Mutual Fund Analytics Capstone Project

Project Overview

This project is a Mutual Fund Analytics Capstone Project focused on building a complete data analytics pipeline for the Indian Mutual Fund industry.

The project includes:

• Data ingestion from CSV datasets
• Live NAV data collection using MFAPI
• Data quality validation
• Data cleaning and transformation
• Exploratory Data Analysis (EDA)
• SQL-based analysis
• Dashboard development
• Business insights and reporting

---

Project Structure

MUTUAL_FUND_ANALYTICS
│
├── data/
│   ├── raw/
│   └── processed/
├── EDA_charts/
├── notebooks/
├── Fund Performance Analytics outputs/
├── database/
├── sql/
├── dashboard/
├── reports/
│
├── data_ingestion.py
├── data_cleaning.py
├── database.py
├── live_nav_fetch.py
├── requirements.txt
├── data_dictionary.md
├── README.md

---

Datasets

The project uses the following datasets:

1. Fund Master
2. NAV History
3. AUM by Fund House
4. Monthly SIP Inflows
5. Category Inflows
6. Industry Folio Count
7. Scheme Performance
8. Investor Transactions
9. Portfolio Holdings
10. Benchmark Indices

---

Task 1 Deliverables

Project Setup

• Created project folder structure
• Initialized Git repository
• Connected project to GitHub

Data Ingestion

• Loaded all 10 datasets using Pandas
• Displayed dataset shapes
• Inspected column names
• Verified data types
• Displayed sample records

Data Quality Checks

• Checked missing values
• Checked duplicate records
• Validated AMFI codes across datasets

Fund Exploration

• Identified unique fund houses
• Analyzed categories and sub-categories
• Examined risk classifications

Live NAV Data Collection

• Connected to MFAPI
• Parsed JSON responses
• Converted API data into Pandas DataFrames
• Saved NAV history as CSV files

---

Key Findings

• Successfully loaded all datasets.
• No duplicate records found.
• All AMFI codes in fund_master exist in nav_history.
• Missing values were found only in yoy_growth_pct within monthly_sip_inflows.csv.
• Dataset contains 10 mutual fund houses.
• Dataset includes Equity and Debt categories with multiple sub-categories and risk levels.

---

Task 2 Deliverables

Data Cleaning & Preprocessing

• Cleaned all raw datasets and created processed versions.
• Standardized column names across datasets.
• Converted date columns to proper datetime format.
• Removed duplicate records.
• Handled missing values using appropriate techniques.
• Corrected inconsistent data types.
• Standardized categorical values.
• Validated AMFI codes across related datasets.
• Created derived columns where required (e.g., year from date).
• Saved cleaned datasets into the processed folder.

Processed Datasets

• 01_fund_master_cleaned.csv
• 02_nav_history_cleaned.csv
• 03_aum_history_cleaned.csv
• 04_monthly_sip_inflows_cleaned.csv
• 05_category_inflows_cleaned.csv
• 06_industry_folio_count_cleaned.csv
• 07_scheme_performance_cleaned.csv
• 08_investor_transactions_cleaned.csv
• 09_portfolio_holdings_cleaned.csv
• 10_benchmark_indices_cleaned.csv

Key Outcomes

• Improved data consistency across datasets.
• Removed invalid and duplicate records.
• Prepared datasets for visualization and analysis.
• Established a clean data pipeline for downstream EDA and SQL analysis.

---

Task 3 Deliverables

Exploratory Data Analysis (EDA)

Created multiple visualizations to analyze trends, fund performance, investor activity, and portfolio allocation.

Visualizations Created

1. Daily NAV Trend Analysis (Plotly)
2. AUM Growth by Fund House (Seaborn)
3. Monthly SIP Inflow Trend (Plotly)
4. Category-wise Net Inflow Heatmap (Seaborn)
5. Mutual Fund Folio Count Growth
6. NAV Return Correlation Matrix
7. Sector Allocation Donut Chart

Key EDA Insights

• Daily NAVs showed an overall upward trend across most schemes.
• AUM increased consistently across major fund houses.
• Monthly SIP inflows demonstrated sustained growth over the analysis period.
• Equity-oriented categories attracted comparatively higher inflows.
• Mutual fund folios increased steadily, indicating growing retail participation.
• Daily NAV returns of several equity schemes showed strong positive correlations.
• Financial Services and Information Technology represented significant portions of portfolio allocations.

