🇺🇸 IBKR — US equities Phase 2 jump to Futu ↓
📈 Portfolio vs market
Indexed to 100 at the first trading day. S&P 500 proxied by SPY, Nasdaq by QQQ — topped up daily from Futu OpenD so the lines never lag a day. The final point is an intraday price while the US session is open.
Daily P&L
💼 Positions
| Symbol | Qty | Avg cost | Mark | Unrealised |
|---|---|---|---|---|
| QQQ | 120 | 703.43 | 722.50 | $2,288 |
| NVDA | 120 | 211.10 | 222.05 | $1,314 |
📊 Bot statistics
By rule: momentum_breakout × 12, mean_reversion × 40
Strategy & decision inputs
| Playbook | Entry conditions |
|---|---|
| Momentum breakout | 5-min EMA9 > EMA21, price > session VWAP, RSI14 crossing up 52, RSI < 75 |
| Mean reversion | 5-min RSI14 ≤ 32 and daily close > SMA200; entries only before 14:30 |
Both require the daily close above SMA200 (never long against the trend). Universe (5):
| Symbol | Role |
|---|---|
| QQQ | Nasdaq-100 ETF |
| SPY | S&P 500 ETF |
| NVDA | NVIDIA (AI chip leader) |
| IWM | Russell 2000 small-cap ETF |
| AAPL | Apple |
Each cycle reads: session gate (IB calendar) | NAV, cash, positions, orders today, realised P&L | 5-min indicators (price, EMA9, EMA21, VWAP, RSI14, ATR14, day high/low) | daily SMA200 and last close | AI risk review (deepseek-chat, EXECUTE/VETO only, veto needs confidence ≥ 0.7) | size = NAV × 0.5% ÷ (entry − stop), capped at 20 shares and US$20,000 | exits: stop ATR×1.5, target ATR×2.5, momentum fade, flat by 15:50 ET
🧾 Recent closed trades
| Time | Symbol | Qty | Price | P&L |
|---|---|---|---|---|
| 2026-09-18 13:10 | SPY | 20 | 759.43 | $23 |
| 2026-09-18 13:05 | IWM | 20 | 283.15 | $4 |
| 2026-09-18 11:39 | AAPL | 20 | 335.47 | $40 |
| 2026-09-18 11:29 | SPY | 20 | 759.73 | $13 |
| 2026-09-18 10:28 | AAPL | 20 | 333.75 | -$35 |
| 2026-09-18 10:13 | IWM | 20 | 283.17 | -$9 |
| 2026-09-18 10:13 | SPY | 20 | 759.83 | -$19 |
| 2026-09-18 09:47 | IWM | 20 | 284.64 | -$15 |
| 2026-09-17 14:36 | SPY | 20 | 762.78 | -$2 |
| 2026-09-17 14:36 | IWM | 20 | 286.33 | -$6 |
| 2026-09-17 14:21 | AAPL | 20 | 335.88 | -$1 |
| 2026-09-17 13:36 | AAPL | 20 | 335.71 | -$2 |
🇭🇰 Futu — HK equities Phase 2 back to IBKR ↑
📈 Portfolio vs HSI
The portfolio curve is reconstructed from order history + daily closes (起點 2026-06-08), anchored to the real total assets on the last day.
💼 Positions
| Stock | Qty | Cost | Mark | Value | P&L |
|---|---|---|---|---|---|
| 腾讯控股 HK.00700 | 100 | 427.80 | 419.00 | HK$41,900 | -HK$880 |
| SPDR金 HK.02840 | 350 | n/a* | 3,160.00 | HK$1,106,000 | n/a* |
* Broker cost basis is anomalous (negative), so cost and P&L are not shown.
📊 Bot statistics
🧠 Strategy & decision inputs
Strategy & decision inputs
| Playbook | Entry conditions |
|---|---|
| Momentum breakout | 5-min EMA9 > EMA21, price > session VWAP, RSI14 crossing up 52, RSI < 75 |
| Mean reversion | 5-min RSI14 ≤ 32 and daily close > SMA200; entries only before 14:30 |
Both require the daily close above SMA200 (never long against the trend). Universe (5):
| Symbol | Role |
|---|---|
| QQQ | Nasdaq-100 ETF |
| SPY | S&P 500 ETF |
| NVDA | NVIDIA (AI chip leader) |
| IWM | Russell 2000 small-cap ETF |
| AAPL | Apple |
Each cycle reads: session gate (IB calendar) | NAV, cash, positions, orders today, realised P&L | 5-min indicators (price, EMA9, EMA21, VWAP, RSI14, ATR14, day high/low) | daily SMA200 and last close | AI risk review (deepseek-chat, EXECUTE/VETO only, veto needs confidence ≥ 0.7) | size = NAV × 0.5% ÷ (entry − stop), capped at 20 shares and US$20,000 | exits: stop ATR×1.5, target ATR×2.5, momentum fade, flat by 15:50 ET
🧾 Recent closed trades
| Time | Symbol | Qty | Price | P&L |
|---|---|---|---|---|
| No closed trades yet — the Futu Phase 2 bot started 2026-09-17. | ||||
Portfolio history starts 2026-09-17 (Phase 2 launch); not enough yet for an HSI comparison. HSI data is already being collected.
🔍 Method & limitations
- IBKR figures are queried live from the IB Gateway (paper); Futu from OpenD (simulate account 2141908).
- **IBKR portfolio values before the latest snapshot are reconstructed** from the Phase 2 journal plus daily closes, anchored to the real NAV on the last day; real snapshots replace them over time.
- S&P 500 is proxied by SPY, Nasdaq by the Nasdaq Composite (closes from Nasdaq's public data; HSI from IB).
- No commissions beyond paper-account behaviour and no dividends are modelled, so the reconstructed curve can differ slightly from the real NAV.
- Bot statistics come from each bot's own journal (IBKR: trades_phase2.csv; Futu: trades_futu_phase2.csv) and use a different scope from account-level realised P&L.
- Both bots run on **paper (simulated) accounts** — not investment advice.
Last updated：2026-09-18（updated automatically by the daily summary crons）