Backtesting Engine by Vivek Venigalla

A high-performance algorithmic trading backtesting engine built from scratch in C++. This project simulates the execution of trading strategies using historical market data with optimized memory efficiency and realistic brokerage rules.

---

Architecture Overview

• CoreTypes.hpp: Defines foundational structures (Bar, Order, Position, Trade) optimized for memory and cache management.
• CSVParser: Parses local CSV files into standard vectors in RAM, utilizing RAII to minimize memory allocations.
• Account: Serves as the primary ledger. It tracks balances and share positions using an unordered map for constant-time lookups while calculating total equity.
• Broker: Manages active orders(market, limit and stop). It utilizes an ID-keyed map to enable constant-time cancellations and handles initial order ingestion from the strategy layer.
• Strategy: Enables compile-time polymorphism so new child strategies can be developed, swapped, and managed via smart pointers

---

Directory Structure

├── include/
│   ├── structures.hpp         #Memory structures (Bar, Order, Position, Trade)
│   ├── csvParser.hpp          #CSV file processing declarations
│   ├── account.hpp            #Portfolio management declarations
│   ├── broker.hpp             #Execution simulator declarations
│   ├── strategy.hpp           #Abstract strategy base declarations
│   ├── createStrategy.hpp     #Strategy creation declarations
│   ├── performanceEval.hpp    #Performance metrics declarations
│   ├── dataFeed.hpp           #Simulation data exporrt declarations
│   └── strategies/            #Subfolder for strategy declarations
├── src/
│   ├── csvParser.cpp          #CSV file processing definitions
│   ├── account.cpp            #Portfolio management definitions
│   ├── broker.cpp             #Execution simulator definitions
│   ├── strategy.cpp           #Abstract strategy base definitions
│   ├── csv_download.py        #Market data collection script
│   ├── plotter.py             #Graph plotting script
│   ├── createStrategy.cpp     #Strategy creation script
│   ├── performanceEval.cpp    #Performance metrics definitions
│   ├── dataFeed.cpp           #Simulation data exporrt definitions
│   ├── strategies/            #Subfolder for strategy definitions
│   └── main.cpp               #Simulation runner and verification
├── data/
│   ├── AAPL_interval_set.csv  #Historical market data
│   └── test.csv               #Simulation data 
└── CMakeLists.txt             #Build system configuration

---

Current Test Framework

Currently the main.cpp serves as the unit testing for method development.

---

Data Acquisition Pipeline

Before running a backtest, historical data must be downloaded. Since yfinance includes dependencies that can alter the environment, it is necessary to create a virtual environment.

To set up and activate the environment, run the following from the project root:

python3 -m venv yf_env
source yf_env/bin/activate

To install yfinance, run the following command:

pip install yfinance

To run the script and deactivate the environment, run the following commands:

python3 src/csv_download.py
deactivate

Compilation and Execution

This project is built using CMake. Ensure you have CMake(3.10 or higher) and a modern C++17 compiler.

To compile run the following commands(this will create a new directory called build and write all build files here):

mkdir build
cd build
cmake ..

To compile the code, execute the following command:

cmake --build .

This will return a executable by the name of runny.

To run the code, execute the following command:

./runny

NOTE: This code will return a csv file in the data directory with the simulation results

Plotting

To plot any graphs using the plotter.py, make sure that matplotlib is installed. If not, activate the same environment and install the library

source yf_env/bin/activate
pip install matplotlib

Make sure that a test.csv exists and is filled with data before exeucting the script:

python3 src/plotter.py
deactivate

Will update with more information soon!