Panduan Backtesting

Pengenalan

Backtesting adalah proses menguji trading strategy anda menggunakan data historical. Sebelum risk real money, test dulu!

"Would you drive a car without testing it first? Then why trade a strategy without testing?"

---

Apa Itu Backtesting?

Definisi

Backtesting = Applying your trading rules to historical data to see how they would have performed.

Why Backtest?

---

Types of Backtesting

1. Manual Backtesting

Process:

• Scroll back in charts
• Find setups manually
• Record hypothetical trades
• Calculate statistics

Pros:

• Free
• Learn price action deeply
• Build intuition

Cons:

• Time-consuming
• Prone to bias
• Can miss setups

2. Automated Backtesting

Process:

• Code your strategy
• Run on historical data
• Get instant results

Pros:

• Fast
• Objective
• Large sample sizes

Cons:

• Need coding skill
• Requires quality data
• May over-optimize

For Most FCPO Traders: Manual

Start with manual backtesting:

• No coding needed
• Learn market behavior
• Affordable (free)
• Good enough for most strategies

---

Manual Backtesting Process

Step 1: Define Your Strategy

Write down clear rules:

Strategy: RBS Entry
Timeframe: 1H
Entry: Long when price retests broken resistance
Confirmation: Bullish candle at level
SL: Below the level (15-20 ticks)
TP: Next resistance or 30 ticks

Step 2: Choose Timeframe for Test

Recommended:

• Minimum 3 months of data
• Better: 6-12 months
• Must include different market conditions (trending, ranging)

Step 3: Set Up Your Recording

Create spreadsheet with:

Step 4: Scroll and Find Setups

Process:

1. Go to starting date
2. Scroll forward candle by candle
3. When setup appears, record it
4. Note if SL or TP hit
5. Continue to end

IMPORTANT: Don't peek ahead! Only trade what you would have seen in real-time.

Step 5: Calculate Statistics

After completing all trades:

Total trades: ___
Winners: ___
Losers: ___
Win rate: ____%
Total ticks: ___
Average win: ___ ticks
Average loss: ___ ticks
Largest win: ___ ticks
Largest loss: ___ ticks
Max consecutive losses: ___
Profit factor: Total wins / Total losses

---

Backtesting Example

Strategy: 20 EMA Bounce

Rules:

• Timeframe: 1H
• Entry: Bounce off 20 EMA in uptrend
• Confirmation: Bullish candle at EMA
• SL: Below EMA (15 ticks)
• TP: 30 ticks

Backtest Period: October 2025

Sample trades:

Results Analysis

Total trades: 10
Winners: 6
Losers: 4
Win rate: 60%

Total wins: 6 × 30 = +180 ticks
Total losses: 4 × 15 = -60 ticks
Net profit: +120 ticks

Average win: 30 ticks
Average loss: 15 ticks
RRR: 2:1

Profit factor: 180/60 = 3.0
Max consecutive losses: 2

CONCLUSION: Strategy is profitable in backtest ✓

---

Backtest Metrics to Track

Essential Metrics

Good Benchmarks

Red Flags

Don't trade strategy if:

• Win rate below 40%
• Profit factor below 1.0
• Max consecutive losses > 7
• Drawdown > 20% of capital

---

Common Backtesting Mistakes

Mistake 1: Hindsight Bias

Problem: "I would have known to skip that trade"

Reality: You wouldn't have known in real-time

Solution: Trade EVERY setup that meets criteria, don't cherry-pick

Mistake 2: Too Few Trades

Problem: "I tested 10 trades, looks good!"

Reality: 10 trades is not statistically significant

Solution: Minimum 50 trades, ideally 100+

Mistake 3: Curve Fitting

Problem: Adjusting rules to fit historical data perfectly

Reality: Over-optimized strategy fails in live trading

Solution: Keep rules simple, test on unseen data after

Mistake 4: Ignoring Costs

Problem: Backtest shows profit, but doesn't include commission

Reality: Commission eats into profit

Solution: Deduct RM 20 per trade from results

Mistake 5: Only Testing Good Markets

Problem: Only backtest trending periods

Reality: Market also ranges and chops

Solution: Include all market conditions in test period

---

Backtest to Live Trading

After Successful Backtest

Don't jump to live immediately!

Process:

1. Backtest (3-6 months data) ✓
2. Forward test (1-2 months paper trading)
3. Small size live (50% normal size)
4. Normal size live

Forward Testing

What is it: Paper trading the strategy in real-time

Why: Confirms backtest results hold in current market

Duration: Minimum 1 month, 20+ trades

Transition Checklist

Before going live:

• Backtest shows profit over 50+ trades?
• Forward test confirms results?
• Commission included in calculations?
• Comfortable with max drawdown?
• Understand max consecutive losses?
• Trading plan written based on results?

---

Backtest Recording Template

Trade Log

Date: ___________
Setup: ___________
Entry: ___________
SL: ___________
TP: ___________
Result: Win / Loss
P&L: ___________ ticks
Notes: ___________

Summary Statistics

PERIOD: _____ to _____

Total Trades: _____
Winners: _____
Losers: _____
Win Rate: _____%

Total Winning Ticks: _____
Total Losing Ticks: _____
Net Ticks: _____

Average Win: _____ ticks
Average Loss: _____ ticks
Actual RRR: _____

Profit Factor: _____
Max Consecutive Wins: _____
Max Consecutive Losses: _____
Max Drawdown: _____ ticks

PASS / FAIL: _____

---

Soalan Lazim (FAQ)

Q: "Berapa lama nak backtest?"

A: Minimum 3 months of data, 50+ trades. Better to have 6-12 months covering different market conditions.

Q: "Platform mana untuk backtest?"

A: TradingView (free) is excellent. You can scroll back historical FCPO data and manually mark trades.

Q: "Backtest profitable tapi live trading loss. Kenapa?"

A: Common reasons:

• Hindsight bias in backtest
• Emotions in live trading
• Execution differences
• Market conditions changed

Solution: More rigorous forward testing

Q: "Perlu coding untuk backtest?"

A: No. Manual backtesting is enough for most discretionary strategies. Only need coding for fully automated strategies.

---

Kesimpulan

Key Takeaways

Your Action Items

1. Write your strategy rules clearly
2. Set up backtest spreadsheet
3. Test minimum 3 months data
4. Calculate statistics
5. Forward test before live trading

---

"Every losing trader wishes they backtested. Every winning trader is glad they did. Which will you be?"