PS-I station: William O Neil , Bengaluru
Student
Name: SAKSHAM BAJAJ . (2021A7PS1315H)
Student Write-up:
PS-I Project Title: Implementing various trading strategy using different indicators as
market signals
Short Summary of work done: Initially we were supposed to implement a moving
average crossover monthly rebalancing strategy using backtrader library in python. And
then we were supposed to implement another strategy similarly which used weekly COT
reports as market signals to take relevant market positions.
Objectives of the project: To use COT reports to implement a weekly rebalancing
trading strategy on our portfolio
Tool used: Python, Backtrader, CVXOPT
Details of Papers/patents:nil
Brief description of the working environment: We were expected to work in a
coordinated manner with our team and be timely with our deadlines. We received
adequate help from our industry mentors with periodic inputs about our work, putting us
in the right direction.
Academic courses relevant to the project: nil
Learning Outcome: Understanding the working of stock market, and how to use
different indicators and their essence.
------------
PS-I station: William O Neil , Bengaluru
Student
238
Name: BHANDARI SOURABH RAHUL . (2021A7PS2412P)
Student Write-up:
PS-I Project Title: Machine Learning-Based Strategy using FRED Economic Indicators
Short Summary of work done: Achieved 85% accurate recession prediction using
Random Forest classifier, based on FRED economic indicators, used this model to
develop a successful trading strategy, and further optimized it using Mean-Variance
Optimization Applied Deep Reinforcement Learning based algorithm Advantage Actor
Critic (A2C) for developing an intelligent stock trading agent, which uses FRED economic
indicators to make decisions, trained the agent on a self made trading environment,
achieving a CAGR value of 31.5%
Objectives of the project: William O' Neil India specializes in stock market data
research, implementing algorithmic trading strategies, and performing quantitative
analysis
Tool used: Skills learnt- Python (Backtrader, Quantstats, Gym, Pytorch, Pandas), Data
Analysis (Correlation, Covariance), Finance
Details of Papers/patents:NA
Brief description of the working environment: It was a very good learning project. We
were assigned a research problem and made to work in teams of two. Liked the work a
lot. The work culture at the company was pretty cool. Will definitely recommend the PS
site to the juniors
Academic courses relevant to the project: Principles of Economics, Ecotrix, Computer
Programming
Learning Outcome: An excellent start for Quantitative Research
------------
PS-I station: William O Neil , Bengaluru
Student
Name: NISHIL ASHUTOSH KULKARNI (2021AAPS2499G)
239
Student Write-up:
PS-I Project Title: Backtesting Trading Strategy designed on COT Reports
Short Summary of work done: Researched what data COT reports provide, and how
this can be used to generate a trading signal. Used python package backtrader to code
the strategy. Worked on optimizing the strategy by tuning various parameters.
Implemented portfolio allocation among different asset classes using Mean Variance
Optimisation, coded in Python using library cvxopt. Generated detailed reports of the
strategy execution to serve as a guide for investors.
Objectives of the project: Design, code, optimize and backtest a trading strategy based
on Commitment Of Traders (COT) reports published by the USFTC every week. The
portfolio must be composed of multiple asset classes, must generate high returns to
investors, and be practically executable.
Tool used: Jupyter Notebooks, VSCode
Details of Papers/patents:None
Brief description of the working environment: This PS was online. Company expects
us to research a lot on the given work, adhere to the timelines and ask relevant questions.
New ways of approaching any financial situation is appreciated. Company also expects
us to be thoroughly versed with all the nuances of the strategy that is designed. I improved
on my data processing, algorithm design and visualisation skills. Apart from this, I
improved on my communication and presentation skills. I also learnt how work division is
to be done on any project.
Academic courses relevant to the project: CP, OOP, SAPM
Learning Outcome: Python, Data wrangling, Portfolio Theory, Derivatives trading
------------
PS-I station: William O Neil , Bengaluru
Student
Name: JAIDITYA SINGH . (2021B3A71147P)
Student Write-up:
240
PS-I Project Title: Algorithm Trading and Trading Strategies
Short Summary of work done: Created two trading strategies in python using the
backtrader library primarily. First one used the concept of moving average crossover in
order to buy or sell and used Mean Variance Optimization for weight allocation. The
second project used COT reports to generate signals for buying and shorting. We used
MVO and Inverse Volatility for weight allocation.
Objectives of the project: Two Projects were given. First project focused on creating a
Moving Average Cross over strategy. Creating and back testing it. The second was to
develop a trading strategy using COT reports for signal generation.
Tool used: Python majorly
Details of Papers/patents:none
Brief description of the working environment: Gives a good insight into the world of
Quantitative Trading and what is it all about. Helps in learning the use of various financial
concepts in the real world. Makes you more knowledgeable about the security exchange
market as a whole.
Academic courses relevant to the project: Derivative and Risk Management
Learning Outcome: 1) Use of various financial instruments.
2) Portfolio management
3) Risk reduction and mitigation
4) Pandas, Numpy, Backtrader, cvxopt libraries in python.
5) Usage and cleaning of large amounts of data.
------------
PS-I station: William O Neil , Bengaluru
Student
Name: TANMAY AGARWAL . (2021B3A73040H)
Student Write-up:
PS-I Project Title: Portfolio Optimization Strategy
241
Short Summary of work done: We devised a signal generation strategy centered
around yield curve inversion. To construct our portfolio, we incorporated data from key
benchmark indices like S&P 500, NASDAQ 100, and NASDAQ Composite Index. The
strategy involved monitoring the 10Y-2Y yield spread data, which acted as an indicator of
an impending recession, prompting us to take short positions based on different Moving
Averages. Additionally, we conducted a monthly rebalance of assets using Mean
Variance Optimization. To implement these methods, we utilized Backtrader and CVXopt
libraries, along with Pandas and Numpy.
Objectives of the project: Backtesting moving average crossover for stock selection &
fixed income assets with yield curve inversion as signal source
Tool used: None
Details of Papers/patents:None
Brief description of the working environment: The working environment was positive,
offering a conducive space for learning. Despite the remote setup, the company provided
comprehensive help and support, making me feel included and valued. Regular video call
meetings, occurring 2-3 times per week, fostered a comfortable atmosphere and allowed
for productive discussions.
Initially, I had expected exposure to various company departments. However, the
internship primarily focused on working closely with my assigned mentor. While this
limited my scope, it provided an opportunity to delve deeply into my area of interest and
develop a strong mentor-mentee relationship.
Under my mentor's guidance, I engaged in various tasks and projects, expanding my
knowledge and honing practical skills. These real-world challenges encouraged critical
and creative thinking, preparing me for future endeavors.
Despite the limitations, the internship provided a valuable learning experience, and I
appreciate the supportive working environment and the opportunity to collaborate with
dedicated professionals. I look forward to applying the knowledge and skills acquired
during this internship in my future pursuits, and I am eager to contribute to a company
that values and nurtures its interns as essential team members.
Academic courses relevant to the project: Computer Programming (CS F111)
Fundamentals of Finance and Accounts (ECON F212)
Derivatives and Risk Management (ECON F354)
Learning Outcome: The objective of our project was to work with Backtrader library of
Python and use Yield Spread Data as well as various other indicators, to be able to predict
the incoming financial recession in order to protect our portfolio from potentially large
drawdowns. This helped us learn more about financial strategies, and acquainted us with
the process of fine tuning parameters for best financial returns on our portfolio.
------------
242
