Technical Analaysis Scanner

Goals

This software attempts to provide a framework that does a combination of (1) scanning and, (2) backtesting to ask and answer such questions as,

• which instruments have responded well to the 4 hr MACD(26, 9, 12) in the last quarter?

• which stocks are getting close to their 20/50 Moving Average Crossover, and of those stocks, which have responded more than +/- 2.5% in the past?

• which instruments in the last 2 weeks after trended out of their opening ranges? What is the 1 and 2 std dev band for each isntrument at 30 minute intervals?

Features

• Pull data from IB
• Apply various indicators against single or multiple instruments
  • Simple Moving Average Crossover
  • MACD Crossover

Digging a little deeper

Technical analysis sometimes works, and sometimes doesn't. The goal of this codebase is to provide means/methods for measuring a universe of instruments and determine which ones are behaving in line with various TA patterns

It's intended to work differently than a traditional backtester (eg, a Quantopian and QuantConnect). From what I undestand about backtesting, the goal is to provide predetermined entry and exit rules, and measure the results for a single or multiple instruments. This software is didferent it that it intends to experiment with the entry and exit rules and see how those adjustments impact results.

Structure

Core Framework lives in this repo, and your secret sauce parameter/configs, research findings live in another one.

This repo will feel like a mono repo to many, and focuses on

• downloading market data (currently using IB API Gateway)
• exposing market data
• running multi-variate simulations
• reporting results

Releasing

Pypi release commands,

python setup.py sdist bdist_wheel
twine upload dist/*