MARKET
Start from zero · 12 min read

What is a blockchain — and which parts actually cost a trader money?

Quick answer. A blockchain is a list of transactions kept in identical copies by thousands of independent computers, each block carrying a fingerprint of the one before it. For a trader only three parts matter: confirmations create the deposit delay — roughly 0.30% of price uncertainty on a 30-minute wait, about $15 on $5,000 — separate chains create the wrong-network mistake, and your exchange balance never touches a chain at all.

Almost every blockchain explainer is written for someone who wants to build on one. If you are here to trade, you need a much smaller and much more practical subset: what the chain is doing while your deposit sits there, why it sometimes costs $2 and sometimes $30, why the same token on the wrong network disappears, and the fact that surprises most people — that when you buy and sell on an exchange, no blockchain is involved in the trade at all.

Four pages of records in a row, each linked to the next by a chain ring, the newest one highlighted in gold
Each page grips the page behind it. Break one link and everything after it stops fitting.

KEY TAKEAWAYS

  • A block is a batch of transactions plus a fingerprint of the previous block. That single link is the whole idea.
  • The newest block can still be replaced, which is the entire reason confirmations — and deposit delays — exist.
  • A 30-minute deposit wait carries roughly 0.30% of price dispersion at 40% annualised volatility — about $15 on $5,000, and double that once in twenty times.
  • That cost scales with volatility, so it is biggest at exactly the moments you feel the urge to hurry.
  • Trading on a centralised exchange writes nothing to any chain. Your balance there is a database row and an IOU.

What is a blockchain, in one paragraph?

A blockchain is a shared list of transactions that thousands of independent computers keep identical copies of. New transactions are collected into a block — think of a page in a ledger — and every block includes a short fingerprint, called a hash, of the block immediately before it. That is the entire trick. Because each page carries a fingerprint of the previous page, editing an old page changes its fingerprint, which invalidates the next page, and the next, all the way to the present. You cannot quietly alter history; you can only be caught.

The word “chain” is doing real work here. It is not a metaphor for “connected”. It describes a literal cryptographic link, one block reaching backwards to grip the one behind it.

What a blockchain is not is a place where anything is stored. Your coins are not files sitting on a chain. What exists is a history of transfers, and the current balance of any address is simply the result of adding that history up. There is nothing else there.

How blocks grip the block behind themFour blocks in a row, each one a rounded box containing a list of transactions, a fingerprint of the previous block, and its own fingerprint. An arrow runs from each block's own fingerprint to the previous-block slot of the next block. Beneath, a note explains that editing block two would change its fingerprint, so block three's previous-block slot would no longer match and every block after it would be rejected.EACH BLOCK CARRIES A FINGERPRINT OF THE ONE BEFORE ITBlock 1prev: (none)A pays B 0.4C pays D 1.2own: 7f2aBlock 2prev: 7f2aB pays E 0.1D pays A 0.9own: c41bBlock 3prev: c41bE pays C 0.1A pays F 2.0own: 9ed0Block 4 (newest)prev: 9ed0F pays B 0.5your depositown: 1a8cEdit one transaction in Block 2 and its fingerprint stops being c41b.Block 3 still says prev: c41b, so it no longer fits — and neither does anything after it.Rewriting history means rewriting every block since. That is the only security there is.
The fingerprint link is the whole invention. Everything else — mining, nodes, gas — exists to protect it.

Why thousands of copies instead of one database?

Because one copy needs an owner, and an owner is exactly what the design is trying to avoid. A single database is faster, cheaper and easier in every respect — the moment you are willing to trust whoever runs it. Every blockchain is a deliberate trade of efficiency for the removal of that one trusted party.

The cost of that trade is worth being blunt about. A chain that thousands of machines must agree on can process a tiny fraction of the transactions a single ordinary server handles, and every participant stores the entire history forever. That is not a bug awaiting a fix; it is the price of the property being bought. Anyone selling you “blockchain” for a problem where a trusted owner already exists and works fine is selling you the cost without the benefit.

For a trader, the practical consequence is capacity. Limited block space is rationed by price, which is why network fees rise when the chain is busy — you are bidding for room in the next block, against everyone else who also wants in. Your $3 transfer fee is not a price list. It is an auction result.

What is a confirmation, and why does your exchange wait?

