Inflation and Staking Rewards

The previous page followed SOL out of user pockets: rent locks lamports inside accounts to pay for the state they occupy, and transaction fees charge for the compute and blockspace a transaction consumes. Both are transfers — SOL that already exists changing hands or getting parked. This page is about the other direction: where new SOL comes from, and why the protocol prints it at all.

The answer ties straight back to the book’s throughline. A single global state machine running at hardware speed needs thousands of expensive validators to stay honest and online. Somebody has to pay them. Fees alone are far too small and far too spiky to do it. So Solana does what Bitcoin does with its block subsidy — it mints new coins on a schedule and hands them to the people securing the network. That minting is inflation, the coins are staking rewards, and understanding the schedule is understanding how Solana funds its own security.

Two different sources of validator income

Before the schedule, pin down a distinction that trips people up constantly. A validator earns SOL from two entirely separate places, and they behave nothing alike:

   INFLATION (issuance)          |   FEES (usage)
   ------------------------------|------------------------------
   minted by the PROTOCOL        |   paid by USERS
   on a fixed schedule           |   on demand, per transaction
   predictable, network-wide     |   spiky, follows congestion
   funds the SECURITY BUDGET     |   prices scarce blockspace/compute
   paid to STAKERS (via stake)   |   paid to the block's LEADER

Inflation is the protocol saying “here is freshly created SOL, distributed to everyone who has locked up stake to secure the chain.” Fees are users saying “here is my existing SOL to buy priority for this transaction.” As of the mid-2020s, inflation dwarfs fees as a source of validator income — which is exactly why the issuance schedule, not the fee market, is the load-bearing security mechanism today. Hold that thought; it is the same “who funds security” question Bitcoin faces as its subsidy halves toward zero.

The inflation schedule — high, then disinflationary, then a floor

Solana’s issuance is not a flat percentage forever. It is disinflationary: it starts high, decreases every year by a fixed fraction, and asymptotes toward a small long-run rate. Three numbers define the whole curve:

Initial rate — issuance began around 8% per year when inflation was switched on.
Disinflation rate — that annual rate drops by about 15% each year (a relative cut, not 15 percentage points — 8% becomes ~6.8%, then ~5.8%, and so on).
Terminal (long-run) rate — the decline stops at a floor near 1.5% per year, and issuance stays there indefinitely.

   annual inflation rate over time

   8% ┤●
      │ ╲
   6% ┤  ╲●
      │    ╲●
   4% ┤      ╲●●
      │         ╲●●●
   2% ┤             ╲●●●●●●
1.5% ┤                    ●━━━━━━━━━━━━━━━  ← terminal floor, forever
      └────┬────┬────┬────┬────┬────┬────┬──
          yr1  yr3  yr5  yr7  yr9  ...      (each year ≈ prior × 0.85)

The shape matters more than the exact figures. Early on, the network prints a lot of SOL — it needs to bootstrap a large, well-distributed staking base fast, when the token is young and the security budget most needs padding. Over time it prints proportionally less, so the currency stops diluting holders aggressively. But it never reaches zero: unlike Bitcoin’s subsidy, which halves all the way to nothing, Solana keeps a permanent ~1.5% trickle of new SOL to fund security in perpetuity. That is a deliberate, different answer to the security-budget question — Solana bets on a small forever-tax rather than a hard cap.

Each year’s rate is the previous year’s rate multiplied by (1 − 0.15) = 0.85, floored at 1.5%. Starting from 8%:

   year 0:  8.00%
   year 1:  8.00 × 0.85 = 6.80%
   year 2:  6.80 × 0.85 = 5.78%
   year 3:  5.78 × 0.85 = 4.91%
   year 4:  4.91 × 0.85 = 4.17%
   ...
   → keeps shrinking by 15% relative each year
   → clamps at 1.50% once the curve descends past it, and stays there

The rate is applied to total supply, so the number of SOL minted each year is rate × total_supply. Because supply grows while the rate falls, the actual SOL minted per year rises for a while, then declines as the falling rate wins out. Treat the exact percentages as the designed schedule; the precise dates the network crossed each step and the live supply figure are things to check against a current source, not memorize.

