Ian Klosowicz
Finance is one of the highest-paying industries for data analysts and one of the most competitive to break into. The analytical bar is higher, the questions are more quantitative, and hiring managers in finance know immediately when a portfolio project shows genuine understanding of financial data versus someone who downloaded a stock price CSV and made a line chart.
These project angles are strongest when they mirror the type of data questions financial analyst roles ask and which is worth understanding before picking a direction.
This post covers 8 specific project ideas using publicly available financial data, what each one demonstrates, and why each one reads as real analytical work rather than a tutorial exercise. Finance experience helps, but none of these require it — they require understanding what the question is and why it matters.
Finance analytics isn't one job — it splits across several very different environments, each with different analytical problems:
The distinction matters when you're building a portfolio. A project that maps to FP&A reads differently than one that maps to investment analytics. Pick the environment you're targeting and build toward it.
SEC EDGAR is the most important source for financial analytics portfolios. Every publicly traded US company files quarterly (10-Q) and annual (10-K) reports with the SEC. EDGAR makes these available in structured XBRL format through the Financial Data API, which lets you pull income statement, balance sheet, and cash flow data for any company across multiple years. Access at data.sec.gov/api/xbrl.
FRED (Federal Reserve Economic Data) publishes macroeconomic time series: interest rates, inflation, unemployment, GDP, housing starts, credit spreads, and hundreds of others going back decades. Access at fred.stlouisfed.org. Directly downloadable as CSV.
Yahoo Finance and yfinance (Python library) provide historical stock price, volume, and options data. For Python projects, yfinance pulls this data directly into a DataFrame. For SQL and BI projects, historical price CSVs are downloadable from Yahoo Finance directly.
FDIC BankFind Suite publishes quarterly financial data for every FDIC-insured bank in the US: balance sheet, income statement, loan portfolio composition, capital ratios, and performance metrics. Access at banks.data.fdic.gov. Good for banking industry analysis without needing proprietary data.
BLS Consumer Expenditure Survey tracks household spending by income bracket across dozens of categories. Good for personal finance and fintech projects.
IRS Statistics of Income publishes aggregated tax return data by income bracket, state, and year. Useful for personal finance and tax burden analysis projects.
CFPB Consumer Complaint Database publishes every consumer financial complaint filed with the CFPB, including company, product type, issue, resolution, and state. Access at consumerfinance.gov/data-research.
1. Public company financial performance comparison
Question: Across a set of companies in the same sector, which ones have improved their operating margin most consistently over 5 years, and does revenue growth or cost control explain more of the variance?
Data: SEC EDGAR Financial Data API for income statement data (revenue, operating income, net income) across 10 to 20 companies in a sector you can speak to — retail, tech, healthcare, energy.
What it demonstrates: API data pull and normalization across multiple companies and periods, margin calculation and trend analysis, SQL window functions for year-over-year change, a dashboard with company comparison and drill-down by metric. Pulling directly from EDGAR signals that you know where real financial data comes from.
Why it reads as real: equity analysts, corporate strategy teams, and FP&A benchmarking groups run exactly this analysis. Knowing how to calculate and compare operating margin trends across a peer group is a fundamental finance analytics skill.
2. Bank financial health analysis using FDIC data
Question: Which banks in a target state or asset size range have seen the largest deterioration in their loan loss reserve ratios or capital adequacy ratios over the last 3 years, and what does their net interest margin trend suggest about earnings sustainability?
Data: FDIC BankFind Suite quarterly call report data. Covers every FDIC-insured institution with balance sheet, income statement, and regulatory ratio data going back decades.
What it demonstrates: working with regulatory financial data (a real banking analytics skill), ratio calculation and trend analysis, peer group comparison by asset size or geography, a dashboard with institution-level drill-down. The FDIC dataset is completely non-tutorial and signals serious domain awareness in banking analytics.
Why it reads as real: bank examiners, bank analysts, and internal risk teams use call report data for exactly this kind of financial health monitoring. If you're targeting banking analytics roles, this is the right dataset.
3. Macroeconomic indicator relationships using FRED data
Question: How do changes in the federal funds rate historically relate to changes in unemployment, inflation, and housing starts, and how do the lag relationships compare across economic cycles?