---

Task 4 – Fund Performance Analytics

Objective

Performed comprehensive performance analysis of 40 mutual fund schemes using historical NAV data and benchmark indices. Calculated key risk-return metrics, compared fund performance against market benchmarks, and developed a composite fund scorecard.

Tasks Performed

• Loaded processed datasets:
  • Fund Master
  • NAV History
  • Benchmark Indices
• Prepared and merged datasets using amfi_code and date
• Computed daily returns for all mutual fund schemes
• Calculated 1-Year and 3-Year CAGR from NAV history
• Documented 5-Year CAGR as unavailable due to insufficient historical data
• Calculated Sharpe Ratio using a 6.5% annual risk-free rate
• Calculated Sortino Ratio using downside deviation
• Estimated Alpha and Beta using linear regression against the NIFTY100 benchmark
• Computed Maximum Drawdown and identified peak-to-trough periods
• Built a composite Fund Scorecard (0–100) using weighted ranking methodology
• Compared the top 5 mutual funds against NIFTY50 and NIFTY100 over the last 3 years
• Calculated Tracking Error relative to the NIFTY100 benchmark
• Exported final analysis outputs

Key Performance Metrics

• Daily Returns
• CAGR (1-Year & 3-Year)
• Sharpe Ratio
• Sortino Ratio
• Alpha
• Beta
• Maximum Drawdown
• Tracking Error
• Composite Fund Score (0–100)

Visualizations

• Daily Return Distribution Histogram
• Daily Return Box Plot
• Top 10 Funds by Sharpe Ratio
• Top 10 Funds by Sortino Ratio
• Top 10 Funds by Alpha
• Drawdown Curve
• Top 10 Fund Score Comparison
• Top 5 Funds vs NIFTY50 & NIFTY100 (3-Year Performance)
• Tracking Error Comparison

Deliverables

• Performance_Analytics.ipynb
• fund_scorecard.csv
• alpha_beta.csv
• tracking_error.csv
• benchmark_comparison.png

Skills Demonstrated

• Financial Performance Analysis
• Risk-Adjusted Performance Evaluation
• Portfolio Analytics
• Time Series Analysis
• Linear Regression (OLS)
• Benchmark Performance Comparison
• Feature Engineering
• Data Visualization
• Pandas
• NumPy
• Matplotlib
• SciPy

---

Technologies Used

• Python
• Pandas
• NumPy
• Requests
• Matplotlib
• Seaborn
• Plotly
• SQLAlchemy
• Jupyter Notebook
• Git
• GitHub

---

Installation

Clone the Repository

git clone <repository-url>

Navigate to the Project Directory

cd mutual-fund-analytics

Install Dependencies

pip install -r requirements.txt

---

Project Workflow

CSV Datasets
      │
      ▼
Data Ingestion
      │
      ▼
Data Validation
      │
      ▼
Data Cleaning & Preprocessing
      │
      ▼
Processed Datasets
      │
      ▼
Exploratory Data Analysis (EDA)
      │
      ▼
SQL Analysis
      │
      ▼
Dashboard Development
      │
      ▼
Business Insights & Reporting

EDA Charts Generated

The following charts were generated and exported as PNG files:

• NAV Trend Analysis
• AUM Growth by Fund House
• Monthly SIP Inflow Trend
• Category Inflow Heatmap
• Folio Count Growth
• NAV Return Correlation Matrix
• Sector Allocation Donut Chart

All charts are stored in the EDA_charts/ directory.

---

Future Work

• SQL-based analytical queries
• Dashboard development using Power BI
• Advanced business insights
• Performance benchmarking against market indices
• Portfolio performance analytics
• Interactive reporting and visualization

---

Author

Asritha Lalam

Mutual Fund Analytics Capstone Project