Nym Token Economic Highlights
This blogpost gives an insight into the science behind the Nym token economics: high level takeaways of the Nym Reward Sharing for Mixnets…
This blogpost gives an insight into the science behind the Nym token economics: high level takeaways of the Nym Reward Sharing for Mixnets paper, written by Claudia Diaz, Aggelos Kiayias and Harry Halpin. Find the pre-print here.
The Reward Sharing for Mixnets paper outlines the token economics of the Nym mixnet and reports on token economic simulations. It presents the economic model for all Nym participants and the mechanisms for allocating rewards to mix nodes and further distributing these among the node operator and the stakeholders delegating NYMs to the node.
This blogpost describes the overall systemic properties. Formulas and possible parameters for calculating APY are in the paper, and a blog post next week will elaborate on results of simulations for two example scenarios.
For the impatient reader, here are the high-level takeaways. The paper shows that:
- The overall scheme can sustain an economically viable decentralised mixnet: incentivising NYM stakeholders to register (“bond” or “pledge” stake) and support (“delegate” stake) a sufficient number of well-performing mix nodes, ensuring network capacity, scalability, decentralisation and quality of service.
- That the total bonded + delegated stake serves as a mix node’s reputation and that it matters: mix nodes that attract a lot of stake will be selected more often to be active in the network, as well as being rewarded at a higher rate, than nodes with lower reputation.
- The reward scheme rewards well-performing nodes while underperforming mix nodes miss out on rewards. Node performance affects both the rewards of the node operator and of the stakeholders supporting the node with delegated stake, incentivising stakeholders to support nodes that maximise the network’s Quality of Service.
- There is an increase in reward rate for higher bonded stake, incentivising node operators to focus all their stake on running one node. The reward scheme penalises (with lower rewards) stakeholders that split their stake to register multiple mix nodes, thus making Sybil attacks economically disadvantageous.
- Mix node operators receive higher reward rates than delegators (which is fair, given that they are covering the cost of running nodes and putting in labour to ensure operations).
- Even if you have a small token holding, you can earn steady rewards by operating a high-reputation mix node supported by delegation from other stakeholders.
- Decentralisation is incentivised. There is a ‘softcap’ on reputation, after which more stake will not increase rewards so that the network does not centralise on a few nodes.
NOTE: The paper does not say anything about the real world profitability of an individual node as that will depend on the overall distribution of stake among nodes as well as individual node parameters including performance and reputation. To be clear, this is an academic paper and not an investment prospectus or financial advice. We use the term ‘APY’ in a loose generic sense based on ‘Return on Stake’, and are relying on hypothetical scenarios using experimental technology. Actual results using the technology will vary widely. It’s quite possible that this venture could result in loss of monetary funds for anyone who uses the software.
Overall findings from Nym token economic simulations
The paper reports on simulation results and mathematical proofs that the reward scheme has the following properties, namely that the Nym token economics…
- supports node operations
The Nym mixnet reward scheme prioritises, in the first instance, the viability of mix node operators, ensuring that they cover operational costs and can make a sufficient additional profit to be incentivised to operate. Even for nodes that can only initially stake a small bond, as long as their performance is good and their overall reputation is high (meaning that the node is close to saturation thanks to delegated stake), nodes receive a consistent amount of NYM rewards even if the network has low income from paid usage in the first few years. Once sufficient rewards have been dedicated to ensure operator incentives and viability, the scheme compensates delegators proportionally to their support of well-performing, high-reputation nodes.
- supports network scaling
As mixnet usage demand increases, the reward scheme provides mechanisms for funding the scaling of the network with usage fees. When the network demand grows, so does the network income, as well as the number of rewarded nodes, which increases as needed to meet user demand. Inversely, the stake saturation point decreases, as stake becomes further spread over a larger number of nodes, meaning that fewer tokens go a longer way in increasing a node’s reputation. The simulations indicate that in high-usage growth scenarios the per-node rewards remain stable, while the per-staked-token returns increase proportionally to the additional network income.
- supports good quality of service
Mix nodes do not just get rewarded for running the Nym mixnet, they get rewarded for running it well. Nodes that are unavailable or that fail to follow the routing protocols miss out on rewards, which incentivises delegators to move their support (delegated stake) to better performing nodes. The underperforming node will thereby have lower reputation and be selected less frequently to the network and will lose out on rewards. Delegated stake thereby serves as a form of decentralised quality control.
- supports decentralisation
There is a stake saturation point, meaning a point after which more stake does not economically benefit a node. This is to encourage delegators to spread their delegations across a broader number of well-performing nodes.
The saturation point thereby defines a ‘softcap’ on how much stake it makes sense to accumulate on a single node. Until reaching this stake saturation, a node’s rewards increase steadily as the node increases its reputation, meaning that new delegates increase the node’s rewards with their delegated stake (and receive a part of the extra income themselves). Once saturation is reached, however, rewards max out and additional delegated stake does not further increase node rewards. Instead, when a node is oversaturated the same node rewards have to be shared among more stakeholders, making the returns per staked token lower than nodes that are not oversaturated. The system thus disincentivises the concentration of stake on too few nodes.
Delegators maximise their returns when they provide maximum support (stake at the point of saturation) to a target number of nodes with perfect performance, with the target number being the number of rewarded nodes (720 nodes in the reward set). This distribution of stake over nodes corresponds to the system equilibrium, which is the configuration in which all stakeholders maximise their returns.
- economically disincentivises Sybil attacks
The reward algorithm rewards nodes with larger operator bonds at a slightly higher rate. Given two nodes with identical performance and reputation, the node with the larger bond will receive higher rewards than the node with the smaller bond.
As a result, stakeholders running nodes maximise their rewards by registering a single node with the largest possible bond (up to the stake saturation level of one million NYM), rather than splitting their holdings into multiple nodes with small bonds.
This property of the reward algorithm incentivises that each node is run by an independent party that invests all their stake on one node. Stakeholders who split their stake among many nodes, in an attempt to control a large part of the network, are on the other hand penalized with lower rewards.