Newly issued SOL is paid out as staking rewards

Here is the mechanism that turns “the protocol prints SOL” into “the network stays secure.” Solana secures itself with Proof of Stake: validators are ranked by how much SOL is staked behind them, and staked SOL is at risk — it backs the validator’s promise to follow the rules. The protocol needs people to lock up SOL as stake. So it makes staking the only way to receive the newly issued SOL.

   protocol mints new SOL each epoch
              │
              ▼
      distributed as staking rewards
              │
        ┌─────┴─────────────┐
        ▼                   ▼
   validators          delegators
   (their own stake)   (SOL they delegate to a validator)

If you hold SOL and do nothing, you receive none of the new issuance. If you stake it — either by running a validator or by delegating your SOL to one — you earn a share of the freshly minted rewards proportional to your stake. The incentive is deliberate: issuance is the carrot that pulls SOL into stake, and stake is what makes attacking the chain expensive. New supply and network security are the same lever.

How rewards accrue per epoch

Rewards are not continuous; they are paid per epoch. An epoch is a fixed span of slots (on the order of a couple of days) after which the protocol tallies up how each validator performed and pays out. The rough flow:

Over the epoch, each validator votes to confirm blocks (this is its actual security work). Its earned rewards scale with its total stake and how reliably it voted.
At the epoch boundary, the protocol computes the total rewards to mint (from the current inflation rate) and each validator’s slice of it.
That slice is split between the validator and its delegators.
Rewards are typically auto-compounded into the stake account — your active stake grows, so next epoch earns on a slightly larger base.

   epoch N  ─────────────────────────────►  epoch boundary  ─►  epoch N+1
   [ validators vote, produce blocks ]        [ tally + pay ]     [ larger stake ]

The validator’s commission

A delegator does not run hardware; the validator does, and hardware plus operations cost real money (the next page is entirely about those costs). So a validator takes a commission: a percentage of the inflation rewards earned by the stake delegated to it, skimmed off before the rest flows to delegators.

   suppose delegated stake earns 100 SOL of inflation rewards this epoch
   validator commission = 8%
      → validator keeps  8 SOL
      → delegators share 92 SOL (split pro-rata by how much each delegated)

Commission is the validator’s cut for doing the work; the rest rewards the capital that delegators put at stake. A validator choosing its commission is trading off revenue against attractiveness — set it too high and delegators move their stake elsewhere, because they can see everyone’s commission on-chain. This is a real, visible market: your yield as a delegator is (your share of inflation rewards) minus (the validator’s commission).

The staking ratio — why non-stakers get diluted

Now the piece that makes inflation more than a technicality. Because new SOL goes only to stakers, inflation is not a uniform tax on everyone. It is a transfer from those who don’t stake to those who do.

Let the staking ratio s be the fraction of all SOL that is staked. New issuance equals inflation_rate × total_supply, but it is divided among only the s fraction that is staked. So a staker’s real yield is boosted, and a non-staker’s holdings are diluted by the full issuance with nothing to offset it:

   inflation rate                = i        (say 5%)
   fraction of supply staked     = s        (say 65%)

   nominal yield to a staker     ≈  i / s   = 5% / 0.65  ≈  7.7%
   real dilution of a NON-staker ≈  −i      = −5%        (they get 0 of it)

Two consequences fall straight out of this:

The lower the staking ratio, the higher each staker’s yield — the same minted SOL is split among fewer people. This is a self-balancing pressure: low participation makes staking more lucrative, pulling more SOL in.
Not staking is a slow bleed. Idle SOL is diluted every epoch relative to staked SOL. There is no neutral position; holding without staking is a decision to be diluted. This is the engine that keeps a large fraction of supply locked in stake — precisely what PoS security needs.

A recurring confusion, well-documented across 2023–2024 as staking dashboards proliferated: people quote the headline staking APY (say ~7%) as if it were free money. It is not. If inflation is 5% and you are staked, a big chunk of that “yield” is just you keeping pace with dilution — the network minted SOL and you received your share instead of being diluted by it. Your yield relative to a non-staker is real; your yield relative to the total supply is much smaller. The lesson: always separate nominal staking yield from real yield net of inflation, or you will badly overestimate what staking earns you.

A second, subtler point from the same period: inflation rewards are not a validator’s only income. MEV (value extracted from ordering transactions, often routed through systems like Jito) and priority fees have grown into a meaningful slice of validator and staker returns — sometimes rivaling inflation rewards during busy periods. So “staking yield” is really inflation share + fee/tip share + MEV share, and the mix shifts with network activity. Treat any single APY number as a snapshot, and date it.

Under the hood — inflation vs. the security budget

Step back to the throughline. A chain that runs at hardware speed is expensive to secure: validators need beefy machines, and they must be paid enough that honest operation beats attacking. That total pay is the security budget. Where does it come from?

