In a series of precedent papers, we have presented a comprehensive methodology, termed Field Economics, for translating a standard economic model into a statistical field-formalism framework. This formalism requires a large number of heterogeneous agents, possibly of different types. It reveals the emergence of collective states among these agents or type of agents while preserving the interactions and microeconomic features of the system at the individual level. In two prior papers, we applied this formalism to analyze the dynamics of capital allocation and accumulation in a simple microeconomic framework of investors and firms. Building upon our prior work, the present paper refines the initial model by expanding its scope. Instead of considering financial firms investing solely in real sectors, we now suppose that financial agents may also invest in other financial firms. We also introduce banks in the system that act as investors with a credit multiplier. Two types of interaction are now considered within the financial sector: financial agents can lend capital to, or choose to buy shares of, other financial firms. Capital now flows between financial agents and is only partly invested in real sectors, depending on their relative returns. We translate this framework into our formalism and study the di¤usion of capital and possible defaults in the system, both at the macro and micro level. At the macro level, we find that several collective states may emerge, each characterized by a distinct level of average capital and investors per sector. These collective states depend on external parameters such as level of connections between investors or firms' productivity. The multiplicity of possible collective states is the consequence of the nature of the system composed of interconnected heterogeneous agents. Several equivalent patterns of returns and portfolio allocation may emerge. The multiple collective states induce the unstable nature of …nancial markets, and some of them include defaults may emerge. At the micro level, we study the propagation of returns and defaults within a given collective state. Our findings highlight the significant role of banks, which can either stabilize the system through lending activities or propagate instability through loans to investors.