Skip to content

Put your trading idea to the test with real data.

Write it in your own language, the way you would explain it to another trader. Arbit simulates it on years of real market data, shows you what would have happened, even if it loses, and optimizes it by testing hundreds of configurations.

REAL CASES

An idea that sounds good is not an idea that works.

Same coin, same period and same costs. Only the idea changes.

The idea almost everyone tries

“Buy when the 9 EMA crosses above the 21 EMA and sell when it crosses below.”

BTC · 4h candles · 2 years

−12.3%simulated result
Trades
199
Won per $1 lost
$0.79
Largest drop
15.7%

Real BTCUSDT data, Sep 18, 2024 to Sep 18, 2026: 30% of capital per trade, 0.1% fee, 12% take profit and 6% stop loss. Simulated results on historical data. Not financial advice.

The variant that passed the test in our study

“Enter when the close breaks the 20-candle channel, the A/D line confirms it and the 200 EMA is on your side.”

BTC · 4h candles · 2 years

+16.1%simulated result
Trades
62
Won per $1 lost
$1.41
Largest drop
7.0%

Real BTCUSDT data, Sep 18, 2024 to Sep 18, 2026: 30% of capital per trade, 0.1% fee, 12% take profit and 6% stop loss. Simulated results on historical data. Not financial advice.

5:26See it in actionFrom an idea to a tested strategy, in 5 minutesAn idea told to Arbit in a chat, turned into a strategy and tested with real data, step by step.Watch the video
NO MAKEUP

Most configurations fail the test.

We tested 44 popular strategies across about 2,200 configurations and only one variant passed the test on BTC and ETH. Passing it does not guarantee it will win either, which is why we show you what fails. ArbitJar figures as of November 7, 2026.

simulations run
205,000+
optimizations completed
145
ended as “not ready”
59%
reached “solid”
15%
HOW IT WORKS

From one sentence to an answer.

No code and nothing to install.

  1. Write

    Tell Arbit your idea. If something important is missing, it asks.

  2. Check

    See every trade: what it won, what it lost and its worst drop.

  3. Watch

    If you are convinced, turn on the alert. It arrives when the candle closes and you decide whether to trade.

VERDICT

A good result is not enough. We look for what fails.

Test enough combinations and one will win by luck. Arbit tests them all and checks whether the result holds up before giving it a score from 0 to 100.

  • Every combination is a full backtest you can open
  • Checks whether nearly identical configurations also work
  • Shuffles the order of the trades to see how far you could have dropped
  • Keeps unused data aside to see whether the idea holds up outside the test

OPTIMIZATION · REAL RUN

46/100

Not ready yet

BTC · 1h candles · 45 combinations

  • Winning zoneOnly 20% of nearly identical configurations sustain the result.
  • Number of tradesThe winner closed 35: enough to read the metrics, not enough to trust them.
  • Order of tradesShuffled 1,000 times, the worst drop was 7.1%.

Run from July 20, 2026. Simulated results on historical data. Not financial advice.

ZONES

No idea yet? Start by reading the market.

Zones measures where the levels that matter are and how much room is left to reach them. It does not predict up or down: it helps you measure the risk before you enter.

  • Support, resistance and liquidity walls, each with a strength from 0 to 100
  • What to watch if you already hold a position, and if you do not
  • Bitcoin included for free; more assets and on-chain on paid plans
See the Bitcoin zones
BTCInside the zone
  • 80,845 – 82,339+7.4%
    Strength 100
  • 79,500 – 79,890+4.9%
    Strength 49
  • Price75,944
  • 74,315 – 77,476price is here
    Strength 91

Real Zones data from November 7, 2026. It measures risk and reach; it does not predict direction.

NO NEED TO START FROM SCRATCH

Already have strategies? Bring them.

And once an idea convinces you, Arbit stays with you.

  • Paste your TradingView Pine Script code
  • Upload a PDF, a Word file or a screenshot with up to 5 strategies
  • Describe an indicator and Arbit builds it
  • Ask Arbit why a trade lost
  • Get alerts by Telegram, email, browser or webhook
  • Learn step by step with videos of complete strategies
START FREE

Before you risk money, test the idea.

Free with no card: 50 credits a month to test on the last 60 to 90 days. Paid plans, from $20 a month, extend the history to several years.

Investing in crypto assets carries a high risk of total loss of capital. Past performance does not guarantee future results.

Frequently asked questions.

Do I need to know how to code?

Not at all. You have two ways in. You can build it step by step: pick the asset, the indicators, the entries and the risk, and Arbit helps you at each step. Or you can write the idea in your own language, the way you would explain it to another trader. If you already have TradingView Pine Script code, you can paste it too.

How does Arbit turn my idea into a strategy?

Arbit is ArbitJar's AI. You describe the strategy in your own words and, if something important is missing, such as the timeframe or the stop loss, it asks you. Then it picks the indicators, writes the entry and exit rules, runs the backtest and hands you a strategy you can see on the chart and edit.

Can I create my own indicators?

Yes. You get more than 60 ready-made indicators, such as EMA, RSI, Supertrend or Ichimoku. If yours is not there, you create it by pasting its Pine Script code or describing it in words. It is saved in your library and works just like the others in any strategy.

Can I bring strategies I already have?

Yes. Upload a PDF, a Word file, text or screenshots. Arbit detects each strategy, shows you what is missing to test it and creates them, up to 5 per import.

Does ArbitJar trade for me?

No. ArbitJar builds, tests and watches your strategy, and alerts you by Telegram, email, browser or webhook when it triggers. You decide whether to trade and on which exchange.

Are my strategies private?

Yes. Each user sees only their own. Nothing is shared or used to train models.