Last year a new resource model was proposed by Block.one in order to address issues of network utilization efficiency on EOSIO networks. The issue boils down to this: the existing model reflects a right to network resources (CPU and NET) proportional to the amount of staked core tokens one has, leading to network inefficiencies in the allocation of resources in situations where a token holder is staking and not leveraging their granted resources.
This is problematic as it means that token holders that aren’t leveraging their tokens as a utility continuously are reducing the efficiency of the network for those that are using it, because their “claim” on resources is simply idle and nobody else can claim those resources instead. It also incentivizes “maxing” out those resources in various schemes, such as “CPU mining” in order to yield some return larger than the current REX incentive. Thanks to Dexaran and EOS Asia for breaking this problem down so elegantly in their recent post.
I believe that the new resource model is a step in the right direction, as it attempts to bring the pricing of network resources closer to a market equilibrium. It also enhances the EOS core token with a new “DeFi” element that can bring residual value to token holders who would otherwise be contributing to the idle resource problem in the existing paradigm, by boosting the REX incentive via new fees collected by renting. If the yield is high enough, this should have the effect of providing passive income to token holders, while decoupling that mechanism from network resource allocation as is in the current REX model. Our hope is that this will lead to value accrual for token holders in a paradigm that also maximizes the efficiency of resource allocation to useful decentralized apps, so that token holders don’t seek to leverage these resources for yield instead.
There is a philosophical element at play regarding the changing of what an EOS token represents to current token holders, who may have purchased their tokens under the assumption that they will always provide a grant on network resources. Those token holders may not necessarily want these assumptions to change in a manner that costs them more money to submit their transactions to the network. At the crux of the issue on EOS is the idea that schemes such as CPU mining, which creates a large burden on network operators who must store a large volume of redundant or similar state, should be curbed by changing the way network resources are priced and allocated. Although it may be painful for the current token holders who purchased their EOS under specific assumptions and are using them in these sorts of ways, the overall longevity of the network as well as the cost of operating it over time can be made more efficient by putting in place measures that reduce this sort of CPU mining behavior. After analyzing Dexaran’s feedback, I tend to agree that the exponent on the pricing function in the new model should be formula driven rather than fixed, in order to better achieve an equilibrium and reduce the nothing at stake problem that can exacerbate submitting a high volume of “low value” (quoted as this is subjective) transactions.
At the end of the day EOS is a DAC/DAO; the EOS token holders are aligned in that their interests regarding the value of the EOS network and it’s corresponding token are intertwined. As a DAO, it is expected for the protocol and network to evolve over time, and in fact beneficial for it to do so when confronted with challenges that inhibit the value of the network. The fact of the matter is that CPU mining and incentivizing value extraction through actions that bloat the chain state has created a situation where potential adopters have had to look elsewhere due to cost constraints of building their business on EOS, while also increasing the cost to operate the network. This has created both a PR challenge as well as lost opportunity for the EOS network, which can begin to be addressed through the adoption of this new resource model with some tweaks.
With proper experimentation, testing, and tuning of the resource model to more accurately reflect the instantaneous price for network resources at any given time, the interests of large token holders and network operators can also be aligned, by setting the network up for long term viability and fair use, and optimizing the traffic on the network for every type of user, stakeholder or not. The more real people deriving real value from the EOS network, the more this network will translate into real utility and value for use cases beyond academia and mining schemes.
Thank you for reading our thoughts. If you’d like to discuss more, please reach out to us on our Telegram group.
CEO, EOS Detroit