Backtest Quantitative Trading strategies from Scratch

Name: Backtest Quantitative Trading strategies from Scratch
Rating: 4.5 (13 reviews)

Building your own backtesters with a 20 year data set and large scale simulation studies.

Created byCemal Arican

Last updated 3/2023

English

English [Auto],

What you'll learn

Cointegration: Engle and Granger Approach
Creating your own backtesters in python
Covering Object Orientated Programming
Python packages: Pandas, Numpy, Scikit, Joblib, Matplotlib
Time Series
Quantitative trading
Moving Average Cross Over
Pairs trading
Stock Pipeline

Course content

6 sections • 33 lectures • 10h 21m total length

Introduction2:58

Part 1: Introduction to the Cross Over strategy and a worked example10:53
Part 2: Coding the Backtester - the Cross Over class16:08
Setting up the class for Cross Over backtester and setting up the constructor method.
Part 3: Coding Backtester part 2- Handle Data35:09
Second part of the coding of the backtester. Creating the Handle data method. This method is designed to create the signals, and obtain the profits determined from each position created.
Part 4: Coding the Backtester part 3 - Backtest22:24
Third part of the coding of the backtester. Creating the Backtester mathod that recursively updates the necessary attributes to for the eventual final part of getting the results
Part 5: Coding the Backtester part 4 - Get results28:30
Fourth and final lecture on the coding of the backtester class. Code provided in the resources.
Part 6: TSLA example33:30
Implementing the Cross Over strategy on a single ticker: Tesla.

Some useful links:

- https://stackoverflow.com/questions/21971449/how-do-i-increase-the-cell-width-of-the-jupyter-ipython-notebook-in-my-browser
- https://www.geeksforgeeks.org/python-sys-module/
Part 7: Long term study on Russell 3000 Index vs SPY37:16
Conduct a long-term backtest study of the Russell 3000 index versus SPY, exploring moving averages and exponential moving averages within a robust simulation pipeline and data quality checks.
Part 8:Long term study part 25:01
Backtest a long term crossover strategy on an index and compare it with the S&P 500 ETF (SPY) over about two and a half years, showing 92% return.
Part 9: High vs Low volatile stocks in the Cross Over part 136:47
Part 10: High vs Low volatile stocks part 222:20
Run and compare a simulation study of high volatility stocks versus low volatility stocks, assessing two-year returns against a buy-and-hold S&P 500 benchmark and visualizing distributions with histograms.
Part 11: Discussion topics0:04

Introduction to Pairs trading11:15
Hello everyone. Welcome to the last section of this course and to the introduction of this section. Please go through the reader and the referenced paper, it is a good guide where to start and how to things work in the Pairs trading literature. Also dont forget to add PART 3 to the already existing directories we have so far. Please make sure that all your directories are in order. Ofcourse, this is advice, if you have done things differently that is up to you :). Hope you enjoy, and please give me as much feed back as possible and lets start more discussions in the comment sections in any findings that you may have.
Set Up part 126:33
Set up data pipeline to backtest quantitative trading strategies by building a stock universe, filtering by high correlations, and using formation period with a squared distance metric to evaluate pairs.
Set Up part 217:14
Implement augmented filters with the ADF approach on a large stock-data dataset, check unit roots, model y with X, compute residuals, and identify stationary relationships via P-values and SSD.
Worked Example15:01
Although you cant use the code of the Backtester, you have it incase you want to use it later.
Orders class part 122:21
define the orders class to track positions for each security, allocate capital via a target percentage, and support long or short entries with entry price for backtesting.
Orders class part 219:12
Develop and backtest an orders class, managing long and short positions, first trades, capital allocation, closing positions, price points, and profit calculations for quantitative trading.
Backtester class part 119:51
Build the backtester package by inheriting from the oldest class to create an empty canvas for any trading strategy, implementing data handling, signals, and results.
Backtester class part 27:04
Develop data handling in the backtester by implementing exponential moving averages and price data derivations, and compute the spread and z-score with normalization for signals and results.
Backtester class part 329:20
Backtester class part 431:20
I noticed during the filming I made some small mistakes, I have added the corrected versions in the resources.
Large scale simulation part 131:20
Large scale simulation part 242:16
Large scale simulation part 37:13

Pairs trading and Holding SPY11:06
Crossing Over13:52
Backtest quantitative trading strategies from scratch by exploring crossing over with short term and long term trends, moving day parameters, and signals in a notebook.
Corrections0:10
Implementations22:58
Pairs Trading0:27
Once you have completed this Task send me your results to cemal.arican@gmx.com.

Remember this is a real assignment at a large trading institution. Hint: this outside the box, and improvise outside the classical pairs trading strategy. For example, rather a fixed standard deviation from mean, why not consider moving averages. From my own experience, short or long term averages work best i.e. 15 or 30 days, or, 125 to 252 days.

See: https://www.researchgate.net/publication/343981808_Bachelor_Thesis_-_Determining_Profitable_Pairs_Trading_Strategies_under_Cointegration

Requirements

A curious mind will do. However, you need to be familiar with some of the elementary Python operations and OOP
Some background in mathematics or econometrics

Description

The essence of this course is a 'from the group up movement' type of course. You will be given a very large dataset with over 30 million rows of data of all tradable equities on the US stock markets from 2001 up until October 2021 - at daily intervals.

Having having said that, we will show how to build your own backtester step by step and apply it to two well known trading algorithms: Moving Average Cross Over strategy and Pairs trading. With a brief reader on time series is also provided in order to help understand some of the mathematical concepts behind pairs trading such as pairs unit root and cointegration.

Besides the implementation of the algorithms, we also look at large scale implementations of the algorithms using a pipeline which allows you to create a stock universe. Which is a class we will go over step by step as well. Moreover, we also look at

Finally we end this section with a take home assignment that is a real life example of an assignment at a trading firm using high frequency data, where data sort and cleaning has to be implemented, determining a cointegrated relationship and determining how you would trade these two instruments.

Who this course is for:

Beginner traders, students and professionals who are interested in their own trading ideas or doing research.

Backtest Quantitative Trading strategies from Scratch

What you'll learn

Explore related topics

Course content

Introduction1 lecture • 3min

Data1 lecture • 4min

The pipeline class2 lectures • 38min

Moving Average Cross Over11 lectures • 4hr 8min

Pairs Trading13 lectures • 4hr 40min

Multi Testing5 lectures • 49min

Requirements

Description

Who this course is for: