XauBot/docs/arsitektur-ai/08-Take-Profit.md

Take Profit (T/P) — Sistem Pengambilan Profit Cerdas

File terkait: src/smc_polars.py, main_live.py, src/smart_risk_manager.py

---

Apa Itu Take Profit di Bot Ini?

Take Profit bukan hanya satu target harga — ini adalah sistem multi-layer yang secara cerdas memutuskan kapan mengambil profit berdasarkan momentum, probabilitas, dan peak tracking.

Analogi: TP di bot ini seperti pemanen buah pintar — tahu kapan buah sudah matang (hard TP), kapan cuaca akan buruk (momentum drop), dan kapan panen sebelum busuk (peak protection).

---

Layer Take Profit

Layer 1: Broker TP           <- Target harga dikirim ke broker (SMC-generated)
Layer 2: Hard TP             <- Software tutup jika profit >= $40
Layer 3: Momentum TP         <- Tutup jika profit bagus tapi momentum turun
Layer 4: Peak Protection     <- Tutup jika profit turun dari peak
Layer 5: Probability TP      <- Tutup jika probabilitas capai TP rendah
Layer 6: Early Exit          <- Tutup profit kecil jika reversal terdeteksi

---

Layer 1: Broker TP (SMC-Generated)

Sumber: smc_polars.py (Lines 654-659, 704-709) Dikirim ke: Broker MT5 sebagai TP order aktif

Perhitungan

# ATR-based TP cap
atr = latest["atr"]                    # Contoh: ATR = $8.50
max_tp_distance = 4.0 * atr            # 4 * 8.50 = $34.00

# Untuk BUY:
risk = entry - sl                      # $4950 - $4937 = $13
tp   = entry + (risk * 2)              # $4950 + $26 = $4976 (2:1 RR)
if tp > entry + max_tp_distance:       # $4976 vs $4950 + $34 = $4984
    tp = entry + max_tp_distance       # Tidak kena cap, tetap $4976

# Untuk SELL:
risk = sl - entry                      # $4963 - $4950 = $13
tp   = entry - (risk * 2)              # $4950 - $26 = $4924 (2:1 RR)
if tp < entry - max_tp_distance:       # $4924 vs $4950 - $34 = $4916
    tp = entry - max_tp_distance       # Tidak kena cap, tetap $4924

Kenapa TP Di-cap 4 ATR?

Sebelum (v2): TP = risk * 2 (tanpa batas)
  -> Bisa sangat jauh ($50+ dari entry)
  -> Jarang tercapai, posisi terbuka terlalu lama

Sesudah (v3): TP = MIN(risk * 2, 4 * ATR)
  -> Dibatasi maksimal 4x ATR
  -> Target lebih realistis, lebih sering tercapai

Dikirim ke Broker

# main_live.py
result = mt5.send_order(
    sl=broker_sl,
    tp=signal.take_profit,    # <- TP dari SMC (ATR-capped)
    ...
)

Jika harga mencapai TP level, broker otomatis menutup posisi — tidak perlu bot online.

---

Layer 2: Hard Take Profit ($40)

Sumber: smart_risk_manager.py (Lines 595-599)

# Profit mencapai $40+ -> langsung tutup
if current_profit >= 40:
    return True, ExitReason.TAKE_PROFIT,
           "[TP] Target profit reached: $40.00"

Kenapa $40? Ini threshold profit yang cukup besar untuk diamankan, terlepas dari kondisi pasar.

---

Layer 3: Momentum-Based TP ($25+)

Sumber: smart_risk_manager.py (Lines 601-603)

# Profit $25+ tapi momentum turun -> amankan profit
if current_profit >= 25 and momentum < -30:
    return True, ExitReason.TAKE_PROFIT,
           "[SECURE] Securing $25.00 (momentum dropping)"

Bagaimana Momentum Dihitung

# PositionGuard.calculate_momentum() (Lines 113-131)
# Melihat 5 profit history terakhir

recent_profits = profit_history[-5:]
profit_change = recent_profits[-1] - recent_profits[0]

# Normalisasi: $10 change = 50 poin
momentum = (profit_change / 10) * 50
# Range: -100 sampai +100

# momentum < -30 artinya profit sedang TURUN cukup cepat

Visualisasi:

Profit ($)
  40 |
  35 |        /\
  30 |       /  \  <- Momentum mulai negatif
  25 |------/----\------ Layer 3 trigger: amankan!
  20 |     /      \
  15 |    /        \
  10 |   /          \
   5 |  /
   0 |_/________________________> waktu

---

Layer 4: Peak Protection ($30+ peak)

Sumber: smart_risk_manager.py (Lines 605-607)

