Premium Sharing Pool Insights

Updated 21 November 2025

Premium sharing pools are mechanisms that allocate a total reward based on participants' contributions, ensuring budget-balance and incentive compatibility.
They are applied in blockchain mining, staking, insurance risk pooling, and dynamic transit systems, using parameters like k and δ to tune fairness and urgency.
Design trade-offs in these pools involve balancing efficiency, Sybil resistance, and bonus allocation through explicit structural and mathematical frameworks.

A premium sharing pool is a general mechanism for collective risk sharing or reward allocation in which participants (agents, miners, riders, insureds, etc.) contribute work, stake, or capital and receive rewards, surplus, or incentives that may include explicit premiums or bonuses. The structure and incentive properties of premium sharing pools are foundational to blockchain mining, proof-of-stake economies, insurance risk allocation, and transportation systems where early or critical participants are rewarded with a premium, or where explicit resilience and incentive tradeoffs are engineered. The literature provides distinct but formally related frameworks for premium sharing in mining, staking, risk pooling, surplus-sharing insurance, and dynamic fare-controlled ride-sharing, all of which employ structural parameters to tune fairness, incentive compatibility, efficiency, and robustness.

1. General Structure and Mathematical Formulation

Premium sharing pools allocate a total reward or surplus $R$ among $n$ participants according to a rule $f$ . The design of $f$ determines how contributions (work shares, pool size, capital, ride participation) and performance (e.g., being first, solving a block) map into payouts. In all formulations, budget-balance is enforced: $\sum_{i=1}^n f_i = R$ The literature distinguishes between:

Discrete share-based pools (e.g., mining): Rewards are partitioned according to share submission order and round-specific fairness axioms (Can et al., 2021).
Stake/capital pools (e.g., PoS, insurance): Participants receive base compensation and potentially a premium based on a leader/operator role or excess contribution (Brünjes et al., 2018, Coculescu et al., 2018).
Dynamic, networked pools (e.g., mobility): Premiums are set via real-time control to incentivize desired pooling behaviors while respecting externalities (Fayed et al., 2023).

The precise definition of "premium" varies: in mining it is an early/last-share bonus; in staking, a margin for pool leaders; in insurance, surplus above fair risk premium; and in urban mobility, a fare discount coupled with privileged access.

The "premium sharing pool" in mining is formalized via a reward-sharing function acting on an ordered multiset of shares $S = \{s_1, \dotsc, s_n\}$ . Two main fairness axioms shape the structure (Can et al., 2021):

Absolute redistribution: Adding a late share reduces each old reward by a fixed amount.
Relative redistribution: Adding a late share scales each old reward by a fixed ratio.

These axioms yield, respectively, the classes of:

Absolute-fair schemes: Parametrized by sequences $\varepsilon(k)$ , yielding payoffs for rank $j$ , round size $n$

$f(j, n; R) = R \left[ \varepsilon(j) - \sum_{i=j+1}^n \frac{\varepsilon(i)}{i-1} \right]$

Relative-fair schemes: Parametrized similarly, but multiplicative

$f(j, n; R) = R\, \varepsilon(j)\; \prod_{i=j+1}^n [1 - \varepsilon(i)]$

Intersection yields the k-pseudo-proportional family: $f(j, n; R) = \begin{cases} \frac{R}{n}, & n < k \ \frac{R-\delta}{k-1}, & n\ge k,\, j<k \ \delta, & n\ge k,\, j = k \ 0, & n\ge k,\, j>k \end{cases}$ This family enables explicit early or solving-share premiums (tuned by $\delta$ ) while remaining budget-balanced and incentive-compatible, interpolating between strict proportionality (no premium) and winner-takes-most (maximal premium). Pool designers can use $k$ to cap latecomers' eligibility, creating urgency, and $\delta$ to split the premium (Can et al., 2021).

For proof-of-stake or collaborative resource pools, premium sharing pools are designed to generate k pools of target size $\beta=1/k$ while balancing Sybil resistance and cost efficiency (Brünjes et al., 2018). The generic reward-sharing mechanism proceeds as follows:

Reward cap: For pool $i$ of total stake $o_i$ and committed stake $\sigma_i$ ,

$r(o_i, \sigma_i) \propto \min(o_i, \beta) + a'\, \min(\sigma_i, \beta)$

where $a$ tunes between efficiency ( $a=0$ ) and Sybil-resilience ( $a\rightarrow\infty$ ).

Premium split: Each operator receives cost reimbursement, a margin $m_i$ of surplus, and a pro-rata share. Delegators share the remainder proportionally: $R_{\mathrm{op},i} = \min\{c_i, r\} + m_i(r-c_i)^+ + (1-m_i)(r-c_i)^+\, \frac{\sigma_i}{o_i}$

$R_{j\to i} = \begin{cases} (1-m_i)\,(r-c_i)\frac{s_{j\to i}}{o_i}, & r>c_i \ 0, & r\leq c_i \end{cases}$

Exact k-pool equilibrium is established by setting optimal margins per operator: $m_i^* = \frac{P(s_i, c_i) - P(s_{(k+1)}, c_{(k+1)})}{P(s_i, c_i)}$ where $P(s_i, c_i) = r(\beta, s_i) - c_i$ and only the top-k potential-profit operators survive. Sybil costs for adversaries scale linearly in $t$ (number of identities).