A confirmation is your transaction being included in a block, and each further block built on top of it counts as one more. Your exchange waits for several because the newest block can still be replaced — until a transfer is buried under a few more blocks, something that looked settled can un-happen. That is the whole source of the delay you have actually experienced.

When you send a transfer it first goes to the mempool — a waiting room of transactions that have been broadcast but not yet included in any block. Nothing has happened yet. A miner or validator then selects it into a block, and that is one confirmation. Each subsequent block built on top adds another.

Why would one not be enough? Because two producers can find a valid block at almost the same instant and briefly broadcast competing versions of the newest page. Only one survives; the other is discarded, and any transaction that lived only in the discarded one goes back to being unconfirmed. This is called a reorg, and it means the freshest block is always the least certain. Each block piled on top makes reversal exponentially harder, because an attacker would have to outrun the entire honest network for every one of them.

So the exchange's rule — credit after N confirmations — is not bureaucracy. It is the exchange deciding how much reversal risk it is willing to carry on your behalf. A chain with ten-minute blocks and a three-confirmation policy means roughly thirty minutes of waiting, built into physics and policy rather than into the support queue. Confirmation requirements differ by exchange and by asset, and they are usually published on the deposit page itself: worth reading once for whatever you actually move.

Where a deposit actually spends its timeA five stage timeline of a deposit. Stage one, you press send and the transaction is broadcast, taking seconds. Stage two, it waits in the mempool competing on fee, taking seconds to hours depending on how much fee was paid. Stage three, it is included in a block, which counts as one confirmation. Stage four, further blocks are built on top until the exchange's required number is reached, the longest and least avoidable stage. Stage five, the exchange credits the balance. A note says the fee you choose only shortens stage two and has no effect on stage four.WHERE A DEPOSIT SPENDS ITS TIMEBroadcastsecondsMempoolseconds to hoursfee decidesIn a block= 1 confirmationWaiting for NN blocks x block timelongest stageCreditedtradeablePaying a higher fee buys you out of the mempool — and nothing else.The confirmation wait is fixed by block time and the exchange's policy. No fee shortens it.This is why "I paid the fast fee" and "it still is not credited" are both true at once.
Two different waits get blamed on one thing. Only the first is for sale.

What does the wait actually cost you?

About 0.30% of price uncertainty on a thirty-minute wait — roughly $15 on a $5,000 deposit at 40% annualised volatility, and double that once in twenty times. That is usually several times more than the network fee, which is the number everyone actually looks at. Almost nobody prices the wait, because it is invisible; the arithmetic is simple enough, so let us do it properly.

While your deposit is unconfirmed, the price is moving and you cannot act. The size of that exposure is not a guess — volatility scales with the square root of time, so a window of t minutes carries dispersion of σannual × √(t ÷ 525,600). Take an annualised volatility of 40%, which is roughly where bitcoin sits in calmer stretches, and a $5,000 deposit:

WaitWorkingPrice dispersion (1σ)On $5,000
5 seconds40% × √(0.083 ÷ 525,600)0.016%$0.80
3 minutes40% × √(3 ÷ 525,600)0.096%$4.78
10 minutes40% × √(10 ÷ 525,600)0.17%$8.72
30 minutes40% × √(30 ÷ 525,600)0.30%$15.11
60 minutes40% × √(60 ÷ 525,600)0.43%$21.37
Price dispersion carried by each length of deposit waitFive horizontal bars whose lengths are exactly proportional to the dispersion figure. Five seconds is nought point nought one six percent, three minutes is nought point nought nine six percent, ten minutes is nought point one seven percent, thirty minutes is nought point three zero percent and sixty minutes is nought point four three percent. Each bar is also labelled with the dollar figure on a five thousand dollar deposit. A note says the bars are drawn strictly to scale and that doubling volatility doubles every figure.DISPERSION CARRIED WHILE YOU WAIT - 40% ANNUALISED VOL, $5,000 DEPOSIT5 sec0.016% = $0.803 min0.096% = $4.7810 min0.17% = $8.7230 min0.30% = $15.1160 min0.43% = $21.37Bars are strictly to scale: 1,123 pixels per percentage point, every bar, no exceptions.Double the volatility and every number here doubles. It does not grow with the fee.
Quadruple the wait and the exposure only doubles — but it starts from the first minute, and it is free to nobody.

