Paper trading track record
A paper-traded portfolio that follows the algorithm’s entry signals — no real money, fully public scoreboard.
Real entry and exit prices, real holding periods, real outcomes — wins and losses both shown. This is not a backtest. How it works →
How this is calculated
We pretend to invest $100,000. Each trade risks up to 2% of that — the position size is set so a 1.5× ATR move against us would lose roughly $2,000, and no single position is larger than 20% of the book ($20,000). For a low-volatility name, that cap can bind first, so the trade risks less than $2,000.
We open a position when the regime classifier moves an asset to CLEARANCE at a session close, with a conviction score of at least 65 (out of 100).
We close a position when any of these happens: the regime turns against us (the asset moves to a caution column), the trailing stop is hit, one really bad day takes the price down further than normal volatility should allow, or the entry conditions no longer hold.
The stop works in two stages. A new position starts on a tight stop 1.5× ATR below the entry price — a trade that never gets going is cut quickly and cheaply. Once it has gained enough to cover that initial risk, the stop converts to a wider 4× ATR trailing below the highest price we’ve seen, so a trade that is working is given room a trade that never started is not. Either way the stop only ever moves up, never back down, and it tightens to 1× ATR if the regime escalates to ELEVATED.
We deduct 0.1% from each side of every trade to mimic real-world slippage. There are no commissions modeled — most retail brokers are commission-free for US equities, but real spreads and execution costs would be slightly higher than what we apply here.
Evaluated at each session close: EU 16:50 UTC, US 21:20 UTC, crypto 10:30 UTC. Long-only. Forward-tracked since Apr 6.