Tinn, KKTinn2018-05-292018-06-17http://hdl.handle.net/10044/1/59721Blockchain technology makes it possible to create immutable smart contracts that are based on reliable and timestamped records of transactions. For this reason, it can eliminate standard frictions such as the need for costly verification, and make raising external financing more accessible. At the same time, it does not eliminate moral hazard, which can even become more severe with the trend towards more faster data analysis and more frequent decision making. I analyse optimal financial contracting in this new environment. I show that whenever there is learning, borrowers and lenders generally benefit from contracts that depend on not just whether, but also on when, cash flows occur -- a feature that timestamped records and blockchain can easily enable. In contrast to some earlier theoretical results which argue that contracting on aggregates is sufficient, I identify conditions when contracting on the cash flow sequence matters. I show that the optimal contract can be expressed as a dynamically adjusting cash flow splitting rule between the lender and the borrowers. Under the optimal contract, external capital is as cheap as internal capital, and more accessible to agents with no internal funds. Equity is optimal only when the potential cash flows are independently distributed over time, and becomes relatively more expensive when there is more learning. Debt is always a sub-optimal contract in this environment and becomes increasingly expensive with more frequent decision making.© 2018 The Author.Conference Paper10.2139/ssrn.3072854