AI-for-Trading

---

Projects completed for Udacity's "AI for Trading" Nanodegree

---

Contents

Project 1 - Trading with Momentum

• Turning price data into daily log returns
• Creating simple trading signal
• Analyzing alpha's effectiveness using T-test

Project 2 - Breakout Strategy

• Creating a mean reversion signal
• Using KS-test (Kolmogorov-Smirnov) to analyze the performance of the strategy

Project 3 - Smart Beta Portfolio and Portfolio Optimization

• Simulating market cap weighted index
• Optimizing portfolio with the cvxpy package

Project 4 - Multi-factor Model

• Creating a statistical risk model using PCA
• Predicting the portfolio risk
• Implementing 5 Alpha factors
• Calculating Sharpe ratios

Project 5 - NLP on Financial Statements

• Pre-processing 10Ks
• Measuring Jacquard- and Cosine-Similarity of financial statements over time
• Creating alpha factor and analyzing its performance

Project 6 - Analyzing Stock Sentiment from Tweets

• Pre-processing, lemmatizing, creating a Bag of Words, and filtering tweets
• Training a LSTM Neural Net
• Using it for sentiment prediction on various stocks

Project 7 - Combine Signals for Enhanced Alpha

• Creating alpha factors, universal quant features, and regime features
• Using Random Forest to predict future returns
• Analyzing RF's performance using cross-validation and various sampling methods
• Analyzing factors' performance

Project 8: Backtesting

• Preprocessing data
• Creating alpha factors
• Calculating risk factor exposures
• Estimating transaction costs
• Optimizing portfolio
• Calculating Profit and Loss