Fees price scarce resources (compute, blockspace, hot accounts). They are excellent at rationing demand — see local fee markets — but they are spiky and small: quiet blocks earn almost nothing, and total fee revenue is nowhere near enough to fund thousands of validators today.
Inflation is smooth and predictable: it mints a known amount every epoch, independent of whether the network is busy. That predictability is exactly what a security budget wants — validators can commit to expensive hardware knowing roughly what they’ll earn.

   security budget  =  inflation rewards  +  fees  +  tips/MEV
                       └── smooth, large ──┘  └─── spiky, growing ───┘

This is the same tension Bitcoin lives with. Bitcoin’s block subsidy (its version of issuance) halves every four years toward zero, betting that fees will eventually carry the whole security budget — an unproven bet that worries a lot of people. Solana takes the opposite side: it keeps a permanent ~1.5% issuance so security is never left depending on fees alone. Neither is obviously right. Both are answers to the same question this book keeps circling: who pays to keep the honest majority honest, forever?

The architect’s lens

Why does it exist? Proof-of-Stake security needs SOL locked up as stake and needs validators paid to run expensive hardware. Inflation exists to mint the reward that pulls SOL into stake and funds the security budget on a smooth, predictable schedule that fees alone cannot provide today.
What problem does it solve? The “who funds security” problem. Fees are too spiky and too small to pay thousands of validators reliably; protocol-minted issuance is steady and network-wide, so validators can commit to the hardware a global state machine at hardware speed demands.
What are the trade-offs? Inflation dilutes non-stakers and mildly dilutes everyone versus a hard cap — the price of a forever-security-budget is a forever-trickle of new supply. The disinflation curve softens this over time but never removes it.
When should I avoid it? You can’t remove issuance and keep PoS-by-reward — but the design choice to avoid is a high, non-decaying rate: permanent heavy inflation would tax holders without limit. The disinflationary taper and 1.5% floor are the guardrails against that.
What breaks if I remove it? Strip issuance and the incentive to stake collapses: the staking ratio falls, the cost of attacking the chain falls with it, and — until fees alone can carry the security budget — the network’s economic security erodes. Issuance is the load-bearing security subsidy, not a nice-to-have.

Check your understanding

State the three parameters that define Solana’s inflation schedule (initial rate, disinflation rate, terminal rate) and describe the overall shape of the curve.
What is the difference between inflation and fees as sources of validator income? Which one funds the bulk of the security budget today, and why?
Newly issued SOL is paid out as staking rewards. Who receives it, and what happens to a holder who never stakes?
Explain how per-epoch rewards are split between a validator and its delegators, and what the validator’s commission is.
Define the staking ratio and use it to explain why a lower staking ratio raises each staker’s yield — and why non-staking is a “slow bleed.”

Show answers

Initial rate ~8% per year, a disinflation rate of ~15% relative per year (each year’s rate is the prior year’s × 0.85), and a terminal rate near 1.5% where the decline stops. The curve starts high, tapers down geometrically, and flattens into a permanent ~1.5% floor that never reaches zero — Solana keeps a forever-trickle of new supply rather than a hard cap.
Inflation is new SOL minted by the protocol on a fixed schedule, paid to stakers; fees are existing SOL paid by users per transaction to the block’s leader. Inflation funds the bulk of the security budget today because it is smooth, predictable, and network-wide, whereas fees are spiky and small — validators need a dependable income to justify expensive hardware, and issuance provides it.
It is received only by those who stake — validators (on their own stake) and delegators (on SOL they delegate to a validator), split pro-rata by stake. A holder who never stakes receives none of it and is therefore diluted every epoch relative to stakers; there is no neutral position.
At each epoch boundary the protocol computes the inflation rewards each validator earned (scaled by its stake and voting reliability), then the validator takes a commission — a percentage of those rewards — and the remainder is distributed pro-rata to its delegators by how much each staked. Commission is the validator’s cut for running and operating the hardware; it’s visible on-chain, so setting it too high drives delegators away.
The staking ratio s is the fraction of total SOL that is staked. Issuance equals i × total_supply but is split among only the staked fraction, so a staker’s nominal yield is roughly i / s — a lower s means the same minted SOL is divided among fewer people, raising each staker’s yield. Non-staking is a slow bleed because idle SOL receives none of the issuance and is diluted by the full inflation rate each epoch relative to staked SOL — pressuring holders to stake.