> ## Documentation Index
> Fetch the complete documentation index at: https://docs.safefoundation.org/llms.txt
> Use this file to discover all available pages before exploring further.

# Rewards

> How SAFE-denominated rewards are calculated and distributed to Safenet Validators and Delegators.

<Note>
  The reward parameters described on this page are proposed values, pending SafeDAO approval, and may change before Safenet Aegis launches.
</Note>

Safenet Validators and Delegators can receive SAFE-denominated rewards through the Validator staking system.

Rewards are calculated in recurring **two-week reward periods**. SafeDAO allocates **5,000,000 SAFE over 12 months** to Validator and Delegator rewards.

This page defines the Validator and Delegator reward calculation and distribution rules for Safenet Aegis.

## Who can earn rewards

Validators can earn rewards on their self-staked SAFE. SAFE holders who delegate SAFE to a Validator can earn Delegator rewards.

A Validator must reach at least **75% participation** during a reward period for the Validator and its Delegators to receive rewards for that period.

Below 75% participation:

* The Validator receives no reward for the period.
* Its Delegators receive no reward for the period.

At or above 75%, rewards scale with the Validator's participation.

Validators receive a fixed **5% commission** on rewards attributable to delegated SAFE. The remaining 95% is distributed among their Delegators.

Validators must also maintain a minimum time-weighted self-stake of **3,500,000 SAFE** over the reward period. Below this threshold, a Validator earns no commission on delegated stake for that period, though rewards on its own self-stake are unaffected. Delegators remain eligible for rewards provided the Validator still meets the participation requirement.

Validator and delegated SAFE are not currently subject to slashing. Validator participation and behavior can affect rewards, but do not reduce staked SAFE.

## Reward calculation

Rewards depend on:

* **Time-weighted stake:** The amount of SAFE staked and how long it remains staked during the reward period.
* **Stake concentration:** Additional stake contributes less reward weight when a Validator is backed by substantially more stake than the network average.
* **Validator participation:** Validators must reach 75% participation, and higher participation above the threshold increases reward weight.
* **Minimum self-stake:** Validators must maintain at least 3,500,000 SAFE self-staked to earn commission on delegated stake.
* **Validator commission:** Validators receive 5% of rewards attributable to delegated stake.
* **Reward eligibility:** Validators must only participate in eligible attestations.

**Approximate reward rate**

A rough annualized network-wide reward rate is:

$$
\text{Approximate reward rate}
=
\frac{5{,}000{,}000\ \text{SAFE}}
{\text{Total SAFE staked across the network}}
$$

For example, if 50,000,000 SAFE is staked across the network, the unadjusted network-wide rate would be approximately **10% per year**.

This is only an approximation. Individual rewards depend on the factors described below.

**Reward flow**

At a high level, the reward pool is allocated to Validators according to their effective reward weight. Each Validator's allocation is then split between rewards attributable to self-stake and delegated stake.

```mermaid theme={null}
flowchart TD
    classDef neutral fill:#555555,stroke:#555555,color:#ffffff
    classDef green   fill:#00855a,stroke:#00855a,color:#ffffff
    classDef blue    fill:#1a6fa8,stroke:#1a6fa8,color:#ffffff

    A["Reward pool"]:::neutral -->|"Proportional to effective reward weight"| B["Validator share"]:::neutral

    B --> C["Self-stake rewards"]:::green
    B --> D["Delegated rewards"]:::neutral

    C --> E["Validator"]:::green

    D -->|"5% commission"| E
    D -->|"95%"| F["Delegators"]:::blue

    F -->|"Pro rata by time-weighted stake"| G["Delegator A"]:::blue
    F -->|"Pro rata by time-weighted stake"| H["Delegator B"]:::blue
```

*How rewards flow from the pool to a single Validator and its Delegators in a reward period.*

**Time-weighted stake**

Rewards use the average amount of SAFE staked during the reward period rather than a single snapshot.

SAFE therefore contributes to rewards only for the time it was actually staked or delegated.

For example, SAFE delegated halfway through a reward period contributes for approximately half of that period rather than receiving the same reward as SAFE delegated for the full period.

**Stake-concentration weighting**

Reward weighting discourages excessive stake concentration with a single Validator.

Reward weight grows linearly up to the average stake backing a Validator. Above that point, additional stake increases reward weight more slowly.

This means additional SAFE delegated to an already heavily staked Validator can contribute less reward weight than the same amount delegated to a smaller Validator.

