those relations based on financial series. In this paper, we develop a data-driven econometric framework that promotes an understanding of the relationship between financial institutions using a nonlinearly modified Granger-causality network. Unlike existing literature, it is not focused on a linear pairwise estimation. The method allows for nonlinearity and has predictive power over future economic activity through a time-varying network of relationships. Moreover, it can quantify the interlinkages between financial institutions. We also show how the model improves the measurement of systemic risk and explain the link between Granger-causality network and generalized variance decompositions network. We apply the method to the monthly returns of U.S. financial Institutions including banks, broker and insurance companies to identify the level of systemic risk in the financial sector and the contribution of each financial institution.