XTZ sits at the control surface of Tezos, not just the fee layer

Tezos is built to change itself. The protocol’s rules can be replaced through an on-chain amendment process rather than a social-layer hard fork, which makes “tokenomics” a moving target by design.

That design choice matters for XTZ holders because XTZ is both security collateral and governance weight. Voting power is exercised by delegates (bakers), and a delegate’s voting power is tied to its staking balance, which includes staked tez plus delegated tez.

On the product side, XTZ is the unit that gets locked to secure consensus and earn protocol-issued rewards via staking. Staked tez cannot be spent while staked, and unstaking becomes spendable only after a delay.

The market-structure consequence is straightforward. Tezos does not have a single, static “supply story.” It has a policy surface that can be tuned, and that tuning changes who holds liquid XTZ, who holds locked XTZ, and who receives new issuance first.

Supply is uncapped, and issuance is now responsive to staking participation

Market data sources list Tezos’ maximum supply as infinite.

Those same sources also report roughly 1.1 billion XTZ as circulating or tradable today. For a contrasting supply narrative, compare this with our Quant tokenomics review.

Mechanically, new tez can be created through three channels in the economic protocol: participation rewards (consensus and related duties), Liquidity Baking (if enabled), and protocol “invoices” issued at protocol migration, as described in the issuance channels documentation.

The headline tokenomics change in the last cycle was Paris. Paris was activated on June 5, 2024, and it introduced adaptive issuance and staking as a major Proof-of-Stake overhaul, per the protocol upgrade timeline.

Adaptive issuance ties the regular issuance budget to the global staked ratio, with the protocol nudging staking participation toward a target of 50% starting in the Paris protocol.

In the adaptive issuance documentation, the adaptive issuance rate is explicitly described as the sum of a static rate and a dynamic rate, bounded by a minimum and maximum to keep nominal emissions within a range.

The protocol’s bounds are not cosmetic. In Paris, the issuance minimum and maximum begin near the pre-adaptive regime and then widen over time. The documented parameters include an initial minimum of 4.5% and initial maximum of 5.5%, transitioning toward a global minimum of 0.25% and a global maximum of 10%.

The same documentation anchors the transition context: at activation, the staked ratio was around 7.5% and the pre-activation issuance rate was around 4.6% under Oxford, which adaptive issuance was designed to transition from smoothly.

Then Quebec tightened the upper tail. Quebec was activated on January 20, 2025, and introduced an “adaptive maximum” issuance bound.

The adaptive issuance documentation describes this adaptive maximum as preventing undesirably high issuance when staking is already near the 50% target, falling to a 1% maximal issuance rate when the staked ratio is at 50% or above.

Rio changed cadence and reward routing. Rio was activated on May 1, 2025, reduced cycles to roughly one day, and allocated 10% of participation rewards to Data Availability Layer (DAL) participation once DAL is active.

Rio’s DAL incentive only becomes fully “in the money” under a coordination condition. DAL becomes active only when at least 66% of total baking power participates.

From a microstructure lens, Tezos’ emission profile is now a feedback controller. When staking participation is low relative to the 50% target, issuance pressure can rise toward the documented upper bounds. When staking participation rises, issuance pressure can compress toward the lower bounds and, under Quebec’s adaptive maximum, the ceiling itself slopes down near target.

Genesis distribution was broad in wallet count, concentrated in size, and included time-locked pools

Tezos’ genesis allocations were explicitly published by the Tezos Foundation ahead of launch, and the genesis allocations disclosure lays out the genesis mint and the vesting treatment for key pools.

The same disclosure states an initial supply of 763,306,929.69 ꜩ in the genesis block, with 20% locked in vesting contracts.

The wallet-count distribution looked broad. The Foundation reports 30,317 wallets funded.

The size distribution looked concentrated. The same post reports a Gini coefficient of 0.878 for contributions.

