Today Coinbase’s Brian Armstrong shared out that there were rumors that exchanges offering staking services to retail users may be made illegal in the US.

https://twitter.com/brian_armstrong/status/1623459203150131201

I agree with Brian’s thread, but also have the opinion that if exchanges are marketing staking products purely as yield generating financial machines, they’re not doing their retail customers a valuable service in the grand scheme of things. Let’s unpack this a bit.

First of all, it’s worth noting that every network is different, and there’s no such thing as one-size-fits-all staking models. In the case of networks like Ethereum, where staking needs to be accompanied by node operation, and only a limited amount of stake can sit on a validator, there’s a better argument that more stake actually provides more security and resilience, and providing this sort of exchange based staking service can actually help the network. In this case the biggest risk seems to be centralization risk, but a competitive exchange environment can help balance this out. Users, at the end of the day, are still giving up a portion of their returns to a centralized middle man toll collector, but that can be justified in terms of the value they’re providing by actually operating the node and securing the network.

In the case of delegated staking on work networks like Livepeer however, the picture gets murkier. In this case, delegators (token holding user who delegate stake towards another node) aren’t required to run a node, but they still provide value through two methods:

1. Increasing the security of the network. More stake on a given node means more at risk, and the more trust a user of the network can have in contracting with that node.

2. Quality assurance. They route more work by default towards the nodes with more stake, essentially providing a higher quality of service to the end users.

A delegator who is optimizing for their own financial growth in the network, is an active participant who is moving stake around according to where they see the best, honest work being done, the most fees being generated, and the largest available returns for moving stake in the short term. If the delegator behaves optimally, then they get their best personal outcome, but the network itself is also optimized in the process, settling on an equilibrium work distribution where nodes are maximizing fees based on capacity, location, performance, and consistency.

An exchange running one node in order to offer a delegated staking landing spot for the users of its staking product, typically has not yet participated in this way, nor passed through the opportunity to their users. Instead, they typically run a lightly (or non) contributing node purely for the purposes of passive yield generation.

This in-and-of-itself is not necessarily as bad as it sounds. It’s more like “unoptimized”, or an inevitable by-product of delegated staking incentives. One could argue the nodes themselves are not helping the network, but there is nothing to stop them from not even bothering with the node, and instead just delegating all stake to another operator. In fact they are incurring more cost to operate the node itself. Their penalty, again, is unoptimized returns by not actively participating to maximize fees, nor increase the network’s long term utility and value. But they’re not causing any harm. On the positive side, it could be argued that they’re creating awareness, and onboarding thousands of users to the project at the very top of the funnel - some of whom will do their research, join the community, participate more directly, become evangelists, join the team, or even build on the network itself. All of which are positive contributions that further the success of the project, even if the actual node operation and delegation isn’t helping very much directly.

What might better look like?

Rather than marketing staking products to retail as passive yield generation, they could go one step further and both communicate the active participation possibilities, and instead proxy the staking experience in a way that is helpful to the network. This would both generate better outcomes for the exchange user and for the network itself. If the exchanges were thinking long term, those two outcomes would be better for the exchange as well.

Today Coinbase’s Brian Armstrong shared out that there were rumors that exchanges offering staking services to retail users may be made illegal in the US.

https://twitter.com/brian_armstrong/status/1623459203150131201

I agree with Brian’s thread, but also have the opinion that if exchanges are marketing staking products purely as yield generating financial machines, they’re not doing their retail customers a valuable service in the grand scheme of things. Let’s unpack this a bit.

First of all, it’s worth noting that every network is different, and there’s no such thing as one-size-fits-all staking models. In the case of networks like Ethereum, where staking needs to be accompanied by node operation, and only a limited amount of stake can sit on a validator, there’s a better argument that more stake actually provides more security and resilience, and providing this sort of exchange based staking service can actually help the network. In this case the biggest risk seems to be centralization risk, but a competitive exchange environment can help balance this out. Users, at the end of the day, are still giving up a portion of their returns to a centralized middle man toll collector, but that can be justified in terms of the value they’re providing by actually operating the node and securing the network.

In the case of delegated staking on work networks like Livepeer however, the picture gets murkier. In this case, delegators (token holding user who delegate stake towards another node) aren’t required to run a node, but they still provide value through two methods:

1. Increasing the security of the network. More stake on a given node means more at risk, and the more trust a user of the network can have in contracting with that node.

2. Quality assurance. They route more work by default towards the nodes with more stake, essentially providing a higher quality of service to the end users.

A delegator who is optimizing for their own financial growth in the network, is an active participant who is moving stake around according to where they see the best, honest work being done, the most fees being generated, and the largest available returns for moving stake in the short term. If the delegator behaves optimally, then they get their best personal outcome, but the network itself is also optimized in the process, settling on an equilibrium work distribution where nodes are maximizing fees based on capacity, location, performance, and consistency.

An exchange running one node in order to offer a delegated staking landing spot for the users of its staking product, typically has not yet participated in this way, nor passed through the opportunity to their users. Instead, they typically run a lightly (or non) contributing node purely for the purposes of passive yield generation.

This in-and-of-itself is not necessarily as bad as it sounds. It’s more like “unoptimized”, or an inevitable by-product of delegated staking incentives. One could argue the nodes themselves are not helping the network, but there is nothing to stop them from not even bothering with the node, and instead just delegating all stake to another operator. In fact they are incurring more cost to operate the node itself. Their penalty, again, is unoptimized returns by not actively participating to maximize fees, nor increase the network’s long term utility and value. But they’re not causing any harm. On the positive side, it could be argued that they’re creating awareness, and onboarding thousands of users to the project at the very top of the funnel - some of whom will do their research, join the community, participate more directly, become evangelists, join the team, or even build on the network itself. All of which are positive contributions that further the success of the project, even if the actual node operation and delegation isn’t helping very much directly.

What might better look like?

Rather than marketing staking products to retail as passive yield generation, they could go one step further and both communicate the active participation possibilities, and instead proxy the staking experience in a way that is helpful to the network. This would both generate better outcomes for the exchange user and for the network itself. If the exchanges were thinking long term, those two outcomes would be better for the exchange as well.