We develop a new theory of banking that provides a view of liquidity creation that differs significantly from the existing paradigm. The theory links modern-day banks to their historical origin as warehouses. When the warehouse accepts deposits, it issues “warehouse receipts,” claims on deposits it safeguards, as “proof” of deposits, but this suffices neither to create liquidity nor to make the warehouse a bank.

Our second main result is that interbank markets, i.e. inter-warehouse markets, for the entrepreneur’s debt are sufficient to support this enforcement mechanism even if the entrepreneur can borrow from one warehouse and store his output in another. This is because this other warehouse can buy the entrepreneur’s debt in the interbank market, thereby obtaining the right to seize the entrepreneur’s deposits. As a result, the entrepreneur ends up repaying in full no matter which warehouse he deposits in. This finding can shed light on why successful banking systems throughout history, such as those operated by Egyptian granaries and London goldsmiths, as well as those in existence today, have indeed developed interbank networks. Our third main result is that warehouse banks make loans even if they have no initial deposits to lend out. In fact, they lend the constrained-efficient amount by making loans in new fake warehouse receipts—i.e. receipts to redeem deposits that are not backed by current deposits.8 5 Empirical evidence, that seems to support this result appears