How Cryptocurrency Prices Are Actually Set: Order Books, Liquidity and Demand

A cryptocurrency’s price is not a value someone calculates. It is simply the price of the most recent trade on an exchange order book, and it rises when buyers are willing to pay more than sellers currently ask. That is the entire mechanism. Everything else — news, sentiment, adoption, regulation — only matters because of how it changes what people are willing to bid and ask. Understanding that plumbing is what stops you from being surprised by a price move.

We sell mining hardware, so we care about this for one specific reason: the Bitcoin price is an input to what a miner earns. It is not the whole story, and the rest of that story is in what a miner actually earns.

How the price is actually set

An exchange keeps an order book: a list of bids (what buyers offer) and asks (what sellers demand). The quoted price is where the two last met.

• When buying pressure is stronger, buyers lift the lowest asks in turn and the price walks upward.
• When selling pressure is stronger, sellers hit the highest bids in turn and the price walks downward.

Nothing more mystical is happening. And because the price is the last trade, it says nothing about how much could be bought or sold at that price — a distinction that catches people out constantly.

Liquidity, from the side of the trade a miner is on

Liquidity is the depth of the order book: how much you can trade before you move the price against yourself. A market can display an attractive price and have almost nothing available at it. Every explanation of this is written for someone deciding whether to buy. A miner is on the other side, and permanently.

Your electricity bill arrives in dollars, monthly, whether or not you feel like selling. Your income arrives in coin. That makes you a structural seller — not someone making one trade they can time, but someone placing a standing sell order for as long as the machines are running. Every argument about waiting for a better price runs into the fact that the power company will not wait with you.

Which is why the depth question is sharper for a miner than for an investor:

• You sell on a schedule you do not choose. A thin book costs you a little on every one of those sales, month after month, and it compounds quietly into your actual realised revenue.
• Your sale is correlated with everyone else’s. Miners face the same bills at the same time and are squeezed hardest by the same events — a price fall, a difficulty rise, a cold snap on the power market. The moment you most need to sell is the moment other miners are selling too.
• Scale makes it worse, not better. One machine’s output is noise on any book. A fleet’s output on a minority chain is not.

The scale of the difference is easy to check and rarely checked. On 30 July 2026 Bitcoin’s market capitalisation was about $1.28 trillion against Bitcoin Cash’s $4.2 billion — a factor of roughly 300, on two chains that use the same algorithm and the same hardware and, that day, paid within 0.1% of each other per unit of work. The payout was identical; the exit was not. That comparison is worked through in mining Bitcoin Cash with SHA-256 hardware.

Check the coin’s real daily traded volume and its market capitalisation before you check a mining calculator — we say the same thing in cryptocurrency mining hardware beyond Bitcoin, because it is the risk that is least often priced. A calculator will happily quote you daily earnings in a coin you cannot sell.

What changes supply and demand

• Supply mechanics. For Bitcoin, new supply arrives on a fixed schedule that halves roughly every four years, described in where new bitcoins come from. No one can decide to issue more, which makes supply the one genuinely predictable side of the equation.
• Demand. Everything else. Adoption, macroeconomic conditions, interest rates, regulatory news, exchange access, and plain speculation. This side is not predictable, and anyone who tells you otherwise is guessing with confidence.
• News and sentiment. Coverage moves prices because it moves the willingness to bid and ask, and it works in both directions — negative coverage can trigger selling that feeds on itself. Sentiment is real, but it is a description of behaviour, not a forecast of it.
• Concentrated holders. Large holders can move a thin market substantially in one order. That is covered in who actually moves the price.

Why we do not publish price predictions

We sell hardware, and it would be easy to make our machines look better by attaching an optimistic price forecast to them. We do not, for a straightforward reason: nobody can forecast this, and a business plan built on someone else’s price target is not a plan.

