Financial markets constitute the backbone of modern economies, intermediating resources from those who have them (i.e., lenders) to those who can put them to productive use (i.e., borrowers). The defining feature of these markets is that they entail the exchange of goods today for a borrower’s promise to deliver goods in the future. These promises are sustained by guarantees, which are akin to the amount of future income that borrowers can credibly pledge to lenders. I refer to this pledgeable income as an economy’s stock of collateral. This stock determines the amount and type of promises that can be traded in an economy and, in turn, this set of promises determines the transactions that can be carried out. Intuitively, when this set of promises is large, resources find their most productive uses and efficiency is high.This raises a general question: what are the key determinants of the set of promises that an economy is able to sustain, and why does it vary?In macroeconomic models, it is commonly assumed that all promises are backed by only one kind of collateral, i.e. usually that of private borrowers, and that this collateral is ‘fundamental’, i.e. it consists of output. Real-world financial markets, however, rely on many types of collateral to guarantee promises. In this proposal, I focus on three such types. First, collateral may be ‘bubbly’, i.e. promises can be backed by nothing else but the income that the sale of new promises is expected to bring in the future. Second, collateral need not be private, as government promises are sustained by pledging public income. Third, collateral need not be homogenous, as it may differ in quality or type across entrepreneurs, and this quality may not be perfectly observed by all. I address the following broad questions. How do economies produce these different types of collateral? How do they interact with one another? Is there a role for policy in maintaining the efficient level and composition of collateral?