HostNexFlow › Library › Cardano vs Polkadot staking
Cardano vs Polkadot staking: no lock-up versus 28 days
Cardano lets you delegate without locking a single coin and never slashes. Polkadot locks your DOT for four weeks and can, in the worst case, take all of it. Those are not two flavours of the same thing — they are opposite answers to the question of what a delegator should be on the hook for.
The two chains are usually grouped together as "the research-heavy proof-of-stake designs", which hides how differently they treat the person delegating. Cardano's Ouroboros makes the delegator a passive party: your ADA stays in your wallet, spendable at any moment, and the pool you delegate to can only earn you less, never cost you principal. Polkadot's nominated proof of stake makes the nominator an active participant in security: you bond DOT, you pick up to sixteen validators, your stake backs them, and if one of them misbehaves the protocol takes a share of what you bonded.
Every row in the table below was read on 2026-09-27 from the chain's current protocol parameters — Cardano's via a live Koios read of epoch 658, Polkadot's from the Polkadot Wiki's published runtime constants — and names its source. Where a live value could not be verified we say "not published" rather than paste a figure from memory. We have filed for a stake-pool lane on Cardano and a validator lane on Polkadot; both are filings, not live operations, and our Cardano page and our Polkadot page say exactly that.
The parameter table
| Parameter | Cardano | Polkadot |
|---|---|---|
| Lock-up on delegated stake | None — delegated ADA stays liquid in the wallet and can be spent at any time; spending simply reduces the delegated amount (Cardano docs, "Delegation", as of 2026-09-27). | Bonded DOT is locked; unbonding takes 28 eras, and an era is 24 hours, so 28 days (Polkadot Wiki, "Staking" — BondingDuration, as of 2026-09-27). |
| Slashing of principal | None in the protocol — a bad pool earns fewer or no rewards, delegators' principal is untouched (Cardano docs, "Stake pools", as of 2026-09-27). | Yes — offences such as equivocation and unresponsiveness slash the validator and every nominator backing it, proportionally; the penalty scales with how many validators offended at once and can reach 100% in a large coordinated offence (Polkadot Wiki, "Slashing", as of 2026-09-27). |
| Minimum to delegate | No protocol minimum; registering a stake key costs a refundable 2 ADA deposit (key_deposit = 2,000,000 lovelace, Koios epoch 658 parameters, as of 2026-09-27). | Direct nomination requires bonding above a dynamic minimum active nomination threshold: current value not published here. Nomination pools accept a 1 DOT minimum bond (Polkadot Wiki, "Nomination pools", as of 2026-09-27). |
| How many operators you back | One pool per stake key (Cardano docs, "Delegation", as of 2026-09-27). | Up to 16 validators per nominator; the election algorithm decides which of them your stake actually backs each era (Polkadot Wiki, "Nominator", as of 2026-09-27). |
| Operator fee model | Fixed cost per epoch of at least 170 ADA taken first (min_pool_cost = 170,000,000 lovelace), then a variable margin the pool sets (Koios epoch 658 parameters, as of 2026-09-27). | Validator sets a commission percentage on era rewards, 0–100%, no protocol floor (Polkadot Wiki, "Validator", as of 2026-09-27). |
| Anti-concentration | Saturation at 1 / k of total stake with k = 500 (optimal_pool_count); rewards to a saturated pool stop growing. Pledge influence a0 = 0.3 (Koios epoch 658 parameters, as of 2026-09-27). | Phragmén election spreads nominations to maximise the minimum backing across the active set; per-validator payouts are equal regardless of stake, so over-backed validators pay each nominator less (Polkadot Wiki, "NPoS election", as of 2026-09-27). |
| Issuance and treasury | Monetary expansion 0.3% of the remaining reserve per epoch (monetary_expand_rate = 0.003), with 20% of each epoch's rewards to the treasury (treasury_growth_rate = 0.2) (Koios epoch 658 parameters, as of 2026-09-27). | Issuance follows the runtime's inflation curve, which targets an ideal staking rate; current parameters: not published here (Polkadot Wiki, "Inflation", as of 2026-09-27). |
| Reward timing | Epochs are 5 days; a new delegation earns from the epoch after it takes effect and that reward is paid two epochs later — first rewards land roughly 15–20 days in (Cardano docs, "Delegation cycle", as of 2026-09-27). | Eras are 24 hours; rewards are payable per era once someone triggers the payout, and unclaimed eras expire after a history depth of 84 eras (Polkadot Wiki, "Staking rewards", as of 2026-09-27). |
| Pool operator deposit | 500 ADA refundable registration deposit (pool_deposit = 500,000,000 lovelace, Koios epoch 658 parameters, as of 2026-09-27). | No fixed deposit; validators must self-bond and attract enough nominations to be elected (Polkadot Wiki, "Validator", as of 2026-09-27). |
| Our footprint | Stake-pool lane filed, testnet queue (/networks/cardano/, as of 2026-09-27). | Validator lane filed (/networks/polkadot/, as of 2026-09-27). |
Liquidity: the biggest difference on the page
Cardano's delegation is a pointer, not a transfer. You register a stake key, point it at a pool, and the ADA in the associated addresses counts toward that pool's stake — while remaining exactly as spendable as it was before. Sell half tomorrow and the pool's stake falls by half at the next snapshot. There is nothing to unbond because nothing was bonded. This is why the phrase "lock-up" never appears in Cardano's documentation for delegators; the protocol's security comes from stake weighting at snapshot time, not from holding your tokens hostage.
