A shared list that nobody owns
Start with the ordinary case. When you hold money at a bank, the bank keeps a record of what
you have. That record is the balance. If the record says a different number tomorrow, then
that is your balance, because the institution keeping the record is the authority on it.
A cryptocurrency replaces that single record with one kept simultaneously by thousands of
independent machines. They exchange every new transaction, check it against the rules, and
append it. No one of them is the authority; agreement between them is. That single structural
change is what produces every other property people associate with crypto, including the
inconvenient ones.
- Kept by
- Many machines
- Authorised by
- A private key
- Settled after
- Confirmation
- Reversible
- No
Ownership is the ability to sign
There is no account in the banking sense: no name attached, no password, nobody to appeal to.
What exists is a pair of numbers. One is private and only you have it. The other is derived
from it and can be shared freely, and it produces the address funds are sent to.
Spending works by signing. Your wallet uses the private key to produce a signature over the
transaction you want to make, and anyone can check that signature against the public half
without ever learning the private one. That check is the whole security model. It means
whoever holds the key can spend, and nobody else can, including you, if you lose it.
What a confirmation means
A transaction announced to the network is not yet settled. It waits in a queue with everyone
else’s until it is bundled into a block and appended to the ledger. That first inclusion is one
confirmation. Each subsequent block built on top makes reversing it more expensive, which is
why services wait for several before treating funds as final.
The fee you pay buys priority in that queue. When the network is busy, fees rise because
everyone is bidding for limited space, and a transaction with a low fee can sit unconfirmed
for a long time. This is not a malfunction; it is the queue working as designed, and it is the
most common reason a transfer appears to be stuck.
Why the same word means three different things
“Crypto” is used for the technology, the asset class and the industry, often in the same
sentence. Separating them helps: the technology is a way of agreeing on a record, the asset
class is what people trade, and the industry is a set of companies with commercial
interests in both. A claim that is true of one is frequently false of the others.
Why the price and the technology are only loosely connected
Nothing about a shared ledger determines what a unit on it should cost. Price comes from
what buyers will pay, which is driven by expectation, access and available supply. A
technically excellent network can have a falling price, and a technically unremarkable one
can have a rising price, for extended periods.
Why “decentralised” is a spectrum, not a state
Networks differ enormously in how distributed they actually are: who runs the software, who
can change it, how concentrated the holdings are, and how many independent parties would
have to cooperate to censor a transaction. The word gets used as a binary label, and the
useful version of the question is always “decentralised in which respect”.
Where the fee comes from, and why it moves
Every transfer pays a fee, and the fee is not a price set by anyone. Space in a block is
limited, transactions compete for it, and whoever writes the next block takes the ones offering
most. When few people are transacting the competition is weak and the fee is small. When many
are, the same transfer costs several times more for no change in what it does.
This catches people out because it inverts the intuition that a payment network charges a
published rate. There is no rate. There is an auction that clears continuously, and a wallet
quoting a fee is estimating what the auction will accept in the next few minutes. Estimates
can be wrong in both directions, which is why a transfer sometimes confirms immediately and
sometimes sits waiting.
The practical handle is that fees depend on the size of a transaction in data rather than in
value. Sending a large amount does not inherently cost more than sending a small one. What
raises the cost is a transaction assembled from many small previous receipts, because it takes
more space to describe.
What is public, and what is not
The ledger is readable by anyone, which is what allows the whole arrangement to work without a
central record keeper. Every transfer, every balance and every address is visible from the
first block onwards, permanently, to anybody who wants to look.
What is not published is the connection between an address and a person. That link is the whole
of the privacy, and it is more fragile than most first-time readers assume. It is established
the moment an address is used somewhere that knows an identity, and once established it applies
to the entire history of that address in both directions.
Calling this anonymous is therefore wrong in a way that matters. Pseudonymous is the accurate
word: the names are missing, the record is complete, and the two are joined by a single link
that is easy to create and impossible to undo afterwards.
What this makes possible, and what it costs
The gain is a payment that does not require permission from an intermediary and cannot be
reversed by one. For some uses that is transformative; for most everyday payments it is
irrelevant, because the intermediary was providing services (dispute resolution, fraud reversal, credit) that people value more than they realise until they are absent.
The cost is that responsibility moves entirely to the holder. There is no forgotten-password
flow, no fraud department and no recall. Every guide on this site that seems excessively
concerned with backups and test transfers is concerned with exactly this, and the concern is
proportionate rather than cautious.
Be the first to comment