Hash-linked timestamping is the key feature behind blockchain. It enhances trust as it enables contracting parties to have common and reliable records of transactions and of their timing. I develop a model of raising external financing where, due to this new technology, some traditional contracting frictions are not present. However, there are informational frictions whereby borrowers learn from data frequently, which affects their effort incentives. I identify conditions under which contracts benefit from being "smart", i.e., self-adjusting based on timestamps. I further show that increasing the frequency of learning makes traditional assets, e.g., debt and equity, costlier and more restrictive.