Polkadot bonds. The DOT you nominate with is locked in the staking system, and when you want it back you call unbond and wait 28 eras — 28 days — before you can withdraw it. During that window the tokens earn nothing and remain slashable for offences committed while they were backing a validator. Nomination pools inherit the same 28-day unbond. Liquid staking derivatives exist on both chains and change this picture, but they change it by adding a contract and an issuer between you and the protocol, which is a different risk, not a shorter one.
Slashing: none versus proportional
Cardano does not slash. The worst a pool can do to its delegators is under-perform — miss blocks, or set a margin so high nothing is left — and even then the delegator loses only potential rewards. The design bet is that pool operators are disciplined by delegators leaving, which on a chain with no lock-up is instant and cheap. Critics call this weak; defenders point out that the delegator was never asked to underwrite the operator in the first place.
Polkadot's nominator is explicitly underwriting. When a validator equivocates or goes unresponsive, the protocol slashes the validator's bond and every nominator's bond behind it, in proportion. The slash percentage depends on how many validators committed the same offence in the same era — a single isolated equivocation is a small fraction, while a large correlated offence can be slashed to 100%. There is a governance path to cancel slashes judged to be caused by bugs rather than malice, but that is a safety valve, not a promise. The practical consequence is that choosing sixteen validators on Polkadot is an act of due diligence with money attached, whereas choosing a pool on Cardano is closer to picking a savings rate.
Rewards are not a headline number on either chain. Cardano's per-epoch rewards depend on the pool's performance, its fixed cost and margin, its pledge and how close it is to saturation. Polkadot's depend on which of your sixteen validators were elected, their commission, and how much other stake was backing them. Neither reduces to a percentage that would be honest to print, so we do not.
Pool economics and how the fees actually work
Cardano's fee model is two-part and visible on-chain. A pool takes a fixed cost each epoch — the protocol floor is 170 ADA — before anyone else is paid, then a variable margin on what remains, and only then are rewards split among delegators pro rata. The fixed cost matters most for small pools: 170 ADA off the top of a small pool's rewards is a larger fraction than the same 170 ADA off a large one. Saturation works the other way: with k = 500 a pool holding more than 1/500 of total stake stops earning more rewards, so delegating to an oversaturated pool dilutes everyone in it. The pledge influence parameter a0 = 0.3 tilts rewards toward pools whose operators have put up more of their own ADA.
Polkadot pays each active validator roughly the same era reward regardless of how much stake backs it — the reward is for the work, not the stake. The validator takes its commission, and the rest is divided among its nominators in proportion to backing. That is why a validator with a huge backing pays each nominator less per DOT than a lightly backed one, and why the Phragmén election exists: it distributes nominations to even out backing across the active set. As a nominator you name sixteen candidates and the algorithm decides, era by era, where your stake actually lands.
Who should pick Cardano, who should pick Polkadot
Pick Cardano staking if you want to keep every coin spendable, cannot tolerate any principal risk from an operator's mistake, and are content with rewards that arrive on a five-day epoch cadence after an initial two- to three-week delay. The trade is that your security contribution is passive and your operator is disciplined only by exit — and that the two-part fee structure rewards you for reading a pool's cost and saturation before you delegate.
Pick Polkadot staking if you accept that nominating is underwriting: you are prepared to research up to sixteen validators, live with a 28-day unbond, and carry proportional slash exposure in exchange for a direct stake in the chain's security. If that research burden is too much, a nomination pool lowers the entry to 1 DOT but does not remove the lock-up or the slash exposure.
Whichever you choose, the operator matters more than the chain. Neither of our lanes on these two networks is live yet — Cardano is in the testnet queue and Polkadot is a filing — and we would rather say so than sell a pool we do not run. When they go live the state will be readable on the validator desk alongside the chains we already operate. Until then, read the pool's parameters on-chain, and if the staking you are offered arrives as a token rather than a native delegation, run it through the token checker first.
Frequently asked
Is ADA locked when you stake on Cardano?
No. Delegation on Cardano is a pointer from your stake key to a pool; the ADA stays in your wallet and can be spent at any time, as of 2026-09-27. Spending reduces the delegated amount at the next snapshot. There is no unbonding period because nothing is bonded.
How long does it take to unbond DOT on Polkadot?
28 eras, and a Polkadot era is 24 hours, so 28 days as of 2026-09-27. Bonded DOT earns nothing during unbonding and remains slashable for offences committed while it was backing a validator. Nomination pools have the same 28-day unbond.
Can you lose your stake on Cardano or Polkadot?
Cardano: not through slashing — the protocol has none as of 2026-09-27; a bad pool only earns you less. Polkadot: yes — offences slash the validator and its nominators proportionally, and a large correlated offence can be slashed up to 100%, as documented in the Polkadot Wiki's slashing section.
What fees do Cardano pools and Polkadot validators charge?
Cardano pools take a fixed cost per epoch — protocol minimum 170 ADA as of epoch 658, 2026-09-27 — then a variable margin they set. Polkadot validators set a commission percentage on era rewards with no protocol floor. Both are visible on-chain before you delegate.
How much do you need to start staking on each chain?
Cardano has no protocol minimum; registering a stake key takes a refundable 2 ADA deposit as of 2026-09-27. Polkadot direct nomination requires bonding above a dynamic minimum active nomination threshold; nomination pools accept 1 DOT.
Does NexFlow run a Cardano pool or a Polkadot validator?
Not yet. As of 2026-09-27 we have filed for a Cardano stake-pool lane (testnet queue) and a Polkadot validator lane; both are filings, not live operations, and our network pages say so. Live state for the chains we do run is on the validator desk.