# Profit pernah $30+ tapi sekarang turun ke 60% dari peak
if guard.peak_profit > 30 and current_profit < guard.peak_profit * 0.6:
    return True, ExitReason.TAKE_PROFIT,
           "[LOCK] Securing profit (was $35 peak)"

Cara Kerja Peak Tracking

# Setiap evaluasi, update peak profit
guard.peak_profit = max(guard.peak_profit, current_profit)

# Contoh:
# Peak: $35 -> 60% = $21
# Current: $18 (turun dari $35)
# $18 < $21 -> TUTUP, lindungi sisa profit

Visualisasi:

Profit ($)
  35 |    * <- peak_profit = $35
  30 |   / \
  25 |  /   \
  21 |./.....\....... 60% threshold ($21)
  18 |        \* <- current = $18, TUTUP!
  15 |         \
  10 |          (kehilangan lebih banyak dihindari)

---

Layer 5: Probability-Based TP ($20+)

Sumber: smart_risk_manager.py (Lines 609-611)

# Probabilitas capai TP rendah + profit cukup -> ambil sekarang
if tp_probability < 25 and current_profit >= 20:
    return True, ExitReason.TAKE_PROFIT,
           "[PROB] Taking profit $20 (TP prob: 15%)"

Cara Hitung TP Probability

# PositionGuard.get_tp_probability() (Lines 133-168)
# Score 0-100% berdasarkan 4 faktor:

Factor 1: Progress ke TP (0-40 poin)
  progress = (current_profit / target_tp_profit) * 100
  -> Makin dekat ke TP = skor tinggi

Factor 2: Momentum (0-30 poin)
  -> Momentum positif = skor tinggi

Factor 3: ML Confidence Trend (0-20 poin)
  -> ML confidence naik = skor tinggi

Factor 4: Time Penalty (0-10 poin DIKURANGI)
  -> 2 poin per jam (makin lama = makin rendah)

probability = factor1 + factor2 + factor3 - time_penalty

---

Layer 6: Early Exit (Profit Kecil + Reversal)

Sumber: smart_risk_manager.py (Lines 617-627)

# Profit $5-$15 + momentum sangat buruk + ML reversal
if 5 <= current_profit < 15:
    if momentum < -50 and ml_confidence >= 0.65:
        if ml_signal berlawanan dengan posisi:
            return True, ExitReason.TAKE_PROFIT,
                   "Early exit - reversal detected"

Logika: Lebih baik ambil profit kecil ($5-$15) daripada menunggu profit hilang karena reversal.

---

Prioritas Exit (Urutan Pengecekan)

1. Hard TP ($40+)              <- Paling prioritas
2. Momentum TP ($25+, mom<-30)
3. Peak Protection ($30+ peak, <60%)
4. Probability TP ($20+, prob<25%)
5. Early Exit ($5-15, reversal)
6. Broker TP (harga hit level)  <- Independen dari software

Catatan: Broker TP berjalan independen — jika harga hit TP level di broker, posisi tertutup otomatis meskipun bot offline.

---

Contoh Skenario

Skenario 1: TP Broker Hit

Entry BUY @ $4950, TP broker @ $4976
  -> Harga naik ke $4976
  -> BROKER TP HIT -> Tutup otomatis
  -> Profit: ~$26 (0.01 lot = $2.60)

Skenario 2: Software TP Lebih Cepat

Entry BUY @ $4950, TP broker @ $4990
  -> Harga naik ke $4990 (profit $40)
  -> Software: profit >= $40 -> HARD TP
  -> Tutup sebelum broker TP level

Skenario 3: Momentum Drop

Entry BUY @ $4950
  -> Profit naik: $10 -> $20 -> $28 -> $25
  -> momentum = -35 (turun)
  -> Software: profit $25 + momentum < -30
  -> MOMENTUM TP: amankan $25

Skenario 4: Peak Protection

Entry BUY @ $4950
  -> Profit naik: $15 -> $25 -> $35 (peak!)
  -> Profit turun: $35 -> $30 -> $22 -> $19
  -> 60% dari $35 = $21
  -> $19 < $21 -> PEAK PROTECTION: amankan $19
  (tanpa ini, profit bisa turun ke $0 atau bahkan loss)

---

Tabel Ringkasan Layer TP

Layer	Trigger	Profit Min	Kondisi Tambahan
1. Broker TP	Harga hit level	-	Otomatis, independen
2. Hard TP	profit >= $40	$40	Tidak ada
3. Momentum TP	profit >= $25	$25	momentum < -30
4. Peak Protection	peak > $30	~$18+	current < 60% peak
5. Probability TP	profit >= $20	$20	TP probability < 25%
6. Early Exit	profit $5-15	$5	ML reversal + momentum < -50