Big Data in Finance

Big Data in Finance – Assignment 1

Algorithmic Trading Assignment

Objective: Develop and perform algorithmic trades and their strategies using big data in finance.

Requirements:

You are required to do the data analysis in Python. The purpose of this document set is toperform Big Data Science and artificial intelligence in financial data mining and ind outthe similarity and differences between your findings and the results of other researcherin journal papers.Introduction: Algorithmic trading has become an ncreasingly important tool in the financial markets,allowing traders to leverage advanced data analysis and decision-making capabilities togenerate profits. In this assignment, you will be tasked with developing and evaluatingseveral algorithmic trading strategies for the Chinese or Hong Kong stock market, or thecurrency exchange market or commodity products in different commodity exchanges inthe world.Procedures:

1. Selection of Investment Portfolio for initial capital $1,000,000:1.1 Select one business sector in accordance with Global Industry ClassificatioStandard in Appendix 1. Each student selects his/her own business sector and nobusiness sector should be repeated. Design with explanation at least 3 combinationsof investment portfolio in the selected business sector including at least10 relevant industrial stocks in China (ie. Shanghai, Shenzhen or HonKong Stock Markets), for example,Internet Software & Services (8-digit number only.HK - Alibaba Group Holding Ltd.s.

2. Trading Strategies
3. Design with explanation the trading strategies as follows:
4. Single Indicator-Based StrategyDevelop a trading strategy that relies on a single technical indicator, such as thhanghai Composite Index's 50-day moving average, the Hang Seng Index'sRelative Strength Index (RSI), or the USD/CNY exchange rate's StochastiOscillator. Explain the rationale behind your chosen indicator and how it can beused to generate buy and sell signals.

2. Multiple Indicator-Based StrategyCreate a trading strategy that combines multiple technical indicators to maketrading decisions. For example, you could use the 20-day and 50-day movingaverages of theShenzhen Component Index, along with the MACD indicator, tgenerate trading signals. Discuss how you selected the indicators and how yointegrated them into a cohesive decision-making framework.

4. Simple Neural Network AI StrategyImplement a simple neural network-based trading strategy using stock data fromthe Shanghai Stock Exchange or the Hong Kong Stock Exchange, or currencyexchange rates. Describe the architecture of your neural network, the input featuresused (e.g., price, volume, technical indicators), and the training process. Explahow the neural network generates trading signals.
5. Hybrid Indicator-Based and Neural Network AI Strategy 代 写 Develop a hybrid trading strategy that combines traditional technical indicators(such as the 200-day moving average of the CSI 300 Index) with a neural networkbased . Discuss the rationale for this approach and how the two componentsare integrated to make trading decisions.

5. Simple Deep Learning AI StrategyDesign a deep learning-based trading strategy, such as using a recurrent neuralnetwork (RNN) or a convolutional neural network (CNN) to analyze the historicalthe deep learning model is used to generate trading signals.

4. Hybrid Indicator-Based and Deep Learning AI StrategyImplement a hybrid trading strategy that integrates traditional technical indicatormodel. Explain the benefitsof this approach and how the two components worktogether to make trading decisions.Page 2Page Customized StrategiesCustomize at least one trading strategy tofind out the optimal trading strategy inyour investment combinations. More than one trading strategy would be in the bonus marks.
5. BacktestingFor each of the trading strategies developed, perform a comprehensive backtestinprocess using at least two-years historical data from the Chinese or Hong Kongstock market, the currency exchange market or different commodity exchanges.This should include:Data Preparation: Obtain and preprocessthenecessary historical marketdata for your trading strategies.Backtesting Methodology: Describe the backtesting methodology you wiluse, including the time period, the evaluation metrics (e.g., returns,drawdown, Sharpe ratio), and any assumptions or constraints.Backtesting Analytical Results: Present the backtesting results for eachtrading strategy, including performance metrics, visualizations (e.g., equity), and a comparative analysis of the strategies. For example,Sharpe ratioDrawdowWin/loss ratioOptimization and Sensitivity Analysis (Optional): Discuss anyyou used to improve the performance of yourtrading strategies, and conduct asensitivityanalysis to understand theimpact of key parameters on the strategy's
6. Real-Time Live SimulationTo further evaluate the effectiveness of your trading strategies, implement a realtime live simulation using current market data theChinese or Hong Kongstock market, or the currency exchange market. This should involve:Data Feeds (Yahoo Finance): Integrate real-time market data feeds intoyour trading system.Order Execution: Develop a mechanism to execute trades based on thgenerated by your trading strategies.Performance Monitoring and its Analysis: Continuously monitor theperformance of your trading strategies in the live market, tracking kemetrics and risk-adjusted performance. For example,Total returnSharpe ratioDrawdown/loss ratioAdaptation and Refinement: Discuss how you would adapt and refineyour trading strategies based on the insights gained from the real-time livesimulation.* Students need to suggest their own business sector. No business sector should bAbstractanBackgroLiterature Review (OptionaInvestment PortfoliTrading Strategies ***Backtesting and its analysis *** Comparison between Backtesting and the results of Real-Time Live SimulationDiscussion (ApplicatioandImplications of Relationship found) Limitations (Any issue related to the Big Data Science / Artificial Intelligence in thisstudy)ConclusionsReferences (the supporting journal and /or conference papers for your findings with

ferences (pdf files))

15. Appendices ****This section “Research Design and Methodology” should include the Big Data Scienc/ Technical Analysis / Artificial Intelligence methods andPython should be usedforprogramming.**** Python code should be attached in the appendices.Bonus:Bonus marks can be obtained as follows:

Except the requirements in Selection of Investment Portfolio in p.1, one additional Investment Portfolio used. (5 marks each max 5 marks)Except the requirements in trading strategy, one additional Artificial Intelligence,

another quantitative analysis method used not mentioned in this subject with submission of python code, data and analysis results. However, the bonus method cannot be the same as in other assignments of Big Data in Finance. (5marks each)All bonus marks are justified in acceptance of above offers in accordance with thequalityof references and data. Maximum bonus marks = 20. Requirements: Students are required to present their topic (at least 10 mins per student) and to write an

Submission: Submit all files online with the following: (I:\Terence\ Big Data in Finance\...):

1. An article (at least 10 pages per 1 student, font 12, single line spacing – count textigures, tables only) – English for English classes or English in both

1. Word andmd (Obsidian) formats (using Word to md)

1.  
2. A presentation file with speaking note and audio (please add the notes below thepowerpoint slides) (at least 5 mins per student) – English powerpoint 2019 or later

3. Python code in Python Format (py files) – 1 master py file with all trading strategies3 py for 3 investment portfolios

4. Data Files in Excel / CSV Format (xlsx/CSV) with web address of data source
5. AI prompt for Python code generation (txt file)
6. Analysis Result Files in Excel Format (xlsx)
7. All References (full text journal paper in pdf files)