Smart contracts cleave to Lawrence Lessig’s coinage Code is Law — you code legal rights into the electronic operating parameters between the parties: they physically constrain what you do to each other, rather than being abstract metaphysical considerations that float above your messy interactions which may or may not reflect those abstract terms.

Notwithstanding breathless distributed ledger technology chat to the contrary, we’ve had smart contracts for a while: wherever counterparties interact electronically (as in collateral posting arrangements under a CSA for example: it is the algorithms and thresholds set in the technology, rather than the abstract ones set out in paper, which govern what, when and how much collateral the parties exchange: it is hardly a big leap to ditch the need for abstract textual reflections of operatring parameters in separately executed "legal terms".

Where the contract is of a highly operational nature — and much of the machinery of finance is, in fairness — smart contracts aren’t so much a revolutionary idea as an utterly necessary one. But the problem is this: while 80% of your contract delivered by electronic interaction and is suitable for digital governance, the balance typically isn't. These are the close out rights, termination triggers and events of default. They concern matters not ascertainable by electronic transmission between the parties, and they require judgment and evaluation, both of their existence (is that adverse change truly material?) and in terms of one’s response: Ok, the NAV trigger has been hit. But do we want to close out? How is our overall position? What do we think of the counterparty's forward prospects? What are the pros and cons?

No smart contract will be able to measure these things, much less make an executory decision about them.