Did you know? A blockchain wallet can show you a token balance without proving that you can transfer, sell or withdraw that amount immediately. The number on the screen is a report about a particular kind of recorded or calculated balance. Spendability is another question. It depends on the correct network, control over the account, token rules, transaction fees and sometimes a platform’s own policies.
Imagine Alicia opening a fictional wallet that displays 50 science-club credits. She tries to send ten to Tricia, but the action fails. Kai Kai says the app must be broken. Not necessarily. The app may be displaying a token that needs separate network-fee funds, a time-locked position, a contract balance that cannot be directly transferred, or information from the wrong network. Even a true balance need not be available in the way Alicia expects.
This guide explains how blockchain technology works through the difference between what an interface displays and what the owner can actually use. We will compare native assets with tokens, look at fee requirements, account control, token restrictions, staking, bridges, custodial balances and Bitcoin’s transaction-output model. Everything can be understood with fictional examples. You do not need a live wallet or real money.
Quick answer: A balance is a measured quantity associated with a defined account, asset and network. Spendable balance means the account has the necessary authorisation, unlocked or transferable value, transaction resources and applicable contractual permissions to perform a specific action. Those values can differ even when the wallet’s displayed number is accurate.
This is a standalone reader’s guide. For the foundation, read What Is Blockchain?. For wallet credentials, use Why a Wallet Password Is Not a Recovery Phrase. Here our focus is one practical question: why can a visible balance be unusable right now?
The simplest distinction: recorded, available and spendable
An amount can be recorded without being immediately available for a new action. Think of a school library account that shows twelve books borrowed over a year. The account history is real, but it does not mean the student has twelve books available to borrow again today. Similarly, a club budget can show 100 approved credits, with 60 already committed to upcoming activities.
The analogy is imperfect—blockchain balances are not library records—but it teaches a useful three-part check. Recorded describes what data or state says. Available concerns how much remains uncommitted under the relevant rules. Spendable concerns whether a specified authorised operation can actually use it. Some wallets label these categories separately; others place one broad figure on the home screen.
On an ordinary Ethereum externally owned account, ETH balance and ERC-20 token balance are not the same field. The native balance is part of the account state. A token’s reported balance comes from a particular token contract. A wallet might display both side by side, but the mechanisms controlling each can differ.
For students, write a complete balance sentence: “On network N, account A has X units of asset T under the stated block and contract rules.” That sentence is longer than “I have X,” but it makes missing context visible. It invites the next question: “Which operation do you want to make with this particular balance?”
Never infer that a website’s total portfolio valuation is a directly withdrawable amount. It may combine several assets, market price estimates, restricted positions and networks. A tidy currency total is a calculation by an interface, not a universal on-chain cash balance.
A token transfer can need a different asset for fees
One of the most common surprises is an Ethereum-compatible wallet displaying many units of an ERC-20 token while having little or no native asset to pay the fee for a direct on-chain transaction. On Ethereum mainnet, transaction execution ordinarily consumes gas paid in ETH. Having 100 units of a token does not by itself mean the account can pay the ETH fee required for sending that token.
Alicia’s fictional science-club token balance says 50 ART. She wants to send ten. The token contract may permit the transfer, and Alicia may control the signing key, yet the requested transaction can still fail to be submitted or accepted if the account cannot meet its fee requirements. The token amount and the processing fee belong to different categories.
A wallet may show a fee estimate before signing. That estimate can change with network conditions and the operation being performed. Some applications sponsor transactions or use specialised account abstractions, so “every token move always needs the user to hold ETH personally” would be too broad. The correct statement is that someone must cover the relevant network resource cost under the particular mechanism.
This is why a helpful interface should show the intended recipient, transferred asset, network, fee-paying asset and quoted fee as separate fields. The wallet’s headline number cannot answer the fee question on its own.
Read Why Network Fees Change When It Gets Busy for the fee calculation and Ethereum Gas and Fees for the network-specific technical model. Neither reference should be turned into universal rules for Bitcoin or every layer-two system.
