This dissertation explores how and why four Latin American democracies -- Argentina, Brazil, Chile, and Mexico -- have regulated commercial banks since the late 1970s. The objective is to explain variation across countries and time in the regulatory regimes of the commercial bank sector, meaning governments' decisions about the levels of restrictions imposed on both the structure and the risk-management behavior of banks. I find that traditional approaches to regulation -- based on interests, ideology, and institutions -- are insufficient to elucidate the political economy of banking regulation in Latin America. In offering a new typological -- the embedded-agency -- theory of regulation, I argue that variation in bank regulatory regimes depends on the nature of the principal-agent (P-A) relationship among regulation-makers and its interaction with what I call 'conjunctural' determinants (i.e., systemic banking crises, international pressures, and technological advances). While the nature of the P-A relationship among regulators offers insights into who the relevant actors in bank regulation are and how a consensus among them can be achieved, the conjunctural determinants allow us to identify which regulatory objective is assigned priority. The interaction between these two sets of variables indicate the circumstances under which each regulatory scheme -- the cost-padding, the laissez-faire, the prudential, or the over-protective -- is more likely to occur. Given that the regulation of commercial banks affects the level of performance of countries' financial systems, and the performance of financial systems in turn affects economic development, the consequences of deficient bank regulation and supervision are too grave to be disregarded.