Not just an idea, it is under active research. There’s multiple ways to directionally steer this effort, with various degrees of (nega/posi)-tive consequences. This type of work needs very extensive research, and we should use this channel to discuss.
The most promising method is to augment the current staking functionality to have a “dynamic destination” field. Currently, when you do staking, you do three things: first you establish your NodeID, then you point your AVAX to that NodeID to stake for some specified period of time, and finally you specify a recipient address for the rewards. However, the stake is always returned back to the original owner of the stake.
You can augment this process by specifying a “destination” for the returned stake as well. Effectively, this means that at the end of the operation, the stake isn’t returned to some third-party destination address.
This enables us to create on-chain contracts that are able to promise delivery of coins upon completion of the stake. Since you’re now able to programmatically specify where the stake ends up, you can create a derivate product that can prove ownership of some UTXOs tied to the destination of the stake. You can freely trade this product back and forth.
Beyond whether this is a good idea we also have to explore how to precisely construct this product. There’s some weirdness with it, in that it would be really difficult to keep track of if it is fungible. This means that you’d basically be trading non-fungible IOUs (worth a particular block of stake with some expiry time into the future).