Beat Stock Index Return: use of Technical Analysis, Machine Learning, Sentiment and Back-Testing
parthasen |
7
Description:
OBJECTIVE
An intermediate level talks to test ideas of getting excess return. Ideas are important for finding market inefficiencies. We have ideas without any system to test. My talks for them and this can be extended farther to build trading platform by interested and experienced persons. Equity Research analysts, Investors,engineering and management students will be benefited but not limited, we all have unlimited Ideas!
OVERVIEW
Technical Analysis, Machine Learning, application of tweets for sentiment analysis,strategy building and Back-Testing are important steps to follow to get excess return from stock market. Ordinary charting software are not able to do these steps but Python can perform in comparison. This language is catching faster in investment world as we find many fund houses and proprietary trading desks have made their set up already for high frequency and medium frequency trading. This language is very common to finance domain for its steep learning curve, open source, good community support etc.
OUTLINE
There would be four parts in this talk. I will use ETF (NIFTYBEES.NS) daily data for this talk. My talks would be 8 minutes for each part ( total 4 parts) in addition 8 minutes for Question and Answer session. Talks would be in brief and not to explain the coding rather by showing the results using python and completeness of python for doing complete process to test idea like Technical Analysis, Machine Learning, application of tweets for sentiment analysis,strategy building and Back-Testing.
Part A: 8 min
Introduction: EMH and frequency of data.
Downloading open source data from yahoo.
Use of pandas,numpy to read data, analysis and input-output in csv format from hard disk.
Technical analysis and plotting data using Matplotlib
Part B: 8 min
Downloading tweets and application of tweets for sentiment analysis
Application of scikit-learn for machine learning
Selection of best technique.
Regression analysis for prediction of price.
Part C: 8 min
Building simple trading strategy using MA and RSI
Back testing of strategy using PyAlgoTrade
Part D: 8 min
- Comparing compounded annual return with index return, annual volatility and draw down with NIFTY.
Questions & Answers: 8 min
Prerequisites:
- Interest in stock market and beginner knowledge of finance.
- Intermediate knowledge of Python is expected, only application will be in talks. These references are useful:
http://docs.scipy.org/doc/numpy/reference/
http://pandas.pydata.org/pandas-docs/version/0.15.2/tutorials.html
http://matplotlib.org/
http://scikit-learn.org/stable/
http://gbeced.github.io/pyalgotrade/docs/v0.17/html/tutorial.html#trading
http://review.chicagobooth.edu/magazine/winter-2013/eugene-fama-efficient-markets-and-the-nobel-prize
Content URLs:
http://www.slideshare.net/parthasen/slides-for-pycon2016
Speaker Info:
SEBI registered research analyst,founder partner of PREDICSENSE.COM, a start up for Indian Investors and analysts has 10 years of experience in stock market.
Speaker Links:
https://www.linkedin.com/in/parthasendotnet
https://twitter.com/parthasen
https://www.researchgate.net/profile/Partha_Sen/publications
http://indianjournalofcapitalmarkets.com/current_issue.php
https://www.youtube.com/watch?v=5MvmkrXVbOk