Data: FRED time series for fed funds rate, CPI, unemployment rate, and housing starts going back to the 1970s. All downloadable as CSV from fred.stlouisfed.org.
What it demonstrates: time series joins on date, lag analysis, correlation across economic periods, a multi-series line chart dashboard with recession shading and period selector. Working with 50+ years of monthly economic data at multiple series simultaneously is a scale and complexity signal most entry-level portfolios don't reach.
Why it reads as real: economists, macro strategists, and FP&A teams building economic scenario models use FRED data constantly. Knowing how interest rate changes transmit to the real economy is foundational knowledge for anyone in finance analytics.
4. Sector equity performance attribution
Question: Over the last 5 years, how much of a sector ETF's return came from price appreciation versus dividends, how did that compare to the broader market, and which individual holdings drove the outperformance or underperformance?
Data: Yahoo Finance historical price and dividend data for a sector ETF (XLF for financials, XLK for tech, XLV for healthcare) and the S&P 500 (SPY). yfinance in Python or direct CSV downloads.
What it demonstrates: total return calculation (price + dividends), benchmark comparison, contribution analysis by holding, a performance dashboard with rolling return comparison. Total return vs. price return is a distinction most people without finance backgrounds get wrong — getting it right signals domain knowledge.
Why it reads as real: portfolio analysts, fund accountants, and investment consultants build performance attribution reports constantly. Even a simplified version of this analysis shows you understand how investment returns are measured and decomposed.
5. Consumer financial complaint trends
Question: Which financial products and companies generate the most unresolved complaints, how has the complaint volume changed over time, and which issues are most likely to result in monetary relief to the consumer?
Data: CFPB Consumer Complaint Database. Over 3 million complaints with company name, product type, issue, sub-issue, company response, consumer disputed status, and state. Directly downloadable as CSV.
What it demonstrates: large dataset aggregation (3M+ rows), text-based grouping and categorization, resolution rate analysis, time series by product and company, a dashboard with company and product filters. The CFPB database is completely non-tutorial and maps naturally to compliance, risk, and consumer analytics roles at financial institutions.
Why it reads as real: compliance teams and consumer protection analysts at banks and fintechs track complaint trends actively. Regulators use this data to identify systemic issues. Building this analysis signals awareness of the regulatory environment financial institutions operate in.
6. Personal income and tax burden analysis by state
Question: How does effective tax burden (federal + state) vary across income brackets and states, and which states offer the largest tax advantage for high-income earners versus middle-income earners?
Data: IRS Statistics of Income (SOI) data by state and income bracket, combined with state income tax rate tables (publicly available from the Tax Foundation or state revenue departments).
What it demonstrates: multi-source join on state and income bracket, effective rate calculation vs. marginal rate, bracket-level comparison across states, a dashboard with state and income filter. Distinguishing effective from marginal tax rates is the kind of detail that signals real financial literacy.
Why it reads as real: wealth management analysts, tax advisors, and personal finance platforms use exactly this analysis to advise clients on state residency decisions and tax optimization strategies.
7. Credit union vs. bank performance comparison
Question: How do credit unions compare to commercial banks on loan growth, deposit rates offered, loan-to-deposit ratios, and charge-off rates, and has that gap widened or narrowed over the last 5 years?
Data: FDIC BankFind for bank data, NCUA Call Report data for credit unions (available at ncua.gov/analysis/credit-union-corporate-call-report-data). Both publish quarterly financial data in comparable formats.
What it demonstrates: cross-institution-type comparison using 2 regulatory data sources, financial ratio calculation and trend analysis, peer group normalization by asset size, a comparative dashboard with institution type toggle. Working with NCUA data alongside FDIC data is a non-obvious move that signals real depth in financial services analytics.
Why it reads as real: financial industry analysts, consultants, and institution strategists compare credit union and bank performance regularly to identify competitive positioning and market share shifts.
8. Revenue and expense variance analysis for a public company
Question: For a company you know well, how did actual quarterly revenue and operating expenses compare to analyst consensus estimates over the last 8 quarters, and what does the pattern of beats and misses suggest about management's guidance accuracy?