**Participation weighting**

A Validator's stake weight is adjusted by its participation rate.

Below **75% participation**, its effective reward weight is zero.

At or above 75%, the Validator's effective reward weight increases with participation.

**Minimum self-stake requirement**

A Validator's commission eligibility is separately gated by its own self-stake.

Below **3,500,000 SAFE** in time-weighted self-stake, a Validator earns no commission on delegated stake for that period. Rewards on the Validator's own self-stake, and rewards for its Delegators, are unaffected.

**Allocating rewards**

Each Validator receives a share of the reward pool proportional to its effective reward weight.

That allocation is attributed between:

* Rewards generated by the Validator's self-stake.
* Rewards generated by SAFE delegated to the Validator.

The Validator receives all rewards attributable to its self-stake. If the Validator meets the minimum self-stake requirement, it also receives a fixed **5% commission** on rewards attributable to delegated stake, with the remaining **95%** distributed among Delegators in proportion to their time-weighted stake. If the Validator does not meet the minimum self-stake requirement, the full delegated-stake reward is distributed among Delegators instead.

## Eligibility and exclusions

Meeting the participation threshold is not sufficient on its own. Validators must also participate only in FROST signing ceremonies for results that are eligible under the Safenet result model.

A Validator is subject to a reward exclusion if it participates in a FROST signing ceremony for a result that is not an eligible unanimous secure Sentinel result.

This includes signing ceremonies for:

* An insecure Sentinel result.
* A split Sentinel result.
* An incomplete or otherwise unreliable result.
* A result that entered arbitration.
* A mismatched Safe, network, or transaction context.
* Any other result that is not an eligible unanimous secure Sentinel result.

If a Validator participates in such a signing ceremony, the following rewards are excluded for the **entire two-week reward period**:

* The Validator's self-stake rewards.
* The Validator's commission.
* Delegator rewards attributable to that Validator.

An insecure or otherwise ineligible result does not trigger an exclusion by itself. The exclusion applies as soon as the Validator participates in a signing ceremony for such a result, whether or not that ceremony produces a final attestation.

Rewards excluded because a Validator participated in a signing ceremony for an ineligible result are redistributed among the remaining qualifying reward recipients.

## Distribution and claiming

Rewards are calculated every two weeks using the applicable staking and Safenet participation state for the reward period.

Validator self-stake rewards and commission are allocated to the Validator's Staker address registered at the end of the reward period.

A staking interface can show rewards associated with a staking position and provide the available claiming flow.

Reward calculation is performed offchain using open-source tooling that anyone can run and independently verify. Proposing, challenging, and executing accepted distributions are permissionless.

The calculation produces a deterministic reward package containing the information needed to verify and publish the distribution, including:

* Recipient allocations.
* Merkle proofs.
* Merkle root.
* Reward dataset.
* Transactions required to publish the distribution.

See the [safenet-staking-scripts repository](https://github.com/safe-fndn/safenet-staking-scripts/) for the reward calculation and distribution implementation.

The accepted distribution is executed through the **Reward Distribution Safe on Ethereum Mainnet**. The Safe controls the rewards-contract Merkle root and the `rewards.safenet-gov.eth` ENS reference.

TODO: Add the Reward Distribution Safe address once deployed.

Publishing a distribution follows a public flow:

1. The reward calculation and corresponding dataset are produced and published.
2. Anyone can independently reproduce and verify the calculation.
3. Anyone can propose the distribution through SafeSnap.
4. Anyone can answer or challenge the proposed distribution on Reality.eth.
5. Once a `Yes` answer is final and the cooldown has elapsed, anyone can execute the accepted distribution.
6. Execution through the Reward Distribution Safe updates the canonical onchain Merkle root and the ENS reference to the corresponding reward dataset.

Safe Foundation does not approve or execute individual reward distributions.

The onchain Merkle root determines which reward claims are accepted by the distribution contract. `rewards.safenet-gov.eth` references the CID of the corresponding canonical reward dataset. Anyone can independently reproduce the calculation, verify the dataset, and mirror or pin it.

There is **no minimum payout amount**. Every positive reward allocation can be distributed.

Reward distributions are subject to standard sanctions and AML checks. Recipient addresses are screened against sanctions lists maintained by the UN, EU, UK, US (OFAC), and Switzerland (SECO). See the [sanctions-address-lists repository](https://github.com/safe-research/sanctions-address-lists) for the address lists used.

Rewards are not guaranteed. The amount received can change between reward periods as stake, Validator participation, stake distribution, and reward eligibility change.

