Project 3: Simple Moving Average Crossover Backtester

Build a backtester that evaluates an SMA crossover strategy.

Project Overview

Attribute Value
Difficulty Level 2: Intermediate
Time Estimate Weekend
Main Language Python
Alternative Languages R, JavaScript
Knowledge Area Backtesting
Tools CSV data, plotting
Main Book “Quantitative Trading” by Ernest Chan

What you’ll build: A simulator that generates buy/sell signals based on two moving averages and evaluates returns.

Why it teaches quant: You learn to separate signals from execution and measure strategy performance.

Core challenges you’ll face:

  • Defining signal logic cleanly
  • Preventing lookahead bias
  • Computing returns and drawdowns

Real World Outcome

You will run a backtest and get equity curves and summary metrics.

Example Output:

$ python backtest.py --symbol AAPL --fast 20 --slow 50
Total return: 18.2%
Max drawdown: -12.4%
Sharpe: 0.81
Saved equity curve to charts/AAPL_sma.png

Verification steps:

  • Confirm signals align with moving average crosses
  • Check that trades occur on the next bar

The Core Question You’re Answering

“How do I test a trading rule without accidentally cheating with future data?”

Backtesting is only useful if it avoids bias.

Concepts You Must Understand First

Stop and research these before coding:

  1. Moving averages
    • How do fast and slow averages generate signals?
    • Book Reference: “Quantitative Trading” by Ernest Chan, Ch. 4
  2. Lookahead bias
    • Why do you trade at the next bar instead of the same bar?
    • Book Reference: “Advances in Financial Machine Learning” by Marcos Lopez de Prado, Ch. 3
  3. Performance metrics
    • How do you compute drawdown and Sharpe ratio?
    • Book Reference: “Quantitative Trading” by Ernest Chan, Ch. 7

Questions to Guide Your Design

  1. Signal timing
    • When exactly does a crossover trigger a trade?
    • How will you handle missing data?
  2. Transaction costs
    • Will you include commissions or slippage?
    • How do costs change results?

Thinking Exercise

Signal Timing

Given a fast MA crossing above a slow MA at today’s close, should you buy today or tomorrow? Why?

Questions while working:

  • How does trading on the same bar create bias?
  • How do you handle market open vs close prices?

The Interview Questions They’ll Ask

Prepare to answer these:

  1. “What is lookahead bias?”
  2. “How do you compute drawdown?”
  3. “Why does Sharpe ratio matter?”
  4. “How do transaction costs affect strategy results?”
  5. “What is overfitting in backtesting?”

Hints in Layers

Hint 1: Starting Point Compute MAs and plot them over price.

Hint 2: Next Level Generate signals and shift them by one bar.

Hint 3: Technical Details Calculate returns only when in position.

Hint 4: Tools/Debugging Print trade log to verify entry/exit timing.

Books That Will Help

Topic Book Chapter
MA strategies “Quantitative Trading” by Ernest Chan Ch. 4
Bias “Advances in Financial Machine Learning” by Marcos Lopez de Prado Ch. 3
Metrics “Quantitative Trading” by Ernest Chan Ch. 7

Implementation Hints

  • Keep data aligned by index to avoid shifts.
  • Add a trade log for inspection.
  • Include transaction cost settings.

Learning Milestones

  1. First milestone: You can generate MA crossover signals.
  2. Second milestone: You can backtest without lookahead bias.
  3. Final milestone: You can interpret performance metrics correctly.