Home › Learn › Lesson 5

How the crypto market actually works — supply, demand and market makers

By TradingPrimer · Updated 11 Sep 2026 · Published 28 Aug 2026

The crypto market has no price-setting authority. The number on your screen is simply the last trade that crossed the spread on that one exchange — one buyer who agreed to pay a seller's asking price. Everything else in this lesson follows from that single fact: why your fill is worse than the price you clicked, why market makers exist and what they are actually paid for, and why "the price" is different on every venue at the same instant.

Price is not published — it is the residue of the last trade

Ask most beginners where the price of Bitcoin comes from and you get a vague answer about supply and demand. That's true but useless, in the same way "planes fly because of physics" is true but useless. Here is the mechanically precise version.

Every exchange maintains a list of unfilled orders called the order book. On one side are bids: people who have committed to buy at a specific price and are waiting. On the other are asks (or offers): people committed to sell at a specific price, waiting. The highest bid and the lowest ask are the two numbers that matter. The gap between them is the spread, and inside that gap there is no price at all — just an empty space where nothing has traded.

Nothing in that book moves the price. A million dollars of resting bids at $59,000 changes nothing while it sits there. The price changes only at the moment someone becomes impatient and crosses the spread — sends an order that immediately consumes the best offer on the other side. That crossing is the trade, and that trade's price becomes "the price."

This is the first thing that separates people who understand markets from people who watch them: the chart is not a measurement of belief or value. It is a log of the moments impatience beat patience. Every candle you will ever read in the rest of this curriculum is built from these events and nothing else.

What a market order really costs you

Beginners assume a market order buys at the price on the ticker. It doesn't. It buys the cheapest offer available, then the next cheapest, then the next, until your quantity is filled. You pay the average of everything you consumed.

Take the book in the diagram above and send a market buy for 5 BTC:

Your average fill is $300,139.50 ÷ 5 = $60,027.90. You clicked at $60,000 and paid $27.90 more per coin — $139.50 of slippage, or 0.047% of the trade, before a single cent of exchange fee. On a liquid pair like BTC that's trivial. On a thin altcoin book, where the same five levels might span 2% instead of 0.1%, the identical order costs you twenty times more — slippage goes from 0.047% to about 0.93%, or $2,790 instead of $139.50 — and it does so silently: no error message, no warning, just a worse entry that quietly shifts your stop-loss distance and your risk-reward for the life of the trade.

The practical consequence, and it is the whole point of this lesson: your fill quality is not a detail your strategy can ignore. It is set before you enter, by which market you chose and what size you sent — and it is the one input you can measure in advance. Look at the book, not just the chart, and your position sizing stops being fiction.

What market makers are actually paid for

Somebody has to be sitting on the other side of your market order, willing to sell instantly at a fair-ish price. That somebody is usually a market maker: an automated participant quoting both a bid and an ask at all times, aiming to buy at the bid and sell at the ask over and over, ending the day roughly flat.

The arithmetic of that business explains their behaviour better than any conspiracy theory. Suppose a maker quotes $59,995 bid / $60,005 ask — a $10 spread on a $60,000 asset, which is 0.017%. On 0.1 BTC, one completed round trip earns $1. To make $2,000 in a day they need roughly 2,000 round trips: one every 43 seconds, all day, every day. Their entire profit lives in tiny margins repeated at enormous frequency.

Now notice what that implies. A market maker's nightmare is not being wrong about direction — it's adverse selection: the trades that fill fastest are the ones where the other side knows something. When news hits and price is about to move 3%, the maker's resting bid gets hit instantly, and the inventory they just bought is worth less a second later. Their only defences are to widen the spread or pull the quotes entirely.

So the spread is not a fee for existing — it is the price of the maker's risk. And it widens precisely when the market is moving fastest, which is precisely when you feel most urgent about clicking market. That is the second-order consequence almost nobody explains: volatility does not just move the price against you, it makes the act of trading more expensive at the same moment. The trader who plans entries in advance with limit orders pays that tax rarely. The trader who reacts to candles pays it constantly.

Only one of those five groups is trading because it wants to. Understanding that the other four are executing mandates — spread capture, arbitrage, order-slicing, forced closure — is what stops you from reading every move as a message about the future.

There is no such thing as "the" crypto price

Traditional equities have consolidated tape: one official reference price per stock. Crypto has nothing of the sort. Binance, OKX, Bybit, Coinbase and hundreds of others each run a separate matching engine with a separate book, and each produces its own last-traded price. Arbitrage bots buy on the cheap venue and sell on the expensive one, which pulls them together — but arbitrage capital is finite, and it thins out exactly when moves are violent.

