Banking - Investment Analysis and Solutions

YOU CAN CHECK OUT MY PROJECT NOTEBOOKS FIRST!

Table of Contents

• Overview
• Tools and Techniques
• Setup
• Credit Card Approvals
• Bank Marketing Campaign
• Hedge Fund Financial Report
• Stock Portfolio Analysis

Overview

Delivered connected insights across financial systems by applying structured data workflows, predictive modeling, and optimization techniques to improve decision-making in banking, marketing, and investments.

1. Credit Card Approvals: Built a logistic regression model that classified credit card applications with accuracy of 0.798, streamlining the decision-making process.
2. Bank Marketing Campaign: Structured marketing data into PostgreSQL-compatible datasets, enabling scalable analysis of campaign performance.
3. Hedge Fund Financial Report: Analyzed leverage and profitability ratios across sectors, providing insights for hedge fund strategies.
4. Stock Portfolio Analysis: Optimized FAANG stock portfolios using mean-variance optimization, achieving a Sharpe ratio of 3.5.

Tools and Techniques

• Python, Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn, PyPortfolioOpt, PostgreSQL.
• Data preprocessing, Data pipeline cleaning, Data wrangling, Feature engineering, Data visualization, Supervised learning, Logistic regression, Hyperparameter tuning, Financial modeling, Ratio analysis, Portfolio optimization.

Setup

git clone <repository-url>
cd <repository-directory>
pip3 install -r requirements.txt
jupyter notebook bank-marketing-campaign.ipynb

Credit Card Approvals

Objective: Automate the credit card approval process using supervised learning.

Key Tasks:

• Preprocessed raw data by replacing missing values and encoding categorical features.
• Trained a logistic regression model to classify applications as approved or rejected.
• Tuned hyperparameters using GridSearchCV to optimize model performance.

Key Insights:

• Achieved an accuracy of 79.8%, improving the efficiency of the credit card approval process.
• Provided a scalable pipeline for automating application decisions.

Bank Marketing Campaign

Objective: Enhanced marketing strategies for the bank personal loan campaigns by analyzing customer demographics and campaign outcomes.

Key Tasks:

• Processed raw data into three structured DataFrames: client, campaign, and economics, each designed for specific analytical purposes.
• Reformatted data to meet PostgreSQL database standards, enabling seamless integration and scalability.

Key Insights:

• Created clean, consistent datasets for future analysis of campaign success rates.
• Enabled better tracking of customer behavior and campaign performance.

Hedge Fund Financial Report

Objective: Assess financial health and risk metrics across industries.

Key Tasks:

• Calculated leverage ratios (debt-to-equity) and profitability ratios (gross margin) for companies in tech, FMCG, and real estate sectors.
• Analyzed sector trends, identifying real estate as having the highest leverage ratio (5.69) and FMCG as the least profitable sector (profitability ratio: 0.51).
• Visualized the positive correlation between leverage and profitability in real estate companies.

Key Insights:

• Provided industry-specific financial insights to guide investment decisions.
• Enhanced understanding of risk and return dynamics in diverse sectors.

Stock Portfolio Analysis

Objective: Analyze FAANG stocks to determine optimal portfolio allocations.

Key Tasks:

• Calculate expected returns and annualized Sharpe ratio for an equally-weighted portfolio.
• Find a portfolio that minimizes volatility using mean-variance optimization.
• Find a portfolio that maximizes the Sharpe ratio using mean-variance optimization.

Key Insights:

• Achieved a Sharpe ratio of 3.5 in the optimized portfolio, showcasing how diversification reduces risk.
• Highlighted efficient investment strategies based on historical stock performance.