The wrong network can make the right number unusable
A user can see the same-looking account address across multiple Ethereum-compatible networks. The asset and balance on one network are not automatically available on another. Mainnet, a layer-two system and a public testnet can keep separate histories even where the wallet displays them within one app.
Suppose Alicia sees 20 fictional credits on network Blue but tries to use an application connected to network Green. The address may look identical. The app might show zero relevant credits because it is querying a different ledger and contract. Switching the selected network can help the user understand the discrepancy, but changing a display selector does not move assets between networks.
Bridging is a separate process involving one or more applications and on-chain events. Sending an asset on a source chain and observing a corresponding representation on a destination chain are not simply two views of one transaction. Each part has its own failure conditions, fees and authorisation model.
A helpful investigation names the exact network and chain identifier where applicable, then identifies the correct token contract. A ticker symbol alone is insufficient because different contracts and chains can reuse names and icons.
The safer conclusion is “the available token is on a different network from the one this action expects,” not “the wallet has lost the tokens.” Some assets are also unsupported by particular applications even though a network observer can see them. Visibility in one wallet and usability in one app are distinct claims.
For comparing displays, see Why Two Block Explorers May Disagree. The explorer may be correctly reporting the asset on a different network from the one a transaction form is using.
What an ERC-20 balance really measures
The ERC-20 interface includes a method named balanceOf that can report a token amount for an account under a particular contract’s implementation. That value is usually presented in human-readable units using the token’s decimal convention, which is a display matter rather than a new source of economic value.
The token contract can have additional rules beyond a bare arithmetic balance. It might limit transfers, impose allowlists, require settlement conditions, pause certain operations or implement fees and other mechanisms. A visible balanceOf result does not automatically prove that every standard transfer will succeed.
Consider a fictional museum ticket token that reports one ticket but allows movement only before the event date. After that date, an interface might still show one token in the account while the relevant contract forbids transfer. Whether the restriction exists depends on the actual implementation, not on the word “token.”
Some assets represent claims on a custodian or underlying property. Seeing units recorded in a token contract does not itself guarantee redemption for cash, delivery of goods or legal recognition of a property right. Those questions depend on the issuer and its rules as well as on-chain state.
It is useful to separate balance, transfer capability, redemption right and market value. These four can change independently. A product can have a positive on-chain balance, be restricted from transfer, lack an active redemption channel and have an uncertain market price.
The ERC-20 standard describes common methods such as balanceOf, transfer and allowance. Understanding what these interfaces do—and what they do not promise—is more useful than assuming a familiar token icon implies cash-like behaviour.
A token can be frozen or a contract can be paused
Some token contracts include features that allow authorised operators to prevent transfers by certain addresses, pause a defined function or restrict movements during specified conditions. Such capabilities may support compliance, emergency response or business processes, but they also make token control different from a simple unrestricted balance.
Alicia’s fictional festival issues 30 access credits. After discovering a duplicate-issuance problem, the event administrator pauses the transfer function under the program’s approved rules. The credits still appear in the account register. Their current transferability has changed. This is a hypothetical contract design, not a statement that every blockchain asset has an administrator.
Students should ask who has authority to impose the restriction, whether the restriction is disclosed, how the decision can be reviewed and which functions are affected. A pause might target transfers but not all possible contract interactions. An allowlist might apply to destinations, senders or specific assets.
The broader lesson is that blockchain applications can include governance and operator powers. Recording a token on a shared network does not guarantee its issuer has no administrative control. The word “decentralised” is not a substitute for reading the contract’s actual permissions.
An explorer might show a successful past transfer even though transfers are now paused. The history remains true, while present authorisation differs. An interface that relies on old information can obscure this change.
For a rigorous system, current contract state, permission roles, upgrade mechanisms and legal or policy terms need to be read together. Readers should avoid assuming that a past balance or past ability to transfer proves the same ability exists forever.