Two things fall out of this table that are worth more than the numbers themselves.

First, the wait usually costs more than the fee. A $15 band of uncertainty on a thirty-minute deposit dwarfs the two or three dollars most transfers cost. People optimise the visible number and ignore the one that is five times larger.

Second, and this is the part that actually matters: the figure scales with volatility, and you only ever rush a transfer when the market is moving. The 40% assumption is a calm-market number. In the hour you are most tempted to hurry money onto an exchange, realised volatility can easily be double or triple that — which makes the same thirty-minute wait carry 0.6% or 0.9%, or $30 to $45 on the same $5,000. The delay is largest exactly when you least want it. That is not bad luck; it is the structure of the thing.

Where this reasoning is wrong. Dispersion is not loss. It is symmetric — roughly half the time the wait moves the price in your favour, and over many deposits the average cost of the delay is close to zero. What you are buying with a faster chain is less uncertainty, not more profit, and that is only worth paying for if the uncertainty is actually large relative to the trade. If you are depositing to hold for a year, a 0.30% band is noise, and switching to an unfamiliar network to avoid it trades a $15 uncertainty for the risk of a $5,000 total loss. That trade is a bad one every single time.

Why do the same coins live on different chains?

Because they are not the same coins. This single misunderstanding destroys more beginner balances than bad trades do.

There is no one blockchain. There are many, each an independent ledger with its own blocks, its own fees and its own address format, and — crucially — none of them can see any of the others. When a token like USDT exists “on Ethereum” and “on Tron”, those are two separate contracts on two separate ledgers that happen to share a name and a price. Moving between them is not a transfer; it is a redemption on one chain and an issuance on the other, performed by a bridge or by an exchange acting as one.

So when a deposit page asks you to choose a network, it is not asking a preference. It is asking which ledger the money should arrive on, and the answer must match what the receiving side is watching. Send on a chain the recipient does not support and the coins are not lost in transit — they arrive perfectly, at an address on a ledger where nobody is holding the key. Sometimes support can recover them. Often they cannot, and the answer is a form and a long silence.

The practical habit is unglamorous and it works: send a small test amount first, wait for it to appear, then send the rest. On a $5,000 transfer, a $10 test transaction costs you a fifth of the dispersion we just calculated and removes the only failure mode on this page that is permanent. Our walkthrough on withdrawing crypto safely covers the checks in order, and how to check a transaction shows you how to read its progress yourself instead of refreshing a balance page.

When you trade on an exchange, is a chain involved at all?

No. And once you see this clearly, a lot of confusing things stop being confusing.

When you buy on a centralised exchange, the exchange decrements one number in its private database and increments another. No block is mined. No fee is paid to any network. Nothing is broadcast anywhere. That is precisely why the trade is instant and why the fee is a fraction of what an on-chain transfer costs — you are not using a blockchain, you are using a company's spreadsheet with an order book attached.

A blockchain enters your life exactly twice: when you deposit, and when you withdraw. Everything in between is an IOU. The exchange owes you the coins; it does not hold them in a box with your name on it. This is the whole content of the phrase not your keys, not your coins, and it is a statement about who bears the risk, not a slogan about ideology.

None of which means self-custody is automatically correct. An exchange protects you from the mistake that is far more common than exchange failure — losing your own seed phrase. The sane split is boring: trading balances on the exchange where they need to be liquid, long-term holdings in your own custody, and a deliberate answer to the question of which is which, rather than a default arrived at by never thinking about it.

SEE IT YOURSELF

Everything above is checkable in about two minutes and costs nothing. Open any exchange's deposit page for a coin you hold, and read the two lines most people scroll past: the network selector, and the required confirmations figure. Then look at the same asset's withdrawal fee on two different networks. The gap between those numbers is the whole of sections three to five on this page, written in the exchange's own words:

Referral links — they never change our assessment. Education only; most retail traders lose money.

What a blockchain is NOT

Not a place your coins are stored. There is no file and no folder. There is a history of transfers, and your balance is what you get when you add it up.

Not anonymous. Every transaction is permanently public and readable by anyone. Addresses are pseudonyms, and pseudonyms are linked to real people constantly, by exchanges, by analytics firms and by the users themselves.

