Bitcoin Daily Trend Hunter

Bitcoin Daily Trend Hunter is a daily trend-following Expert Advisor. It watches six simple setups on closed D1 bars and opens one position for every new signal, with a small stop and no profit target. Most trades lose a small, fixed amount; a few long trends have to pay for all of them. It was designed and tested for Bitcoin.

Please read this before buying

• Low win rate by design. About 30% of the trades were winners in our tests. The average win was about 3.7 times the average loss.
• Long losing streaks are normal. The longest streak in our Bitcoin tests was 27 losing trades in a row, and streaks of 10 or more losses happened 15 times between 2022 and 2026. If you stop the robot in the middle of a streak, you keep the losses and miss the trend that usually ends it.
• Bitcoin only. The same rule was negative on EURUSD, GBPUSD and AUDUSD in 2022-2026 (measured: -0.13 to -0.33 R per trade). On gold it was positive in 2022-2026, mostly thanks to the 2025 rally, but weak with deep drawdowns in older gold data. We do not recommend it for forex pairs.
• Live result so far. The first 12 live trades of this exact configuration (October 14 to November 4, 2026, a sideways Bitcoin market) were negative: 1 winner, -8.9 R in total. Our simulator, run on the same dates, predicted the same picture (12 trades, -9.8 R). In sideways markets this method loses slowly.
• Past results, in backtests or live, do not guarantee future results.

The six setups (each can be switched off)

• Donchian 8, 15, 20 and 55: the close of the bar is above the highest high (buy) or below the lowest low (sell) of the N previous bars.
• Momentum 60: the close is above (buy) or below (sell) the close of 60 bars ago.
• Pullback: the close is above the 200-bar average, the bar touched the 20-bar average and closed up (buy); the mirror for a sell.

A setup gives an event when its signal is not zero and is different from its signal on the previous bar. Only closed bars are used, so a signal never changes after it is given.

Entries

• At the open of the next D1 bar, one market position per event, in the order of the list above. Several setups can fire on the same day.
• Up to 4 positions of the robot on the symbol (input).
• Never both directions: while positions are open, only events in the same direction are taken.
• Allowed directions: buy and sell (default), buy only or sell only.
• Initial stop: 0.5 x ATR(20, Wilder) from the entry. No target.
• The order comment is the name of the setup, so every position can be traced in the history.
• When attached to a live chart, the robot does not enter on a signal that is already old (input).

Exit - evaluated once per closed D1 bar, the stop only moves in the direction of the trade:

1. When the close is at least +1 ATR in profit: the stop goes to the entry plus the spread, or to 2 ATR behind the close if that is higher.
2. When the best profit reached +1 ATR: the stop keeps at least 50% of that best profit.
3. The position closes only on its stop. There is no target and no time exit.

Position size
Each position risks 0.35% of the equity between the entry and the initial stop, rounded down to the lot step. If that lot is below the minimum lot, the minimum lot is used (as in the tested configuration); an optional cap can skip the trade instead. A fixed lot is also available. On a 10,000 USD account with Bitcoin the lot is usually 0.02.

Panel

• Every setup: current signal, last event, date and age.
• Open positions in a table: setup, side, lot, entry, stop and the current result in R (multiples of the initial risk).
• Total open risk to the stops in % of equity, and the profit already locked by the stops.
• Countdown to the next evaluation (the new D1 bar).
• Result of the robot today and in total, number of closed trades, win rate, current and longest losing streak.
• Buttons: pause new entries (open positions keep their stops) and close all positions of the robot (two clicks).

Backtest (our simulation of this exact rule, 10,000 USD account, real spread plus one spread of slippage on every stop)

• BTCUSDm, MT5 data, July 2022 to November 2026: 597 trades, +82%, maximum equity drawdown -13.7%, win rate 30%, longest losing streak 27.
• BTCUSDT, Binance data, June 2018 to June 2022 (years not used to choose the settings): 572 trades, +352%, maximum drawdown -15.8%, win rate 34%.
• BTCUSD, Bitstamp data, same out-of-sample years: 554 trades, +289%, maximum drawdown -21.9%.
• January 2025 to November 2026 only: 251 trades, +18%, maximum drawdown -11.8%.

Most of the profit came from a few strong trend periods, for example the second quarter of 2019, the last quarter of 2020 and the first and last quarters of 2024. The whole calendar year 2025 ended flat (-0.3%) and the second half of 2022 was negative (-3.8%).

Recommended use

• A Bitcoin symbol (for example BTCUSD), any chart timeframe: the signals always use D1.
• Hedging account to hold up to 4 positions. On a netting account the robot holds one position at a time.
• A VPS or a terminal that runs every day: entries happen at the start of each D1 bar.
• Keep the default risk until you have lived through a long losing streak with it.

Inputs

• Signals: timeframe (default D1), each of the six setups on or off, ATR period (20).
• Entries: allowed directions, maximum positions (4), initial stop in ATR (0.5), maximum spread in % of price (off), use an old signal when attached (off).
• Position size: risk % or fixed lot, risk per position (0.35%), minimum lot when the risk lot is smaller (on), cap for the minimum lot risk (off), maximum lot (off).
• Exit: protection start (1 ATR), trailing distance (2 ATR), share of the best profit to keep (50%), peak lock start (1 ATR).
• General: magic number (7301), panel corner and offsets.

Strategy Tester
Use a Bitcoin symbol, any chart timeframe, the model "1 minute OHLC" or "Every tick", and a period of at least one year. The tester charges the swap of your broker's symbol; some accounts are swap-free and some are not, so results can differ between brokers.

Risk warning
Trading Bitcoin with leverage is risky and you can lose money. This robot does not predict prices: it follows trends after they start and loses money while the market moves sideways. Gaps and slippage can make a loss larger than the planned risk. Only trade with money you can afford to lose.