Locked, staked and vested positions are different from free balances
Assets can be placed into systems that restrict immediate transfer. A staking position, time-locked vault, vesting schedule or escrow-like arrangement may record an economic interest while limiting when and how a user can withdraw value. The exact mechanism varies by protocol.
Picture the classroom art club allocating 40 credits to a project that finishes next term. Its dashboard includes the total committed amount in a portfolio summary. Alicia cannot use all 40 to buy supplies today if the governing rules keep them locked until the project condition is satisfied. The amount is not imaginary; its current usability is constrained.
On public chains, different services may represent locked positions through smart-contract balances, derivative tokens, specialised withdrawal queues or internal service records. A single wallet view might summarise them as one figure even though a transfer of the displayed amount cannot be made with a simple token-transfer instruction.
The term staking is especially broad. Ethereum validator deposits and withdrawals have particular network rules; other applications call their own token-locking programmes “staking.” You should not assume that every such programme has identical risks, yield mechanisms or withdrawal conditions.
A careful explanation names the precise instrument and the action required to withdraw. Is there a delay? An unlock date? An administrator? A redemption step? A contract condition? A custodial provider? Does a withdrawal create a separate accepted blockchain transaction?
This distinction is also why a portfolio app’s estimated value may exceed the amount that is immediately liquid or withdrawable. The interface is adding positions together for convenience, not promising instant settlement of every underlying obligation.
Bitcoin: an amount can depend on which outputs are spendable
Bitcoin’s account model differs from Ethereum’s. Rather than a single ERC-20-style contract balance, ordinary Bitcoin spending uses eligible unspent transaction outputs, commonly called UTXOs. Wallet software assembles transactions that consume selected eligible outputs and create new outputs, including possible change.
A balance screen often aggregates relevant outputs for the addresses or keys the wallet tracks. That is useful, but an aggregate may conceal whether each output is spendable under the current consensus rules and the wallet’s signing capabilities. Some outputs may be time-locked, belong to a script with additional conditions or need particular signatures.
Imagine three fictional vouchers of 5, 10 and 15 points. The bookshop reports a total of 30. But one voucher cannot be redeemed until next week, and another requires two signatures from committee members. The total number is real; the immediately spendable amount under the current conditions can be smaller.
For actual Bitcoin, a wallet’s availability calculation also depends on transaction construction, selected inputs and the fee needed to include the transaction. A wallet could have enough total value for a payment but not enough after the selected fee or policy constraints are considered.
A recipient may also care about whether incoming transaction outputs are confirmed rather than merely observed in a local pending pool. Depending on the wallet’s risk policy and transaction structure, unconfirmed outputs may be displayed separately or treated differently.
For the technical model, read the Bitcoin Developer Guide: Transactions and eduKateSG’s UTXO Transactions and Change Outputs. Bitcoin’s eligibility rules should not be replaced by an Ethereum gas calculation.
A balance is not proof that the wallet holds the signing key
Public blockchain data can be viewed by somebody who does not control the relevant address. A website or wallet interface may be configured in watch-only mode, monitoring a public address and displaying activity without having the authority to sign transactions.
This can be useful. A school committee might monitor an official donation address without giving every student the ability to spend donations. An accountant can observe incoming transfers while the actual transaction authority remains with approved custodians. But a person who sees a balance on such a screen cannot automatically send the assets.
A public address is not a recovery key, and a screenshot of that address is not evidence of ownership. Spendability in a conventional key-controlled account depends on possession and appropriate use of the relevant signing authority. Programmable accounts may instead require contract-defined signing factors, guardians or other authorised actions.
Suppose Kai Kai imports a public address into a portfolio tracker. The tracker correctly shows 100 fictional credits. It still cannot spend those credits merely because Kai Kai added the address. The dashboard is performing observation, not providing control.
This is why the phrase “the wallet has the money” can obscure the real question. Which network records the state, and who is allowed to authorise an operation? A custodial platform may add another layer because the service, not the user, controls the underlying keys.
