· four exchange APIs · rebuilt daily
Six decisions, in the order that keeps them from contradicting each other.

The interface presents these as unrelated settings. They are not — each one constrains the next, and doing them out of order is how people end up in positions they did not intend.
Isolated ring-fences a fixed amount of margin to that position. If it liquidates, you lose that margin and nothing else. Cross uses your whole balance as collateral, so the position survives much further — and takes everything with it when it does not.
Cross has legitimate uses for hedged books. Your first trade is not one of them. Start isolated.
Decide the risk budget and the invalidation point first; leverage is what falls out of them.
risk_amount = equity x risk_per_trade (1% is a normal starting point) stop_distance = |entry - stop| / entry position_size = risk_amount / stop_distance margin_required = position_size / leverage
Work it through in the position size calculator. If the position that fits your risk needs more leverage than you have, the answer is a wider stop or a smaller risk budget — never more leverage.
long : liq = entry x (1 - 1/leverage + maintenance_margin_rate) short : liq = entry x (1 + 1/leverage - maintenance_margin_rate)
At 10x with 0.5% maintenance margin, liquidation sits about 9.5% away. At 50x, about 1.5% — inside the ordinary hourly range of most large caps, which makes a 50x position a coin flip against noise regardless of whether the idea is right. Check yours in the liquidation calculator.
Your stop must sit inside your liquidation distance. If it does not, the exchange is your stop-loss, and it is a more expensive one.
A limit order posts and waits, pays the maker fee, and may not fill. A market order crosses the spread, pays the taker fee, and always fills. On MEXC's 0% maker versus 0.02% taker schedule, using limits takes a round trip from roughly 0.12% to 0.04% — a three-fold reduction in cost, for patience. Most retail flow is entirely taker out of impatience rather than necessity.
Post-only rejects the order rather than letting it cross, which guarantees the maker rate. Reduce-only can only shrink an existing position, which is what you want on every exit order.
Not after. The moment the position exists, both exits should exist. A stop you intend to place "once it moves in my favour" is not a stop, and the position is unmanaged until it is.
Perpetuals settle funding every eight hours. A 0.1% funding print is 0.3% a day — on a 10x position that is 3% of your margin daily, before fees. The funding calculator turns the rate into a number.
What the screen actually shows (checked 2026-09-03):
- Open and Close are tabs at the top of the ticket, not a buy/sell pair. - Margin mode is a dropdown reading Isolated, with leverage as a separate dropdown beside it reading 20X. - Order types are tabs: Limit, Market, and Chase Limit Order. - Quantity is entered in USDT by default, with a unit dropdown beside it. - A percentage slider runs 0 / 25 / 50 / 75 / 100 under the amount. - The ticket prints its own fee rate underneath: 'Fee rate Maker 0% / Taker 0.04%' on the contract observed. That is the number that applies to the trade, and it was double the 0.02% on the front page. - Take-profit and stop-loss were not visible in this capture; do not assume where they are.
*Interfaces change. This was account owner's capture of the futures trade panel, logged out; if what you see differs, trust the exchange.*
Small enough that the outcome does not matter. The purpose is to verify that you understand the mechanics — margin posted, fees charged, stop working, funding accruing — not to make money. Treat the first five trades as instrumentation.