Surplus/premium sharing pools in insurance are modeled via coherent risk (utility) measures and the Euler (marginal) allocation principle (Coculescu et al., 2018). Each insured i pays a premium $p_i \geq T_i$ , where $T_i$ is the fair charge under the risk measure, and the overage $p_i - T_i$ constitutes a capital contribution. The pool’s retention $R$ is

$R = k_0 + \sum_{i=1}^N (p_i - T_i)$

Total claims $L$ are capped at $R$ ; the realized surplus is $S = R - \min(L, R)$ . The surplus is allocated in proportion to contributed capital (insurer and insureds); for insured i,

$S_i = \frac{p_i - T_i}{k_0 + \sum_j (p_j - T_j)} \cdot S$

Fair premiums are derived by Euler allocation: $T_i = E_{Q^*}[X_i]$ where $Q^*$ solves the dual for the pool-wide risk measure $\rho(\cdot)$ . This structure ensures that risk capital and surplus are divided equitably and that each insured’s deal is acceptable under coherent utility as long as $p_i \leq \sup_{Q\in M} E_Q[X_i]$ (Coculescu et al., 2018).

For pools of i.i.d. risks $X_1, \dotsc, X_n$ , the per-agent benefit from pooling depends on the underlying risk measure (Knispel et al., 2021). Under law-invariant, rank-dependent utility risk measures (e.g., AVaR, dual utility), the limiting risk premium decays sublinearly,

$\tilde\pi_n = \mu - \rho(\overline{S_n}) = c n^{-1/2} + o(n^{-1/2})$

where

$c = \sigma \int_0^1 \Phi^{-1}(t) p(dt)$

and $p$ is the distortion measure parameterizing risk aversion. Aggregate premium saving grows as $c n^{1/2}$ . In contrast, under classical expected utility with no distortion, premium decay is $O(n^{-1})$ . This quantifies the rate at which collective pooling reduces risk premia in large pools and informs capital pricing and risk allocation in premium-sharing arrangements (Knispel et al., 2021).

Premium sharing in urban mobility context refers to fare discounts (the premium) for ride-pooling, coupled with privileged access (e.g., dedicated bus lanes). A dynamic macroscopic model partitions the city network into vehicle and bus subnetworks (Fayed et al., 2023). Key elements include:

Fare adjustment ΔP(t): Discount for users who agree to pool and use the bus lane.
Macroscopic state variables: Proportions of private cars, solo ride-hail, pooled ride-hail in vehicle/bus subnetworks.
Dynamic control:
- Proportional-Integral (PI) controller adjusts pooling fares to keep bus speeds above setpoint,
- Model Predictive Control (MPC) minimizes global passenger-hours traveled and waiting time while imposing speed constraints.

Simulations show that enabling pooling in the bus lane with fare discount reduces total passenger-hours by up to 25%. Dynamic tuning of premiums (fare adjustments) using MPC enables further reductions and provides operational trade-offs between efficiency and modal delays.

7. Design Trade-Offs and Implementation Considerations

Parameter tuning in premium sharing pools provides multi-dimensional trade-offs:

k (pool window, number of pools): Larger $k$ favors fairness but dilutes urgency; smaller $k$ magnifies speed/premium incentives.
δ (bonus weighting): Larger $δ$ heavily rewards critical/final contributors; smaller $δ$ allocates more to marginal/early contributors (Can et al., 2021).
a (Sybil-resilience): Modulates cost-efficiency versus resistance to Sybil attacks in stake pools (Brünjes et al., 2018).
Regulatory and data constraints: For insurance pools, surplus and premium allocations intersect with capital requirements and calibration of scenario sets for risk measures (Coculescu et al., 2018). In dynamic fare pooling, accurate real-time data and computational capacity for on-the-fly optimization are essential (Fayed et al., 2023).

A plausible implication is that premium sharing pools provide a unifying formal perspective for incentive-aligned collective schemes, with the structure of the premium and the pooling rule determining efficiency, fairness, and robustness against manipulation or moral hazard.

References:

Can, Hougaard, and Pourpouneh, "On Reward Sharing in Blockchain Mining Pools" (Can et al., 2021)
BFT, Kiayias, Koutsoupias, Kyropoulou, "Reward Sharing Schemes for Stake Pools" (Brünjes et al., 2018)
Coculescu & Delbaen, "Surplus sharing with coherent utility functions" (Coculescu et al., 2018)
Knispel, Laeven, Svindland, "Asymptotic Analysis of Risk Premia Induced by Law-Invariant Risk Measures" (Knispel et al., 2021)
Seber, Yildiz, Saberi, "A Dynamic Macroscopic Framework for Pricing of Ride-hailing Services..." (Fayed et al., 2023)