Data: SEC EDGAR for actual financial results, analyst consensus estimates from sources like Macrotrends.net or Wisesheets (free tiers available). Combine actual vs. estimate data manually for a company you can speak to.
What it demonstrates: variance analysis (actual vs. expected), earnings beat/miss pattern identification, guidance accuracy analysis over time, a dashboard showing quarterly results vs. estimates with beat/miss indicators. Variance analysis is the core FP&A skill — building a version of it on public data signals that you understand the budget vs. actuals analytical framework.
Why it reads as real: FP&A analysts and equity research analysts build this analysis every earnings season. If you're targeting corporate FP&A or equity research support roles, this is the most directly relevant project you can build.
If you want a structured approach to picking the right project and building it to a standard that gets interviews in finance roles, the Analyst Hive program covers the project build sequence in Month 1.
Finance interviewers ask follow-up questions that go deeper into the business context than interviewers in most other industries. Know why the metric you're analyzing matters before you walk in.
"Operating margin" means more when you can explain that it measures the efficiency of the core business before the effects of financing and taxes, and that comparing it to gross margin tells you something about operating expense discipline. "Loan loss reserve ratio" means more when you can explain that it's a forward-looking measure of how much management expects to lose on its loan portfolio, and that a declining ratio during an economic downturn is a concern.
You don't need to be a finance professional to say these things. You need to read enough about the metric to understand what it's measuring and why a financial institution would care about it. An hour of research before the interview for each metric in your project is usually enough.
If you have a finance background — accounting, banking, investment operations, insurance, or adjacent — connect the project explicitly to work you've done or decisions you've seen made. That context is irreplaceable and something no other candidate in the pool has.
Do I need to know how to value stocks to build these projects?
No. None of these projects require DCF modeling, stock valuation, or investment theory. They require pulling financial data, calculating ratios and trends, and communicating what you found. The analytical skills are the same ones used in any data analyst role — SQL, aggregation, trend analysis, and clear visualization. The finance context is the domain layer on top of those skills, not a prerequisite to using them.
Is Tableau or Power BI better for finance portfolio projects?
Tableau is slightly more common in investment management, consulting, and larger financial services firms. Power BI is more common in corporate FP&A, banking operations, and mid-market finance teams that run on Microsoft infrastructure. Check the job postings you're targeting. Either tool produces a strong finance portfolio project — the tool matters less than the analytical question and the quality of the data model behind the dashboard.
Can I use stock price data for a portfolio project?
Yes, but a project that just charts stock prices over time isn't analytical — it's a visualization of existing information. The stronger move is to use price data as an input to a calculation: total return including dividends, volatility comparison across sectors, beta calculation relative to a benchmark, or drawdown analysis during specific market events. Any of those turns price data into an analysis rather than a display.
What if I want to target fintech specifically?
Fintech analytics skews toward product and customer analytics more than traditional finance analytics. The CFPB complaint project (#5) maps well to compliance and risk analytics at fintechs. For product and growth analytics at fintechs, the retail and e-commerce cohort retention and RFM projects actually transfer well — fintech companies treat financial products like products, and the analytical frameworks are similar. SQL and Python are both more relevant in fintech than in traditional finance.
Is the SEC EDGAR API hard to use?
It's not a simple CSV download, but it's well-documented and free. The EDGAR Financial Data API returns JSON for each company and filing type. A basic Python script using the requests library can pull income statement data for multiple companies and load it into a DataFrame in an afternoon. If you're not comfortable with APIs yet, Macrotrends.net provides downloadable CSVs of EDGAR financial data for individual companies as a simpler starting point.
Which project is strongest for a corporate FP&A role?
The revenue and expense variance analysis (#8) maps most directly to FP&A work because budget vs. actuals variance analysis is the core deliverable of most FP&A teams. The macroeconomic indicator project (#3) is also strong if the company uses economic scenario planning. Build whichever one you can speak to most confidently — the domain knowledge you demonstrate in the walkthrough is as important as the technical execution.
If you want a structured path through building a finance project that clears the bar, the Analyst Hive program covers the full build in Month 1. Daily tasks, structured around getting hired.
