A friend who was maybe two weeks into trading sent me a screenshot of three prices for the same coin on the same exchange and asked which one was real. ETH in the USDT pair, ETH in the USDC pair, ETH in the BTC pair. Two of the numbers were close but not identical, and the third looked like a typo. All three are real, and the confusion comes from expecting a coin to have one price in the first place. What an exchange shows you is an exchange rate between two specific assets, and once that idea settles in, a lot of things that trip up new traders, from odd price gaps to losing money on trades that went up, stop being mysterious.
Base on the left, quote on the right
Every pair on every exchange follows the same convention. In BTC/USDT, BTC is the base currency and USDT is the quote currency. The base is the asset you are buying or selling. The quote is the asset you pay with when you buy and receive when you sell. Price is always expressed as units of quote per one unit of base, so if BTC/USDT prints 50,000, one bitcoin costs 50,000 tether. That number says nothing about what a dollar or a tether is worth on its own. It only tells you the going rate between those two assets.
Order tickets follow the same logic. When you set a limit price, you set it in the quote currency. When you sell, your proceeds land in the quote currency, and that one detail matters more than anything else in this post. Sell ETH in the ETH/BTC pair and your proceeds arrive as bitcoin, with everything that implies about what your account is exposed to afterward.
The convention takes a while to feel natural because the pair reads left to right while the mental question usually runs the other way. My shortcut is that the base is the thing and the quote is the money. BTC/USDT is bitcoin priced in tether money. ETH/BTC is ether priced in bitcoin money. Clunky phrasing, but it sticks.
Why the same coin has three different prices
Each pair is its own order book with its own buyers and sellers. ETH/USDT and ETH/USDC are separate markets that happen to share a base asset, and nothing forces them to agree except arbitrage traders, who buy in whichever book is cheaper and sell in whichever is richer until the gap shrinks below the cost of the round trip. So the pairs track each other closely, but only to within fees, and a small residual gap is normal.
There is a second reason the stablecoin pairs disagree. The stablecoins themselves trade slightly apart from each other and from the dollar, because USDT and USDC are claims on their issuers and the market reprices those claims continuously. Most days the difference is a few hundredths of a percent and you can ignore it. Under stress it gets much wider. USDC traded meaningfully below a dollar for a weekend in 2023 when one of its reserve banks failed, and every USDC-quoted pair on every exchange repriced with it, even though nothing had happened to the coins trading in those pairs.
The BTC pair looks completely different because the unit is different. ETH priced in bitcoin is a small decimal, and to compare it with the USDT pair you multiply by bitcoin's own dollar price. When a new trader spots what looks like a huge discount in a BTC pair, it is almost always this unit conversion, and the leftover difference after converting is usually smaller than the fees it would take to capture. If you ever do see a genuinely large gap between pairs, especially on a smaller exchange, reach for the boring explanations first. A thin book where the last trade is stale, or a venue with withdrawal problems, where prices drift because arbitrageurs cannot move funds in and out to close the loop. Historically, coins on exchanges with frozen withdrawals have traded at enormous premiums or discounts for exactly that reason, and a gap nobody is arbitraging away is a warning about the venue rather than an opportunity.
Your profit and loss lives in the quote currency
This is the part that actually costs people money. When you hold a position in any pair, you are long the base and short the quote for as long as the trade is open. That sounds like derivatives jargon but it is plain accounting. Buy ETH/BTC and you make money when ether outperforms bitcoin and lose when it underperforms, regardless of what either one does in dollars.
Run a plausible example. You buy an altcoin in its BTC pair. Over the next couple of months the alt gains roughly 20 percent in dollar terms, which feels like a win. Bitcoin gains roughly 30 percent over the same stretch. Your position, measured in the pair you actually traded, is down about 8 percent, because the unit you are measured in ran faster than the asset you bought. Whether you made or lost money depends entirely on which unit you keep score in, and most new traders have never consciously picked one. They keep score in dollars while trading BTC pairs, or the reverse, and end up with a portfolio where every position is silently a relative value bet they never meant to make.
The quote also decides where your money sits after you exit. Sell into USDT and your gains become a tether balance, with whatever issuer risk that carries. Sell into a BTC pair and your gains become bitcoin, which means a good trade can still round-trip to nothing if bitcoin sells off while you sit there feeling clever. Neither choice is wrong. Holding proceeds in BTC is perfectly reasonable if you wanted bitcoin exposure anyway. The mistake is making that choice by accident, based on whichever pair happened to be on screen, instead of on purpose.
Picking the pair, in practice
My default rule is boring. Trade the deepest book unless you have a specific reason to do otherwise, and on most exchanges the deepest book for a given coin is its biggest stablecoin pair, typically USDT on offshore venues and USD or USDC on the more regulated ones. Depth means tighter spreads and less slippage, and for a retail-sized order those two costs usually dwarf everything else, including the fee tier differences people spend hours optimizing.
Reported volume is a poor proxy for depth, because volume is easy to fake and depth is expensive to fake. Wash trading inflates the volume column on plenty of smaller venues, but nobody wash trades resting limit orders near the mid, since real traders can hit those. Judge liquidity from the order book rather than the volume leaderboard. Before entering, I run some version of this checklist.
- Compare the bid and ask spread across the coin's available pairs. Tighter is cheaper, and the difference between pairs is often several times larger than the trading fee.
- Check how much size actually rests within roughly half a percent of the mid price. That is what your market order will eat, and a thin book shows itself here immediately.
- Decide what asset you want to hold after the trade and prefer the pair quoted in it, so your exit does not require a second conversion across a second spread.
- If you deliberately trade a BTC or ETH quoted pair, write down that the position is a bet on relative performance, so future you does not misread the profit and loss.
- On a smaller exchange, sanity check the price against a major venue before treating any discount as real.
The failure mode I have seen most often goes like this. Someone buys a mid-cap coin in its BTC pair because the decimal looks cheap, crosses a wide spread on the way in, watches the coin do fine in dollars while bleeding in BTC terms, then crosses another wide spread on the way out and wonders where the money went. Every individual step looked harmless, and the quote currency choice quietly determined the whole outcome.
Comparing depth across pairs and venues by hand gets tedious, which is part of why we surface per-pair liquidity in Blockcircle's market scorecards instead of a single ticker-level number. But you do not need tooling for the core habit. Read the pair right to left before you trade it. Know what you are paying with, know what you will be paid in, and pick both on purpose. A surprising amount of the expensive confusion in a trader's first year traces back to skipping that step.