For each candidate config, we walk through the full 540-day price series and deploy a 30-day grid only when a regime filter passes (7-day annualized vol < 0.70 AND |drift| < 3%). Capital compounds across deployments. Cash sits idle when filter fails.
| Asset | Range% | Grids | SL | Final ($) | Strategy % | BnH % | Excess % | Deploys | Skips | Win% | Worst |
|---|
| # | Bar Range | Start $ | End $ | BnH % | Strategy % | Capital After | Vol (1w) | Drift | Trades |
|---|
Ran 255,000 backtests across 600 parameter combinations and 126 scenarios (real BTC/ETH price paths + synthetic regimes). A configuration is robust if it achieves positive mean return across all scenario types, max drawdown <20%, beats buy-and-hold in ≥60% of scenarios, coefficient of variation ≤2.0, and stop-out rate <50%.
| Asset | Range% | Grids | Stop-loss | Recenter | Mean Return% | P10% | P90% | Max DD | Beats BnH% | Stop-out% | Score | Status |
|---|
Range=sideways market, Bull=slow uptrend, Bear=slow downtrend, Crash=sudden -25%, Pump=sudden +30%, Choppy=high-vol alternating
Lower stop-loss % = tighter stop. Note: tighter stops increase stop-out rate and often reduce mean return because stops can trigger before price recovers.