Not automatically secure. The chain guarantees that a valid transaction cannot be reversed. It guarantees nothing about whether the transaction was a good idea, whether the contract you signed was honest, or whether the address you pasted was the one you meant.

Not one thing. Chains differ enormously in block time, fees, capacity and how decentralised they actually are. “It is on the blockchain” carries roughly as much information as “it is on a computer”.

Not an investment. Blockchain is a data structure. Thousands of unrelated projects use one, as do private companies with no token at all. Buying a coin is exposure to that specific asset, not to a technology sector.

Where does the "immutable and trustless" claim break down?

Immutable is a probability, not a property. The newest block is genuinely reversible, which is the entire reason confirmations exist. What the chain offers is history that becomes exponentially more expensive to rewrite as it ages — extremely strong after a few blocks, and not absolute at any point. Smaller chains with less total mining or staking power behind them are correspondingly cheaper to attack, and several have been.

Trustless just relocates the trust. You no longer trust a bank. You now trust that the software is written correctly, that the majority of validators behave, that the bridge you used is solvent, that the contract you approved does what its name suggests, and that you typed the address correctly. Most money lost in this industry has been lost at one of those points, not to a broken chain. The chain itself has an excellent record; everything built beside it does not.

Decentralised is a spectrum, and the marketing never says where on it. Whether a chain is genuinely hard to capture depends on how many independent parties run it and how concentrated the stake or hashpower is — measurable things that the word itself conceals.

And the risk that has nothing to do with any of this: understanding blockchains correctly will not make you profitable. The failure mode for new traders is sizing and psychology, not a misunderstanding of consensus. If you take one thing from this page onto a trade, let it be the order in which the decisions have to be made.

Where should a beginner go from here?

You now know the three parts of this subject that touch your money. The useful next steps, in order: read what Bitcoin actually is and why it has a price, which is the same machinery seen from the asset side, then how the market that prices it actually works — who is on the other side of your order.

Before anything moves, two practical pages: withdrawing crypto safely covers the wrong-network mistake in checklist form, and choosing a trustworthy exchange covers what to check before you deposit into anyone's database. And if you would rather start with a position than a transfer, the position size calculator answers the only question that has ever protected a beginner's account.

FAQ

What is a blockchain in simple terms?

A blockchain is a list of transactions kept in identical copies by thousands of independent computers. Transactions are grouped into blocks, and each block carries a fingerprint of the block before it, so changing an old entry would break every block after it. Nobody owns the list and no single computer can edit it alone.

Why does my exchange deposit need confirmations?

Because the newest block can still be replaced. Two miners can find a block at nearly the same moment and only one survives, so a transaction that looked settled can briefly un-happen. Each additional block built on top makes that reversal exponentially less likely. Exchanges wait for a fixed number of blocks before crediting you, which is why the delay exists at all.

Is a blockchain involved when I trade on an exchange?

No. Buying and selling on a centralised exchange changes rows in that company's private database. Nothing is written to any blockchain. The chain is only involved twice: when you deposit and when you withdraw. Everything between those two moments is an IOU from the exchange to you.

Why does sending on the wrong network lose your coins?

Because the same ticker exists as separate tokens on separate chains that cannot see each other. USDT on Ethereum and USDT on Tron are different contracts. Send to an address whose chain the recipient does not support and the coins sit at an address nobody controls. Sometimes support can recover them; often they cannot.

What does waiting for a deposit actually cost?

Not a fee — uncertainty. At 40% annualised volatility, a 30-minute wait carries about 0.30% of price dispersion, or roughly $15 on a $5,000 deposit, with a one-in-twenty chance of twice that. The figure scales with volatility, so it is largest at exactly the moments people rush their transfers.

Risk reminder: education, not advice. Crypto is volatile; most retail traders lose money, and you should never commit money you need.

Block-time, mempool, reorg and confirmation mechanics are general properties of proof-of-work and proof-of-stake chains; the number of confirmations required is set by each exchange and published on its own deposit page — check yours rather than relying on any figure here. The dispersion table and its chart are calculated by TradingPrimer from a single stated assumption of 40% annualised volatility, using the square-root-of-time rule on a 525,600-minute year, with the working shown so you can reproduce or replace it; bars in the chart are drawn strictly to scale at 1,123 pixels per percentage point. Dispersion is two-sided and is not a prediction of loss. Published 17 Sep 2026.