Daily Picks so far: about 0.9 a week (at most two a day; many days none). Each one tells you how much you risk, how big to trade, where to stop and where to take profit.
✓ Free in the beta✓ Works on major exchanges✓ You decide every trade
An example plan, as it arrives on Telegram and the site.
0.9
Daily Picks a week so far (at most 2 a day)
6 years
of past data behind the tests
69
rules tested — 55 rejected
100%
of results public
The main crypto futures are checked day and night: trend, sudden moves and data quality.
When a tested rule finds a good setup, you get the coin, entry, stop, targets and a leverage limit. At most two a day; many days none.
Choose your risk, place the orders on your exchange, and follow the updates until the trade ends.
Entry, stop loss and two take-profit levels, with the distance of each in percent.
Type how much you are happy to lose on the trade. CMI tells you how big to trade.
Each plan shows the highest leverage that keeps liquidation beyond the stop. Lower is always fine.
New plans, take-profits and closes arrive the moment they happen.
See which coins are in an uptrend and which are close to a signal.
Wins and losses are recorded and never deleted. You can check our record any time.
A rule earns trust over many trades, not one good or bad week.
🧪
Tested before going live
Rules are tested on up to 6 years of past data after fees. 55 of the 69 rules tested so far did not pass and were dropped.
📈
Proving itself live
Daily Picks v2 (selection record): 3 of 60 live trades so far. First verdict from 2027-04-13.
🔒
Every result recorded
Every trade is saved day by day in a chained record, so nothing can be changed or removed later. Check it any time.
Which coins are in an uptrend, and how close each one is to a Daily trend signal.
The latest headlines from trusted crypto news sites, updated through the day.
Cointelegraph · 33 min ago
Decrypt · 53 min ago
The Block · 1 h ago
Decrypt · 1 h ago · BTC
Decrypt · 2 h ago
Headlines from the publishers' official feeds, linked to the original articles. For context only — signals never depend on news.
Yes — the beta is free until at least 31 January 2027, with no card needed. You need an invite code to create an account: ask for one on the join page, get one from a member, or find one in the Telegram channel.
Usually a few a week and many days none: at most two Daily Picks a day, plus at most two experimental beta signals with their own record. Waiting for a good setup is part of the method.
Most plans of the Daily trend rule lose: many small losses have to be paid for by a few large winners, so a win rate alone says little. Judge a rule by its average result over many trades — the live record of every rule, losses included, is on the Live tests page.
Take-profit levels. At TP 1 you close half and move your stop to the entry price, so the rest can no longer lose. The other half closes at TP 2, or follows the trend with a trailing stop.
Never more than the plan shows, with isolated margin. Your risk comes from the position size, not the leverage — lower leverage is always fine.
Plans are for USDT perpetual futures, which most major exchanges offer. Use one that is allowed where you live. Prices come from Binance; other exchanges usually differ by a few ticks.
The rules passed their tests on past data and now run on live prices with simulated money. CMI places no real orders. A plan stays marked “Live test” until its rule has proven itself live.
No. CMI shares rule-based trade plans and market information. You decide if and how much to trade.
Join with an invite code, or ask for one — it takes a minute. Want to look first? Read how the rules were tested.
| 0.63110 |
| 0.73250 |
| +16.07% |
| QNT | ▲ Uptrend | 192.160 | 223.820 | +16.48% |
| 1000PEPE | ▲ Uptrend | 0.00487 | 0.00569 | +16.76% |
| ARB | ▲ Uptrend | 0.15796 | 0.20427 | +29.32% |
| INJ | ▲ Uptrend | 7.2500 | 9.4220 | +29.96% |
A long signal needs a daily close above the breakout level (20-day high) while above the 100-day trend line. Measured at the last closed daily candle (); it is information, not a signal.