See Why a Wallet Password Is Not a Recovery Phrase for how keys, app access and recoverability differ. A visible account can remain visible even after the ability to sign with that account has been lost.
Why a pending transaction can make the usable amount look smaller
Wallet software sometimes treats a pending outgoing transaction as a commitment. The chain may not yet have included it, but the wallet estimates that a portion of the balance will be used if the request succeeds. It may therefore show “available” as less than the raw confirmed balance.
Alicia has 50 fictional credits and submits a transaction to send 20. Before inclusion, a careful application might show 50 as the recorded balance and 30 as provisionally available, along with the relevant fee allowance. Another app might keep showing 50 until the transaction is confirmed. Both interfaces need clear labels.
Pending transactions are not one global waiting room. Different nodes can observe different pending requests, and some transactions are replaced, dropped or become invalid. Treating a pending entry as final without checking the network’s accepted history can lead to mistakes.
The wallet might also prevent the owner from submitting a conflicting request based on its own local policy. That user-interface restriction can be helpful, but it is not identical to a universal network rule preventing all alternative signed transactions.
When investigating a mismatch, distinguish confirmed state, pending intent, wallet-reserved amount and actual accepted outcome. Write down the transaction identifier and network only if needed; do not share recovery phrases or sign unrelated actions to test the discrepancy.
For the wider verification sequence, use Why a Transaction Hash Is Not a Payment Receipt. Inclusion, successful execution and finality are distinct stages that affect what a reader should call spendable.
The token symbol and picture might be misleading
Token creators can use names and symbols that resemble established assets. A wallet interface might display a familiar logo even where the underlying contract is different. That can make a balance appear more valuable or more usable than it really is.
Imagine receiving 10,000 fictional “SCHOOL” tokens from a stranger. The large number looks impressive, but the token contract might be unrelated to the school’s authorised credits. There may be no legitimate redemption route or reliable market price. A malicious token can also be designed to lure readers to a deceptive website.
The correct identity of an ERC-20-style asset includes the chain and full contract address, not just its displayed name. Decimals influence how raw units are presented. A wallet’s currency valuation usually depends on an external price source and assumptions about whether the asset can actually be exchanged.
Some portfolio interfaces deliberately hide spam-like tokens; others show all newly detected assets. Neither choice changes the underlying network record. It changes the user’s interpretation of the record. A high displayed balance is not automatically a valuable, transferable asset.
The safe classroom habit is not to test unknown token contracts with real funds. Instead, compare fictional examples: one authorised token, one token with the same symbol but another contract address, and one deliberately unspendable classroom label.
Students can then state the difference precisely: “These two assets have similar presentation but distinct contract identities and rights.” That is more informative than saying the wallet’s balance must be fake because it cannot be sold.
A custodian’s account balance is not the same as a blockchain balance
People may hold digital assets through a custodial platform. The platform can manage underlying keys and maintain its own accounting records showing what it owes each customer. The customer’s login credentials grant access to that service, not necessarily direct authority over a distinct public blockchain address.
An exchange app might show 200 units in a customer account while withdrawals are temporarily unavailable because of maintenance, compliance checks, network support, solvency concerns or a service dispute. Such restrictions are service-level conditions. They should not be described as an inherent property of the blockchain protocol.
The reverse can also occur: a public blockchain address controlled by a custodian may show a large pooled balance that does not map directly to one customer’s personal entitlement. The platform’s own ledger determines the customer’s claimed portion, subject to its agreements and actual financial condition.
This is why “proof of reserves,” “account balance,” “withdrawable balance” and “confirmed on-chain transaction” answer different questions. A visible balance does not by itself establish that a platform will meet every redemption or withdrawal request.
A careful user asks who holds the signing keys, which entity owes the asset or money, where the balance is recorded and what conditions control withdrawals. The response may involve contract law, consumer protections and custodial arrangements beyond the blockchain.
For students, draw two ledgers: a public network balance held by a custodian and a private list of customer entitlements maintained by that custodian. Matching the two is an accounting question, not something one screenshot can guarantee.
