| Strategy | Direction | Lots | Entry | Stop-Loss | Take-Profit | Live Profit/Loss |
|---|---|---|---|---|---|---|
| loading… | ||||||
One proven method runs the whole basket: a Donchian breakout. We draw a "box" around the highest high and lowest low of the last N candles, then wait for price to break out of it. Every trade has a hard stop and a fixed target before it is placed — nothing is left open-ended.
Price closes above the box → momentum is breaking up. We enter long at the box high.
Price closes below the box → momentum is breaking down. We enter short at the box low.
Set ATR × 1.5–2 away from entry and sent to the broker. Caps the loss on every trade — it survives even if the bot is off.
Set at 1.5–2× the risk. Winners are held multi-day to reach it. A few big wins beat many small losses — that's the edge.
Important — Gold Daily uses a 96-day box (~4½ months), not 16 days. The 16-day channel you may have seen is the Gold 4-Hour strategy (96 four-hour candles = 16 days). A 96-day box is very wide, so a fresh breakout is rare — about 6 trades a year. A card can read "BUY" yet be tagged stale: that means the breakout already happened days ago and we're already riding it, so no new order is placed.
| Strategy | Settings | Profit Factor | Win rate | Trades/yr | Tested |
|---|---|---|---|---|---|
| Gold Daily | 96-day · ATR1.5 · RR1.5 | 2.5 | ~60% | ~6 | 22 yrs |
| Gold 4-Hour | 96-bar (16-day) · ATR1.5 · RR2.0 | 1.33 | ~40% | ~12 | 22 yrs |
| Silver 4-Hour | 20-bar · ATR2.0 · RR2.0 | 1.59 | ~40% | ~45 | 19 yrs |
| USD/JPY Daily | 48-day · ATR2.0 · RR2.0 | 1.44 | ~39% | ~2 | 37 yrs |
| S&P 500 Daily | 20-day · ATR2.0 · RR1.5 | 1.61 | ~51% | ~3 | 15 yrs |
| Gold 5-EMA Long · NEW #6 | H1 indicator / M5 entry · Stop-Loss = indicator low · Risk-Reward 2:1 · London-NY only | 1.75 | ~43% | ~60 | 22 mo (demo proof) |
Strategy #6 — Gold 5-EMA Long is the new addition discovered via a 900-combination grid. It looks for an H1 candle that closes entirely below the 5 EMA (the "indicator candle"), then enters long when a 5-minute candle closes above that indicator candle's high — but only between broker 13:00–17:00 (London-NY overlap, ~18:30–22:30 IST). Stop-Loss = indicator low; Take-Profit = 2× the risk (Risk-Reward 2:1). Long-only. Magic 50001. Lot size = 0.10 (10 oz) like the rest of the basket.
Profit Factor = total winnings ÷ total losses. Above 1.0 is profitable; we require ≥ 1.25 and profitability across most years.
Trends powerfully on safe-haven demand, real yields and central-bank buying. Our strongest edge — traded on both the Daily and the 4-Hour.
Moves with gold but in bigger, cleaner swings. Trends on the same macro flows; traded on the 4-Hour.
The one currency that genuinely trends — driven by the carry trade and Bank-of-Japan policy, which create long, ridable moves.
Trends upward structurally as the economy and company earnings grow. A long-biased breakout catches the sustained rallies.
We put 15+ markets through the same honest gate. These failed — each for a specific reason a breakout system can't overcome. Trading them would lose money, so we leave them out. Being selective IS the strategy.
Driven by OPEC headlines, inventory reports and geopolitics. It spikes then snaps back, so a break above the box is usually a fake-out that reverses the next day. Profit Factor stayed under 1.0.
The most violent major commodity — weather-driven, with 10–20% swings in days. Stops get hit constantly and no stable trend ever forms for the box to capture.
It does trend, but MetaTrader carries only a few years of price history. Too little data to trust — a good backtest here is likely a recent-regime artifact, the exact trap we avoid.
The most traded, most efficient markets on earth. Any breakout is arbitraged away within seconds and price mean-reverts back into the box — breakouts fail more often than they work.
Commodity currencies that chop sideways inside ranges for months. The box gets pierced and reversed over and over — many small losses, with no trend to pay for them.
Heavily managed by the Swiss National Bank, which caps and pins the rate. Central-bank intervention smothers the clean trends a breakout needs.
An honest graveyard of everything we tested and rejected, instrument by instrument — what we tried, why it lost, the likely reason, and how to improve if we ever revisit. The point: next time we restart research, we don't re-walk dead ends — we pick up from here. "Tested" = full out-of-sample backtest on the local Data Bank. Profit Factor needs to be ≥ 1.25 and profitable in most years to pass the gate.
Desk view on restarting: the biggest un-tried upside is not new intraday systems — it's better exits (trailing stops / pyramiding) and portfolio sizing on the working positional engine, plus screening new instruments through the same gate. Finish the live demo proof first.