A Price Is Not a Number. It’s a Queue.
What reading an order book taught me about what exchanges actually sell.
For most of my career as a developer, I thought a price was a number.
You look it up, you get a figure, the figure moves. If you want to buy, that’s what you pay. Simple.
Then I pulled a live order book from an exchange API and there was no price in it.
What actually comes back
Most exchanges expose a public order book endpoint that needs no authentication. On Binance it’s GET /api/v3/depth. Hit it and you get this shape back:
{
"lastUpdateId": 99479749980,
"bids": [
["79176.00000000", "1.86948000"],
["79175.99000000", "0.00111000"],
["79175.98000000", "0.00013000"],
["79175.91000000", "0.00007000"]
],
"asks": [
["79176.01000000", "3.67422000"],
["79176.02000000", "0.00090000"],
["79176.03000000", "0.00013000"],
["79176.24000000", "0.00400000"]
]
}Two arrays, trimmed to four rows each here. One side is everyone who wants to buy, sorted from the highest price down. The other is everyone who wants to sell, sorted from the lowest price up. Each row is a [price, quantity] pair, returned as strings.
There is no field called price anywhere in that response.
That bothered me for a while. Then it clicked, and it rearranged how I look at markets.
Price is the top of a stack
The number you see on a chart is derived. It’s the best bid, or the best ask, or the midpoint between them. It’s a summary of the first row of a much larger structure.
And that structure is a queue.
Exchanges match orders by price-time priority and most modern electronic venues run first-in, first-out. If you and I both place a buy order at the same price, whoever arrived first gets filled first. Your position in the queue is a real thing with real consequences.
So when a price “moves,” nothing mystical happened. A market order came in, consumed the resting orders sitting at the best level and the next level down became the new top of the book.
The price didn’t change. The queue ran out.
Once you’ve seen that, you can’t unsee it. A price chart stops looking like a line and starts looking like a log of things being eaten.
The price you see is not the price you get
Here’s the part that actually matters.
If your order is smaller than the quantity sitting at the best price, you pay the best price. Done.
If it’s bigger, you clear that level, then the next, then the next. Your order walks up the book until it’s filled and your average price ends up worse than the number that was on the screen when you clicked.
That gap is slippage. Nobody charged you a fee. It’s arithmetic.
How much worse it gets depends entirely on how much size is resting in the book. That quantity has a name: depth.
Thin book, big order, bad fill. Deep book, same order, price barely flinches.
Which leads to the conclusion that changed how I think about this industry.
What exchanges are actually selling
An exchange doesn’t sell you prices. Prices are public and roughly the same everywhere.
An exchange sells you the ability to trade size without moving the market against yourself.
It sells depth.
And depth is brutally hard to build, because it’s a chicken-and-egg problem wearing a network effect. Traders go where the book is deep. The book is deep because traders are already there. The industry has an aphorism for it: liquidity begets liquidity.
This is why exchanges pay market makers. Rebate programs, fee tiers, formal market-making agreements — all of it exists to solve the cold-start problem of an empty book. Somebody has to stand there quoting both sides before there’s a reason for anyone else to show up.
It’s also why, in derivatives especially, it is extremely difficult to pull liquidity away from an incumbent venue even with better technology and serious capital behind you. It’s a well- documented pattern: well-funded competitors have launched against entrenched exchanges and failed, not because their product was worse, but because they couldn’t reach critical mass in the book.
Get Hirusha Rasanga’s stories in your inbox
Join Medium for free to get updates from this writer.
Worth being precise here, because it’s a common overstatement: this doesn’t mean liquidity always concentrates in one place. US cash equities went the other way entirely. Trading there is spread across roughly a dozen public exchanges and dozens of dark pools, with a large share of volume executing off-exchange. Fragmentation happens. But when a venue does win depth in a product, that lead is very hard to take away.
Two books, one cent apart
Here’s the experiment. No API key needed. I pulled twenty levels from two places: Bitcoin against USDT on Binance and a tokenized IonQ share against USDT on MEXC.
Both came back with a one-cent spread.
On Bitcoin at $79,176, one cent is 0.0000126% of the price. On a $39.56 stock token, one cent is 0.0253%. The same number, two thousand times the meaning.
The rest of the comparison:
I expected the Bitcoin book to simply contain more money. It doesn’t. Adding up all twenty visible bid levels, the tokenized stock book held about $464,000 against Bitcoin’s $152,000. Three times more.
It’s just spread out. Bitcoin packs its liquidity into a price band three thousandths of a percent wide. The stock token spreads the same job across two percent.
That reframed depth for me. Depth is not how much is in the book. It’s how much is sitting at the touch, where your order will actually land.
You can see the market maker
One more thing you notice reading raw books instead of charts.
The tokenized stock ladder went 373.91, 367.64, 364.69, 375.14, 394.18 — near-identical sizes at near-even four and five cent gaps, then a clean step up to roughly 1,000 units at the deeper levels.
Order flow from actual humans doesn’t look like that. That regularity is a quoting system doing its job. I can’t prove it from public data, but the shape is hard to read any other way.
And then, parked at the very front of it, two tiny orders worth about $98 and $83, sitting ahead of the entire structure.
That’s the market’s plumbing, visible in a JSON response, for free.
Why this is the whole game right now
I keep coming back to this because it’s playing out live in traditional assets moving onto crypto exchanges.
In a joint report with CoinGecko published in July 2026, MEXC put numbers on it. TradFi trading volume across six major centralized exchanges rose from $3.46 billion in January 2025 to $393.15 billion in June 2026. US stocks overtook precious metals as the largest category in June, with monthly volume up 337.4% to $189.84 billion.
But the number I’d point at is a different one: perpetual futures accounted for 98.5% of that June volume. Spot was under $6 billion of it.
That tells you something. People aren’t mostly buying tokenized shares to hold them. They’re trading leveraged exposure and leveraged traders are the most execution sensitive users there are. They care about spread, depth and whether the book holds up when things get violent. They will move venues over basis points.
Which is the order book story again, at industry scale.
Listing an asset is the easy part. Any exchange can do it and increasingly every exchange has. Making that asset liquid enough that a serious size order doesn’t get destroyed on the way in — that’s the hard part and it’s the part that decides who’s still standing.
What it changed for me
I came at this from the engineering side. I wanted to understand a data structure.
What I got instead was a business model.
Once you see that a price is a queue, the rest follows fast. Liquidity is inventory. Market makers are suppliers. And an exchange’s real job is filling its own book before anyone shows up wanting to trade.
That isn’t a technical problem. It’s partnerships, incentives and relationships.
I still read the API docs first. But now, when I look at an exchange, the first thing I check isn’t the feature list.
It’s the depth.
Sources
- Binance Spot API documentation,
GET /api/v3/depth— order book response format - Montaka Capital on CME and the ELX case — liquidity network effects in derivatives
- CoinGecko & MEXC, How Crypto Exchanges Are Reshaping Traditional Asset Trading (July 2026)
- FINRA ATS transparency data — US equity off-exchange volume share
Order book snapshots taken from Binance /api/v3/depth and MEXC /api/v3/depth, 20 levels each, pulled a few minutes apart. Both are twenty-level windows, not full books — real depth extends further on both venues. Slippage figures are simulated by walking the visible ask side.
I’m a developer who spends a lot of time reading exchange APIs and thinking about market structure. If you work on the liquidity or business side of this, I’d like to hear where you think I’m wrong.