A successful smart-contract call can still leave the asset unavailable
Smart contracts can perform many different operations. A user might submit a successful transaction to deposit tokens into a vault, register a withdrawal request or receive a receipt representing a claim. A green “success” label confirms the execution outcome under the network rules; it does not guarantee the user now holds freely transferable tokens.
Suppose Tricia puts 25 fictional ART credits into a club vault. The vault records that she has a claim to 25 after the exhibition ends. The deposit transaction can succeed. Yet Tricia’s original wallet balance may fall while her vault position rises. A portfolio app could sum them into one combined total, but she cannot simply spend the original credits before the vault’s conditions allow withdrawal.
Some applications issue a receipt token for deposited assets. The receipt’s transferability and redemption rights are separate questions governed by its contract and underlying platform. A receipt token’s market value or redeemability is not automatically equivalent to instant access to the original asset.
A reader should identify what the transaction actually did. Did it transfer an asset, lock it, mint a representation, authorise a spender or merely emit an event? Contract state and execution logs may be necessary to interpret the change.
The distinction is the same as the first lesson: what appears as a balance may represent a claim or position rather than spendable native funds. Good interfaces say which is which.
A paper exercise with four balances
Draw a classroom ledger for Alicia. It has four categories: 40 ART credits freely transferable, 20 ART credits locked until next term, 10 ART credits pending receipt from Kai Kai and five credits whose public address can be observed but not signed for by Alicia.
A poorly designed dashboard might total them as 75. But the amount Alice can freely transfer in a particular action depends on which account she actually controls, which credits are unlocked, which pending payment has settled, and which fees the action requires. If the wallet shows 75, it should label what the total represents.
Now add a network dimension. Half the credits are on a different fictional chain. They are not automatically usable in a contract expecting the first network. Then add an approval condition: the club’s token contract restricts transfers until its administrator completes a registration check.
Ask classmates to recalculate what is definitely, provisionally or not currently spendable under each new condition. There may not be one universal answer because the requested operation matters. The exercise deliberately keeps the network rules fictional and makes every assumption visible.
Next ask whether an accurate balance can still be displayed without control of the signing key. Yes: an observer can read public data. Ask whether an app can display a false total because a token uses a familiar symbol. It can mislabel information without changing the actual ledger.
Finally ask students to explain why an estimated national-currency value need not be withdrawable cash. Price, liquidity, fees, custody and restrictions are separate dimensions. No real cryptocurrency is needed to learn this lesson.
A six-check approach to a wallet that cannot send
First, identify the network and the exact asset. Make sure the token contract and chain match the intended application. Similar addresses and identical symbols across networks can be misleading.
Second, identify what the displayed amount represents. Is it a confirmed native balance, a token-contract balance, a custodial entitlement, a staked position, a locked claim, a portfolio estimate or a watch-only observation?
Third, confirm the relevant authority. Does the wallet actually have the signing capability required by the account or the smart-account policy? A public address in a tracker does not grant spending control.
Fourth, check protocol and contract conditions. Does the transfer require native-asset fees? Is the token paused, frozen, time-locked or otherwise restricted? Is the current operation permitted under its contract?
Fifth, distinguish pending activity from accepted state. Some wallet interfaces reserve amounts for pending transactions. Use the correct network and accepted status when deciding whether an amount has been spent.
Sixth, separate withdrawal from economic value. A platform may impose withdrawal rules; a token may lack a credible redemption route; an estimated price may not be realisable. “Worth 100” does not equal “100 immediately transferable.”
If a check is inconclusive, describe that explicitly. For example: “The wallet shows a token balance on network A, but the account cannot currently submit the desired transfer because the required fee asset is unavailable.” That is a diagnosis, not speculation about lost money.
Five misunderstandings worth correcting
“The coins are in my phone.” On ordinary public-chain accounts, network state records balances while wallet software manages access and signing authority. Custodial services operate differently.
