DCA Backtesting Framework

A Python framework to backtest Dollar Cost Averaging (DCA) strategies on historical stock data. Visualize how your portfolio would have performed through market crashes, downtrends, and recoveries.

Why I Built This

DCA is often presented as the holy grail of passive investing:

• Stock going up? Great, you're making money.
• Stock going down? Great, you're buying cheaper!

But it feels wrong watching the position turn increasingly red and continue to pour money in. This framework helps to test and visualise DCA with historical data.

Sample Scenarios

The framework includes pre-configured scenarios on real market events, like:

Stock	Scenario	What Happened
LVMH	2022–now	Peak at €900 → crash to €435 → recovery to ~€600. Big fall with some intermediate recoveries.
Pernod Ricard	2023–now	Steady decline from €217 to €73. A "falling knife" scenario.
S&P 500	1995–2010	Dot-com bubble, subprime crisis. Two major crashes in 15 years.

Features

• 3 DCA strategies — fractional shares, fixed whole shares, or whole shares with cash buffer
• Dividend tracking — accounts for dividends received during the investment period
• Transaction fees — includes broker fees in performance calculations
• Visual analysis — plots the average cost (PRU) vs market price over time

Installation

# Create and activate virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

Quick Start

# List all available scenarios
python main.py --list

# Run a specific scenario
python main.py lvmh_down_fractional

Available Strategies

Strategy	Description	Use Case
fractional	Invest a fixed € amount, buy fractional shares	Modern brokers
whole_fixed	Buy a fixed number of whole shares	Traditional brokers
whole_buffer	Invest fixed € amount, buy whole shares only, keep remainder as buffer	Traditional brokers

Creating Custom Scenarios

Edit scenarios.py to add your own:

"my_scenario": Scenario(
    name="My Custom Scenario",
    start_date=datetime(2020, 1, 1),
    end_date=datetime(2024, 12, 31),
    transaction_fees=1.5,          # € per transaction
    strategy="fractional",         # or "whole_fixed", "whole_buffer"
    config_params={
        "amount_to_invest": 200    # € per month (or "nb_shares_to_buy" for whole_fixed)
    },
    dataset_path="data/MY_STOCK.csv",
),

Using Your Own Data

Place a CSV file in the data/ folder with at least these columns:

Column	Description
time	Date in YYYY-MM-DD format
close	Closing price
Dividends per share (common stock primary issue)	Optional: dividend amount

Tip: I'm using TradingView csv exports. If you have a subscription, your can simply copy paste their CSV inside data/.

Output

The framework calculates and displays:

• Total invested — including transaction fees
• P&L — profit/loss in €
• Return % — percentage return on investment
• Chart — your average cost vs market price over time

License

MIT