As a market-structure analyst, I care less about the existence of vesting and more about its shape. This was monthly vesting over four years, so it likely behaved more like a steady drip than a single cliff.

The other structural concentration vector is institutional treasury. The Tezos Foundation remains a major economic actor, and it discloses endowment composition. As of December 31, 2025, it reports holding “XTZ USD 76 million” as part of its asset base.

Fees go to block producers, storage growth burns tez, and upgrades can mint lump-sum invoices

On Tezos, users pay fees for operations, and the fee flow is designed to accrue to the block producer at L1.

Tezos also charges for state bloat. Storage has a cost that can be accounted for either by paying a fee to a baker or by destroying (“burning”) some tez, with a burn cap applied when storage increases.

From a value-capture standpoint, this splits fiscal flows into two pipes. Compute and inclusion fees accrue to validators. Some storage-related costs are explicitly destructive, which can offset issuance at the margin during periods of heavy storage growth.

The more unusual pipe is governance-time issuance. The protocol can issue “invoices,” which are lump sums minted and allocated during protocol migration, with recipients and values defined by the developers of a protocol proposal.

Invoices are not a day-to-day emission. They are structurally similar to an unlock event, because they can create one-off supply at a known protocol boundary. They are also harder to model from a pure “inflation rate” narrative, because they are proposal-scoped rather than mechanically scheduled.

Governance can rewrite the token economy, and it has already rewritten it

Tezos’ amendment process is split into five periods: Proposal, Exploration, Cooldown, Promotion, and Adoption.

The Tezos documentation states that each period lasts 14 blockchain cycles, or about 14 days.

Only delegates vote, and the documentation defines a delegate’s voting power as tied to its staking balance.

After Rio shortened cycles, the governance process duration remains 70 days total, with voting periods still 14 days.

This is the key trade-off. On-chain governance gives Tezos narrative stability around upgrades. You get fewer existential fork fights. The cost is that monetary policy is, in practice, a governed parameter set. Paris and Quebec changed issuance logic. Quebec changed how different forms of XTZ (staked vs delegated) count toward rights and rewards. Rio changed the reward routing by earmarking part of participation rewards to DAL work. For definitions and common framing, see our tokenomics FAQ.

Float is now a policy variable: staking locks, exit delays, and stake caps shape liquidity

Tezos now supports staking at the protocol level, and staked tez remains in the staker’s account but cannot be transferred or spent while staked.

Liquidity comes back with a delay. Unstaking requires waiting up to 4 days for unstaked funds to be made available, due to cycle-based rights computation and slashing windows.

Delegation remains liquid. Delegating incurs no risk to the delegator, retains full control, and tez can be spent at any time, while staking with a baker exposes the stake to punishment if the baker misbehaves.

This difference matters for price behavior because it creates a two-layer float. There is the liquid float that can hit exchanges immediately. There is also the staked float that can only re-enter circulation on a deterministic delay. When risk rises, unstake decisions can cluster, and the unlock becomes a dated liquidity event rather than a continuous one. For a liquid-staking comparison point, see our Lido tokenomics review.

Tezos also started explicitly shaping who can intermediate staking at scale. Quebec raised the limit for how much external stake a baker can accept to 9x the baker’s own staked balance, up from 5x in Paris.

Quebec also reduced the weight of delegated funds toward baking power from half to one-third.

That is a direct market-structure lever. It pushes the system toward “stake for security, delegate for convenience,” and it shifts governance influence toward those willing to accept lockup and slashing risk. It also concentrates operational importance into bakers who can warehouse stake capacity without becoming overstaked.

Rio tightened the tempo. Cycles are now one day long, and this shortens several delays, including unstaking finalization to 4 days and payout cadence to daily for some reward components.

The net effect is a token economy that can move from “low staking, higher issuance” toward “higher staking, lower issuance,” but it does so through incentives that can themselves create sell-pressure. Participation rewards are paid in XTZ. Many recipients will monetize. Tezos’ adaptive issuance mechanism is, intentionally, a security budget. Markets experience it as a stream of inventory.

