Centralized exchanges
A trading venue can move balances instantly in its own ledger while customers depend on its custody and withdrawal systems.
THE ATLAS
Explore the mechanisms, then follow the connections between them. Start anywhere; each topic gives you a path forward.
64 topics
A trading venue can move balances instantly in its own ledger while customers depend on its custody and withdrawal systems.
A DEX settles exchanges through blockchain rules, but its market design, interfaces and administrative controls still matter.
An order book collects offers to buy and sell. Execution depends on the available quantities at each price.
A liquidity pool quotes exchanges from its reserves and a pricing rule instead of waiting for a matching order.
Your trade moves the price; the world can also change between a quote and execution. These are different effects.
LPs lend neither a fixed quantity nor a guaranteed return: they hold a claim on inventory that traders continually reshape.
Divergence loss compares a pool position with holding its original assets. It is a benchmark difference, not a separate fee.
A displayed yield can combine fees, interest and token subsidies. Trace each source before combining the numbers.
Annual rates describe a convention. Compounding, fees, changing token prices and losses determine what a position actually returns.
Consensus staking places assets at risk to support validation. Its rewards are different from interest paid by a borrower.
Delegation separates capital from some operational duties. The exact custody, voting and penalty boundaries depend on the network and service.
A transferable token represents a staking position through additional accounting and redemption rules. Transferability does not make the underlying stake instantly withdrawable.
Restaking commits already staked or represented capital to additional services, adding new obligations and potential penalties.
Lenders supply assets that borrowers can use. Their return comes primarily from borrower interest, with liquidity and loss risks along the way.
Borrowing creates an obligation in the borrowed asset. Collateral provides a recovery path if the position breaches its rules.
Collateral is a recovery asset. Its quantity, market value, liquidity and permitted use all affect how much credit it can support.
Liquidation repays debt by transferring collateral. It can begin before insolvency, and it does not guarantee complete recovery.
Leverage creates exposure larger than the equity supporting it. Small price changes can then produce much larger changes in that equity.
A pool’s rates coordinate borrowers and suppliers. They change with demand, available liquidity and governance parameters.
An oracle brings an external assertion into a blockchain. Consensus over the assertion does not make the outside fact automatically true.
The ability to choose transaction inclusion and order can create value beyond ordinary fees, with different effects on users.
A protocol’s rules include who can change the rules. Votes, software adoption and administrative keys are different sources of authority.
A bridge makes one system act on events from another. Wrapped assets often depend on locked backing and the bridge’s verification rules.
Connecting chains requires agreement about messages, identity and finality, not just compatible addresses.
A layer two executes activity away from a base chain while relying on that base layer for defined parts of security and settlement.
Optimistic rollups accept proposed state changes subject to a dispute process that can demonstrate incorrect execution.
A validity rollup proves that an offchain state transition follows specified rules before the settlement contract accepts it.
A compact commitment can identify data without supplying it. Verification and independent recovery require the necessary data to be obtainable.
A derivative creates a payoff linked to an underlying reference. Exposure, ownership and collateral are separate quantities.
Perpetual contracts have no scheduled expiry. Funding payments encourage their trading price to stay near a reference market.
Failures occur at distinct layers: keys, contracts, economic assumptions, operators and connections. A useful risk map follows the assets through each.
Turn a protocol description into a map of assets, obligations, authorities and failure paths before interpreting its headline metrics.
A hash turns data into a compact fingerprint. It helps expose changes, but it does not make the data true, secret or permanent.
A signature lets a network check that an instruction was authorized by a key, without receiving that secret key.
Keys create spending authority. Addresses identify destinations or spending conditions. They are related objects, not interchangeable names for a wallet.
A recovery phrase can recreate many wallet keys. It is powerful secret material, and its compatibility depends on the wallet’s actual recovery scheme.
A wallet connects human intent to cryptographic authorization and network state. The assets remain recorded by the network, not inside the app.
Custody asks who can authorize transfers and what claim you have if that party fails. A visible account balance does not answer either question by itself.
A threshold of approvals can distribute control. Its resilience depends on independent signers, recoverable information and the exact account implementation.
A transaction is a proposed state change. Signing, broadcasting, inclusion, execution and finality are separate stages with different failure modes.
Nodes run the rules, exchange information and reconstruct network state. Running a node is different from producing blocks or earning staking rewards.
A block packages an ordered set of transactions with commitments that let other nodes verify it. Its producer proposes a result; the network still checks it.
A blockchain joins a shared history, verifiable state transitions and a rule for choosing between competing histories. The chain of hashes is only one component.
Proof of work makes block production a costly search. In Bitcoin, nodes choose the valid chain with the most accumulated work, not simply the most blocks.
Mining combines transaction selection, a proof-of-work search and an economic business. A pool smooths payouts while introducing coordination and operator dependencies.
Proof of stake assigns consensus influence using committed assets and protocol rules. Rewards compensate participation; penalties make certain failures costly.
A validator participates in consensus under a specific protocol. The operator, signing key, withdrawal authority and beneficial owner may be different parties.
Finality describes the strength of settlement under a network’s assumptions. A transaction can be visible, included and successful before it is considered final.
“Fork” can mean a temporary competing history, a change in consensus rules or a lasting network split. Keeping those meanings separate avoids confusion.
Bitcoin combines a native asset, unspent-output accounting, proof of work and independent validation. Its historic design paper explains the starting idea; current rules add implementation detail.
An unspent transaction output is a discrete piece of spendable value with conditions attached. A Bitcoin transaction consumes whole outputs and creates new ones.
Ethereum maintains accounts and shared contract state, with ETH paying for execution and supporting consensus. Tokens built on it add their own rules.
A smart contract is code and state enforced through a network’s execution rules. It can automate an obligation without guaranteeing that the obligation is sensible.
Gas measures execution resources. A gas price converts those units into an ETH fee; it is separate from the amount of value a transaction transfers.
A token standard defines an interface other software can recognize. It does not define the token’s economic value, backing or every administrative permission.
Tokenomics describes issuance, allocation and economic incentives. A supply schedule is an accounting rule; it is not a complete explanation of demand or price.
A stablecoin targets a reference value. The target, market price, backing assets and redemption rights are four separate things.
An issuer-backed token connects onchain transfers to offchain reserves and redemption procedures. The claim is only as useful as those connections.
Collateralized issuance creates a token alongside a debt position. Extra collateral absorbs some price movement, while oracles and liquidation try to limit deficits.
An algorithm can adjust supply or exchange tokens. It cannot create outside purchasing power by rule alone when confidence and market liquidity disappear.
An NFT tracks a distinct token identifier. The token, its metadata, the displayed media and any rights granted to the holder are separate layers.
Tokenization represents an asset or right with a digital token. The representation can improve coordination while adding a boundary between ledger state and the underlying asset.
A pseudonymous address is not the same as anonymity. Ledger data, network metadata and information held by services can combine into a detailed picture.
A zero-knowledge proof can establish that a precise statement holds without revealing the secret witness used to prove it. The statement and public inputs can still be visible.
Looking for one unfamiliar word?
Open the glossary ↗