The consequence is concrete and it costs people money. If a cascade drives your exchange down to $58,900 in a three-second wick while another venue only prints $59,150, a stop at $59,000 is filled — at your exchange's price, on your exchange's thin book. Traders then spend the evening arguing that "the price never got there," pointing at a chart from a different venue. It did get there, where it mattered: in the book their order was resting in. The mechanics of how those wicks form are taken apart in anatomy of a liquidation cascade.

The same fragmentation is why derivatives complicate the picture. Perpetual futures often carry more volume than spot, and their price is anchored to spot not by law but by the funding rate — a payment between longs and shorts that makes the crowded side pay to stay. If you're still deciding which of the two to trade first, the trade-offs are laid out in spot vs futures.

The market cap illusion — a worked example

One last consequence of "price is just the last trade" is worth doing in numbers, because it protects you from an entire category of bad decisions.

Imagine a token with 1,000,000,000 coins in supply, last traded at $1.00. Its quoted market cap is $1 billion. Now suppose the ask side of its book is thin, and a buyer spends $200,000 walking that book up until the last trade prints at $1.20. Market cap is now 1,000,000,000 × $1.20 = $1.2 billion. Two hundred thousand dollars of real money "created" two hundred million dollars of market capitalisation.

Nothing was added to the token. Market cap is not money held anywhere — it's a multiplication using the price of the most recent, possibly tiny, trade. That arithmetic runs identically in reverse, which is why thin assets can lose 70% of "market cap" on modest selling. Whenever you see a market-cap comparison used as an argument ("if it just reached half of X's cap…"), you are looking at a multiplication, not a valuation.

What this changes about how you trade

Three concrete changes you can make today, all of which follow directly from the mechanics above:

1. Check depth before size. Before your entry, look at the book and ask how far your order will walk it. If a position you're planning would consume several price levels, the position is too big for that market — not too big for your account. This is a different question from the 1% risk rule, and it binds first.

Here is the version of that rule you can actually check, using the same book as above and a stop placed 1% away ($600 on a $60,000 entry). Slippage moves your entry against you, so measure it against your stop distance — that is the budget it eats into:

Read the bottom-right cell slowly: on the thin book, the 5 BTC order starts the trade 93% of the way to its own stop before price has moved at all. The trade is not unlucky, it is unsurvivable by construction. A usable test: keep slippage under roughly a twentieth of your stop distance. On the deep book that allows about 5.35 BTC; on the thin one, about 0.94 BTC — the same account, the same risk rule, a position limit 5.7 times smaller because of the venue alone.

Where this test stops being right: it assumes your stop distance is set by structure and your fill is a single market order. If you scale in with limit orders you are supplying liquidity rather than consuming it, and the slippage column largely goes away — a different trade-off, covered in scaling in and out.

2. Default to limit orders; spend market orders deliberately. A limit order makes you the patient side and often earns the cheaper maker fee. Use market orders when being filled matters more than the price — getting out of a losing trade, for instance — and accept the cost consciously rather than by habit.

3. Judge your stop against your own venue. Place stops beyond obvious clusters (round numbers, the exact low of a swing) on the exchange you actually trade, and expect your venue's wicks to be worse than the pretty chart you screenshot from somewhere else.

None of these require predicting anything. That is a pattern you'll see repeatedly through this course, and the reasoning behind sequencing it this way is set out in knowledge capital before trading capital. You can also test all three without money in the practice arena, where fills, stops and sizing behave like a real venue.

Common mistakes at this stage

Trading the chart without ever opening the book. The candle is the output; the book is the machine. Traders who never look at depth are consistently surprised by their own fills. Assuming a big buy "adds money" to an asset. It doesn't — it transfers money to a seller and re-prices the remaining supply. Treating cross-exchange price differences as proof of manipulation. They're the normal state of a fragmented market; arbitrage narrows them, it never erases them. Using market orders for entries and limit orders for exits. That is exactly backwards: pay for urgency when you need to leave, not when you choose to arrive.

FAQ

Who sets the price of Bitcoin? Nobody. The number is the price of the most recent trade on that specific exchange — a record of what already happened, not a valuation anybody publishes.

Do market makers hunt retail stop losses? Most market making is directionally neutral: quote both sides, capture the spread, stay flat. But resting stops are a known pocket of liquidity and some participants do trade toward them. The defence isn't outrage — it's placing stops beyond the obvious levels rather than just under them.

Why is the price different on different exchanges? Each venue has its own order book and therefore its own last trade. Arbitrage keeps them close but never identical, and gaps widen during fast moves. Your stop is triggered by your exchange's price.

Does buying a coin put money into it? No. Your money goes to the seller. Market cap is just the latest price multiplied by supply, which is why small amounts of buying can inflate it enormously.