We construct banking networks using bank-level balance sheet data from 2005 to 2010 from five emerging countries: Argentina, Brazil, Mexico, South Africa, and Taiwan. The network interaction is based on the leverage ratio dependence between each pair of banks within a same country. Despite leverage and accounting rules heterogeneity, the results are robust across countries. The leverage diversity produces financial networks with a modular structure characterized by one large bank community, some small ones, and isolated banks. However, these groups of banks merge together creating a financial network topology that converges to a unique large cluster at a relatively low leverage dependence level. Finally, we simulate the banking system through a model of corporate and interbank loans with credit shocks, where links between banks arise due to insufficient liquidity. The model yields leverage-based networks that are similar to the empirical ones. A model prediction for banks' growth is presented and tested in the data.