See [How to stake](/safenet-aegis/staking/how-to-stake) for delegation and [Lock-up and withdrawals](/safenet-aegis/staking/lock-up-and-withdrawals) for withdrawal rules.

## Mathematical specification

The following defines the reward calculation formally.

**Participants and stake**

* `V`: Set of Validators.
* `N = |V|`: Number of Validators.
* `s̄_i`: Time-weighted self-stake of Validator `i`.
* `d̄_{j,i}`: Time-weighted SAFE delegated by Delegator `j` to Validator `i`.
* `d̄_i = Σ_j d̄_{j,i}`: Total time-weighted delegated stake to Validator `i`.
* `S̄_i = s̄_i + d̄_i`: Total time-weighted stake backing Validator `i`.
* `S̄_total = Σ_i S̄_i`: Total time-weighted SAFE staked across the Validator set.

**Participation**

* `p_i`: Participation rate of Validator `i`.
* `p_min = 0.75`: Minimum participation required to receive rewards.

**Reward parameters**

* `R`: SAFE reward pool for the reward period.
* `T`: Average stake backing a Validator.
* `c = 0.05`: Validator commission on delegated rewards.
* `s_min = 3{,}500{,}000` SAFE: Minimum time-weighted self-stake required for a Validator to earn commission.

**Time-weighted stake**

For any stake balance `x(t)` during a reward period `[t_0, t_1]`:

$$
\bar{x} = \frac{1}{\Delta t}\int_{t_0}^{t_1}x(t)\,dt
$$

where:

$$
\Delta t = t_1 - t_0
$$

**Average stake threshold**

The threshold for linear reward growth is:

$$
T = \frac{\bar{S}_{total}}{N}
$$

**Stake weight**

Reward weight grows linearly up to the average stake threshold and sub-linearly above it:

$$
w(\bar{S}_i) =
\begin{cases}
\bar{S}_i, & \text{if } \bar{S}_i \le T \\
\sqrt{\bar{S}_i \cdot T}, & \text{if } \bar{S}_i > T
\end{cases}
$$

**Effective reward weight**

Participation is applied to the stake weight:

$$
\tilde{w}_i =
\begin{cases}
p_i \cdot w(\bar{S}_i), & \text{if } p_i \ge p_{min} \\
0, & \text{if } p_i < p_{min}
\end{cases}
$$

The total effective reward weight is:

$$
W = \sum_{k \in V}\tilde{w}_k
$$

**Validator allocation**

Validator `i` receives:

$$
R_i = R \cdot \frac{\tilde{w}_i}{W}
$$

The allocation attributable to self-stake is:

$$
R_i^{self} = R_i \cdot \frac{\bar{s}_i}{\bar{S}_i}
$$

The allocation attributable to delegated stake is:

$$
R_i^{del} = R_i \cdot \frac{\bar{d}_i}{\bar{S}_i}
$$

**Validator and Delegator rewards**

If the Validator meets the minimum self-stake requirement (`s̄_i ≥ s_min`), it receives its self-stake reward plus commission on delegated rewards:

$$
R_i^{validator} = R_i^{self} + c \cdot R_i^{del}
$$

Otherwise, it receives only its self-stake reward:

$$
R_i^{validator} = R_i^{self}
$$

Delegators collectively receive the remainder of the delegated-stake reward:

$$
R_i^{delegators} =
\begin{cases}
(1-c) \cdot R_i^{del}, & \text{if } \bar{s}_i \ge s_{min} \\
R_i^{del}, & \text{if } \bar{s}_i < s_{min}
\end{cases}
$$

For Delegator `j`:

$$
R_{j,i} = R_i^{delegators} \cdot \frac{\bar{d}_{j,i}}{\bar{d}_i}
$$

The resulting allocations remain subject to the participation and attestation-eligibility rules described above.

<CardGroup cols={2}>
  <Card title="How to stake" href="/safenet-aegis/staking/how-to-stake">
    Choose a Validator and delegate SAFE.
  </Card>

  <Card title="Lock-up and withdrawals" href="/safenet-aegis/staking/lock-up-and-withdrawals">
    Review the withdrawal process and waiting period.
  </Card>

  <Card title="Risks" href="/safenet-aegis/staking/risks">
    Review the risks of staking and delegation.
  </Card>

  <Card title="Staking" href="/safenet-aegis/staking/overview">
    Return to the Safenet staking overview.
  </Card>
</CardGroup>