Risk analysis: Tezos is modelable, but policy-driven liquidity shocks remain the core hazard

Dominant risk: adaptive issuance reflexivity creates a volatility loop between security incentives and circulating inventory.

Trigger: the staked ratio persists meaningfully below the protocol target of 50%.

Mechanism: the protocol increases the issuance budget to “nudge” staking upward, and the adaptive issuance rate is bounded by a widening min/max band that can reach a documented global maximum of 10%.

Who bears it: holders who stay liquid and do not capture issuance become mechanically diluted. Stakers capture the new issuance, but they also become the first recipients of new inventory that tends to be sold to cover costs or reduce exposure. The market absorbs that flow.

Why this dominates: Tezos’ “inflation” is no longer a single parameter. It is a response function tied to behavior. If the chain fails to attract staking participation, Tezos responds by paying more. That can work for security. It can still be a weak price regime if marginal demand does not scale with the higher issuance stream. In that scenario, adaptive issuance behaves like a governor that keeps opening the fuel line because the engine is stalling, and the exhaust becomes more inventory on the tape.

Quebec’s adaptive maximum helps once the system is near target. It slopes the ceiling down and can compress issuance near 50%, including a 1% max issuance rate at or above target per the documentation.

That still leaves the path dependence. Getting from low staking to high staking requires sustained incentive. Incentive is issuance. Issuance is sellable inventory. The system can stabilize. It can also churn, especially if a large share of stake is intermediated by custodians and large bakers whose reward distribution policies push rewards onto exchanges quickly.

Measurable indicators: (1) on-chain staked ratio versus the 50% target, (2) observed issuance rate relative to the documented min/max bands, (3) concentration of stake capacity given the 9x external stake limit, (4) the liquid-versus-staked mix implied by staking lockups and unstake queues (4-day unlock path). We track similar monitoring patterns in our research reports.

  1. Trigger: a protocol upgrade proposal includes a sizable protocol “invoice” or materially changes issuance parameters.
    Mechanism: invoices are lump-sum tez minted and allocated during migration, and their value and recipients are discretionary to the proposal authors, creating an upgrade-bound liquidity event.
    Who bears it: passive holders and liquid holders absorb dilution and the incremental sell-side. Recipients face execution risk if they liquidate into thin books.
    Measurable indicators: proposal changelogs, governance forum chatter, and any explicit invoice lines in migration documentation.
  2. Trigger: concentrated baker or custodian stress causes correlated unstaking or stake migration.
    Mechanism: staked tez is illiquid until the unstaking delay completes, so risk events can produce dated waves of liquidity returning to the market on a similar timeline, amplifying short-term imbalance.
    Who bears it: stakers with the affected baker bear operational risk and potential penalties; spot holders bear the sell pressure when unlocked funds re-enter liquid float.
    Measurable indicators: stake share by baker, abrupt changes in delegated and staked balances, and changes in unstake requests over time.
  3. Trigger: stake centralization rises because intermediaries can warehouse the 9x external stake capacity and delegated weight is discounted to one-third for rights.
    Mechanism: governance and reward capture tilt toward a smaller operational set, which can reduce the diversity of execution policies for selling rewards, and increase the market impact of a few actors’ treasury decisions.
    Who bears it: smaller holders and delegators bear governance externalities; the broader market bears liquidity concentration risk when large operators rebalance or change payout policies.
    Measurable indicators: top-baker concentration, distribution of staking balances, and shifts in the staked-versus-delegated composition after upgrades that change incentives.

If you are designing incentives on top of Tezos, treat policy surfaces as first-class inputs. For work that needs external validation, a short consulting engagement is often enough to map emission drivers to liquidity windows and define what must be monitored on-chain versus assumed. Keep the model focused on staked ratio, issuance bounds, and the lockup profile, because those are the levers that actually move float.



This article is part of our Tokenomics Deep Dive series.