You do not need a forecast to make a good hardware decision, which is the genuinely useful part. The mining decision can be made entirely on things you can measure today:

• Your all-in delivered electricity rate — total bill divided by total kilowatt-hours
• The machine’s efficiency in J/TH, which is a fixed property of the hardware
• Today’s hashprice, and the fact that difficulty trends upward

Put those together with break-even $/kWh = hashprice ÷ (24 × J/TH) and you know the electricity price at which a machine stops earning — without predicting anything. If it only works on the assumption that prices rise, it does not work. The method is in the ASIC miner buying guide, and the machines are ranked by that number here.

The one part of future revenue you can actually see

There is exactly one forward-looking number in mining that is worth quoting, and it is not the price. It is the next difficulty retarget.

Bitcoin re-calculates difficulty every 2,016 blocks. Part-way through an epoch you can estimate the size of the next change from how fast blocks have actually been arriving — slower than ten minutes means difficulty is heading down, faster means up. On 30 July 2026 the previous retarget had come in at −0.74% and the next was estimated at about −3.6%, roughly 1,400 blocks away.

That is a real, checkable, short-horizon statement about miner revenue: a negative retarget means every machine still running earns a few percent more per terahash, with no change in the coin price at all. It is small, it revises as the epoch runs, and it only looks a week or two ahead. It is also the only prediction on this website, and the contrast with a price target is the whole point — one is arithmetic on blocks that have already been found, the other is a guess about what strangers will feel like paying.

Background: what Bitcoin is and what mining actually does. Stock: the full miner inventory.

Frequently Asked Questions

How does a cryptocurrency price rise?

Through the exchange order book. Buyers post bids, sellers post asks, and the quoted price is wherever the two last met. When buying pressure is stronger, buyers take the lowest asks in sequence and the price walks upward; when selling pressure dominates, it walks down. Everything else — news, adoption, regulation, sentiment — only affects price by changing what people are willing to bid and ask.

What determines the value of a cryptocurrency?

The balance of supply and demand, of which only supply is predictable. Bitcoin’s new supply arrives on a fixed schedule that halves roughly every four years and cannot be changed by anyone. Demand covers everything else: adoption, macroeconomic conditions, regulation and speculation. That asymmetry is why supply-side facts are worth studying carefully and demand-side forecasts are worth very little.

Why does market liquidity matter when choosing what to mine?

Because a miner is a structural seller. Your electricity bill arrives in dollars every month and your income arrives in coin, so you are not making one trade you can time — you are placing a standing sell order for as long as the machines run. On a thin book that recurring sale moves the price against you every time. Check a coin’s real daily traded volume and its market capitalisation before checking any mining calculator: on 30 July 2026 Bitcoin’s market cap was about $1.28 trillion against Bitcoin Cash’s $4.2 billion, a factor of roughly 300, and order-book depth broadly follows that.

Should I buy mining hardware based on a Bitcoin price prediction?

No. Nobody can reliably forecast the price, and a hardware purchase justified by someone else’s target is a bet dressed up as a plan. The decision can be made entirely from measurable quantities: your all-in electricity rate, the machine’s J/TH efficiency and today’s hashprice, combined through break-even rate equals hashprice divided by 24 times J/TH. If a machine only works assuming prices rise, it does not work.

Is there anything about future mining revenue that can be predicted?

One thing, and it is not the price. Bitcoin’s difficulty re-targets every 2,016 blocks, and part-way through an epoch the size of the next adjustment can be estimated from how fast blocks have actually been arriving. On 30 July 2026 the previous retarget had been −0.74% and the next was estimated at about −3.6%, roughly 1,400 blocks out. A negative retarget raises revenue per unit of hashrate for everyone still running. It is a small, short-horizon, revisable estimate — but it is a real one, which is more than any price forecast is.

Does the coin price reach my mining revenue directly?

Only partly, and with a lag. What you earn is hashprice — the network’s payout per unit of hashrate — and that is the coin price divided by the work competing for it. A price rise lifts hashprice immediately, then attracts hardware, which raises difficulty at the next retarget and gives part of the gain back. It is entirely normal for revenue per terahash to fall during a period when the coin price rose.