The Ledger · Public record
VWAP Reversion (15m)
Spec frozenspec fc61…aaec ✓ verifiedby houseForward-tested 1 days15m
Execution model · Paper trading on 15m closed candles, evaluated after the fact by the daily worker (not live or streaming): signals are evaluated on closed bars and filled per the strategy's entry price field with the same slippage assumption as backtests. The deployed spec is frozen (hash-verified); the ledger is append-only.
Original backtest (in-sample, 2016 →)
- CAGR
- -42.57%
- Total return
- -6%
- Max drawdown
- 7.6%
- Sharpe
- -3.92
Live forward test (out-of-sample, since 2026-08-05)
- Forward return
- +0.00%
- Max drawdown
- 0.0%
- Days live
- 1
- Ledger signals
- 6
Forward paper equity
Not enough forward history to chart yet — check back after a few trading days.
Open paper positions
- QQQ — 66 shares since 2026-08-05 @ 717.06
- SPY — 62 shares since 2026-08-05 @ 772.10
Signals ledger
append-only — rows can never be edited or deleted
| Date | Action | Symbol | Price | Shares | Reason |
|---|---|---|---|---|---|
| 2026-08-05T10:30:00 | Entry | QQQ | 723.73 | 66 | RULE_LONG |
| 2026-08-05T11:15:00 | Entry | SPY | 772.10 | 62 | RULE_LONG |
| 2026-08-05T16:00:00 | Stop | QQQ | 716.14 | 66 | STOP_LOSS |
| 2026-08-05T16:00:00 | Entry | QQQ | 717.55 | 66 | RULE_LONG |
| 2026-08-05T17:00:00 | Stop | QQQ | 710.02 | 66 | STOP_LOSS |
| 2026-08-05T17:15:00 | Entry | QQQ | 717.06 | 66 | RULE_LONG |
Frozen strategy rules (immutable since deployment)
{
"backtest_timeframe": "15m",
"description": "15m bars: close 0.3%+ below session_vwap with rsi_3 washed out; exit on VWAP reclaim, 1% stop.",
"entry_price_field": "close",
"entry_rule_long": "(close < session_vwap * 0.997) & (rsi_3 < 25)",
"exit_rule": "close > session_vwap",
"indicators": [
{
"length": 3,
"name": "rsi_3",
"source": "close",
"type": "rsi"
}
],
"max_holding_days": 1,
"max_positions": 2,
"name": "VWAP Reversion (15m)",
"position_size_mode": "notional_pct",
"position_size_pct": 50,
"stop_loss_pct": 1,
"symbols": [
"SPY",
"QQQ"
],
"take_profit_pct": 0
}Forward testing has no survivorship bias — signals are recorded as they happen, on the symbols the strategy actually watches. Historical simulation for research and education. Not financial advice. Past performance does not predict future results.