“If the displayed amount is real, I can withdraw it now.” Not necessarily. Contract restrictions, custody, time locks, fees and account authority can prevent a particular action.
“My wallet has tokens, so fees can come from any token.” Ordinary Ethereum transactions charge gas in the network’s native asset unless an appropriate sponsorship or special mechanism applies.
“A portfolio value is the same as cash.” Token prices can be estimates, and restricted or illiquid assets may not be saleable for the displayed value.
“The same address means the same balance on every chain.” Distinct networks have distinct states. A familiar address string and matching token symbol do not merge those states.
When a wallet screen causes confusion, use more specific nouns: native asset, contract token, locked position, pending amount, available fee balance, account authority and custodian. Most puzzles become easier once the object is named correctly.
Frequently asked questions
Why can I see tokens but not send them?
Possible reasons include inadequate native-asset balance for fees, the wrong network, a token restriction, a locked position, missing signing authority or a stale interface. The exact diagnosis depends on the asset and action.
Can I send an ERC-20 token with no ETH?
On ordinary Ethereum mainnet transactions, the account generally needs ETH for gas unless an application or account mechanism sponsors the cost. Token ownership does not itself pay the network fee.
Why is my portfolio value higher than what I can withdraw?
A portfolio may add together liquid, locked, staked or externally priced positions. The displayed total can include market estimates that differ from immediate transferability or redemption.
Does seeing an address prove that I own its assets?
No. Public addresses can be observed by anybody, and a watch-only interface can show balances without possessing signing authority. Ownership and control can also involve external legal or custody arrangements.
Does a successful transaction mean my tokens are spendable?
Not always. A successful deposit could place them into a vault or create another claim. Read what the contract actually did rather than relying on the transaction’s success label alone.
Are Bitcoin balances spendable in exactly the same way as Ethereum balances?
No. Bitcoin spending relies on eligible UTXOs and transaction constraints, while Ethereum’s native and token-account models differ. Use each network’s own concepts.
Can a token contract freeze my balance?
Some tokens implement administrator powers, transfer restrictions or pauses. Others do not. Read the actual contract’s current rules and governance instead of assuming every token is unrestricted.
Can pending transactions reduce my available balance?
Wallet software may reserve amounts for pending operations to avoid conflicting requests. Pending is not the same as accepted and final. Confirm the outcome under the network’s rules.
Is a token with a large displayed amount necessarily valuable?
No. An untrusted token can copy a familiar name or have little liquidity or redeemability. A portfolio estimate is not proof that somebody will buy the amount shown.
What if a custodial platform shows a balance but will not let me withdraw?
The restriction could be imposed by the platform’s policies or operations rather than the blockchain protocol. Identify the platform’s custody arrangement, obligations and official support route.
Should I share my recovery phrase to fix a wallet balance?
No. Ordinary public balance checks do not require your recovery phrase or private key. Treat anyone demanding those secrets for a routine display or transfer problem as a serious warning.
What is the most important distinction?
A displayed balance answers how much a system reports. Spendability asks what a specified authorised action can actually move under the current network, contract and service conditions.
Primary sources and related articles
For the Ethereum-specific mechanisms, read Ethereum Accounts, Ethereum Gas and Fees and EIP-20: Token Standard. For Bitcoin’s input and output rules, see the Bitcoin Developer Guide: Transactions. These references use their own network terminology and should not be treated as interchangeable.
Within eduKateSG, read Native Assets and Token Standards: Read the Interface Before the Balance, UTXO Transactions and Change Outputs and Why a Wallet Password Is Not a Recovery Phrase. These study the technical balance model, Bitcoin accounting and control of signing authority respectively.
For connected reader questions, continue to Signing a Message Is Not the Same as Sending a Transaction and Why a Blockchain Address Is Not a Verified Identity. Both explain why a familiar screen or string can tell only part of the story.
The key habit is wonderfully simple: ask what the balance represents, who can control it and which rule is preventing the proposed action. That is how a confusing wallet screen becomes a specific and answerable technology question.