If you’re a high school student interested in pursuing research across finance and technology, the research topic you choose will shape nearly every part of the experience that follows. A topic that’s too broad gives you nothing to argue, too narrow gives you nothing to read, and a topic that’s only vaguely connected to your stated interest makes for unconvincing essays and shallow mentorship conversations. Picking something with a precise research question and a defined scope matters because it forces you to engage with real academic literature, work with actual datasets, and arrive at a conclusion you can defend, all skills that college applications and research programs are looking for. FinTech is a particularly productive area to work in because it sits at the intersection of economics, computer science, law, and behavioral science, meaning the published literature is substantial, public data is often freely available, and the questions are actively contested among researchers and policymakers. You don’t need access to a university lab or proprietary financial data to do credible work here. What you need is a question narrow enough to answer in the time and with the resources you have.
How do you pick a good FinTech topic?
Four things determine whether a FinTech topic will hold up as a research project. The first is specificity: your topic needs a defined research question, not just a subject area. So while cryptocurrency is a subject, “whether Bitcoin’s price movements correlated with gold during inflationary periods between 2018 and 2023″ is a question you can actually answer with data. The second is source availability: before you commit to a topic, check whether peer-reviewed papers, government datasets, or primary documents exist for it. Most fintech topics are covered in Google Scholar, and many have dedicated public data sources, such as the CFPB’s consumer complaint database, the FDIC’s unbanked survey, or the World Bank’s remittance cost tracker. The third is personal interest, because you’ll be reading dense regulatory documents and academic papers for weeks, and topics you find compelling produce more coherent writing. The fourth is real-world grounding: the strongest FinTech research papers connect to an ongoing debate, a regulatory development, or a documented market event.
The 25 topics below were selected to cover a range of difficulty levels, from beginner-level projects involving survey design and publicly available reports to advanced case studies that require comfort with reading technical white papers and working with on-chain blockchain data. They also span the main sub-fields within FinTech, such as payments, cryptocurrency, AI in finance, financial inclusion, market structure, and regulation, so regardless of where your interests lie, there’s a starting point here that fits.
1. How Buy Now Pay Later Services Affect Teen Spending Habits
Sub-field: Consumer FinTech / Behavioral Finance
Difficulty: Beginner
Why it’s interesting: BNPL services like Klarna and Afterpay have exploded in popularity among teens and young adults, yet there is ongoing debate among consumer finance researchers about whether deferred payment structures encourage people to spend beyond their means or simply offer a more flexible alternative to credit cards.
Suggested research question: Do high school and college-aged users of BNPL services report higher rates of impulse purchasing than peers who use debit or credit cards?
Key methods/data sources: Academic papers on behavioral economics (Google Scholar), Consumer Financial Protection Bureau (CFPB) reports on BNPL, original surveys distributed to peers, news analysis of industry data from Klarna and Afterpay annual reports.
Good fit for: Students interested in psychology, consumer behavior, or personal finance with no prior research experience.
BNPL is worth studying because the product is engineered to reduce the psychological friction of spending. After all, breaking a $200 purchase into four $50 payments makes the total feel smaller, and behavioral economists call this “pain of paying” reduction. Your research will likely center on a survey you design and distribute to peers, asking about purchase regret, awareness of total cost, and whether BNPL use correlates with going over budget. You’ll want to supplement that with secondary sources like the CFPB’s consumer complaints data on BNPL. Academic papers on payment decoupling can give you the theoretical vocabulary to frame what you’re observing. The core tension to investigate is whether BNPL actually helps cash-flow management or whether its design nudges users toward purchases they wouldn’t otherwise make.
2. Cash vs. Digital Wallets: How Age and Income Shape Payment Preferences in the U.S.
Sub-field: Payments and Consumer FinTech
Difficulty: Beginner
Why it’s interesting: Despite rapid growth in digital payment adoption, cash use has not disappeared, and the gap between who uses each method maps closely onto age, income, and geography in ways that reveal broader inequalities in financial access.
Suggested research question: To what extent does household income predict preference for digital payment methods over cash among American adults?
Key methods/data sources: Federal Reserve’s annual Diary of Consumer Payment Choice, Pew Research Center surveys, Google Scholar for academic literature on payment behavior, basic descriptive statistics, or chart-making in Excel.
Good fit for: Beginners with an interest in social science, economics, or everyday financial habits.
This topic sits at the intersection of economics and social behavior, with plenty of published data available to help answer it. The Federal Reserve releases its Diary of Consumer Payment Choice annually, which tracks exactly what payment methods different demographic groups use across different transaction types. Your job is to dig into those tables, identify patterns across income and age brackets, and then explain them using academic literature on payment psychology and access. You’ll learn to make sense of a rich public dataset, and the analytical skill lies in figuring out which differences are meaningful versus coincidental. A good paper here will also acknowledge what the data doesn’t outright address, like why people in certain groups prefer cash even when digital options are available.
3. How the COVID-19 Pandemic Accelerated Contactless Payment Adoption Across Different Countries
Sub-field: Payments / FinTech Policy
Difficulty: Beginner
Why it’s interesting: The pandemic created a rare natural experiment; countries with different baseline levels of contactless infrastructure saw dramatically different rates of adoption practically overnight, making it possible to study what actually drives payment behavior change.
Suggested research question: Which country-level factors (existing infrastructure, government policy, consumer trust) best predict how quickly contactless payment adoption rose between 2019 and 2022?
Key methods/data sources: World Bank financial access data, Mastercard and Visa public reports on contactless usage, central bank publications from the EU, UK, and Australia, and comparative country-level analysis using spreadsheets.
Good fit for: Students interested in global economics or public policy with an interest in comparative analysis.
The pandemic forced a behavioral shift in how people paid for things, and what makes for a research opportunity isn’t just the fact that contactless use went up, but why it went up faster in some places than others. Countries like Australia and the UK already had widespread tap-to-pay infrastructure and high contactless limits before 2020, while the U.S. lagged partly due to infrastructure gaps and merchant resistance. Your research will involve pulling adoption statistics from central bank publications and Visa or Mastercard public reports, then matching those numbers to country-level variables, like existing NFC terminal density, government messaging, and consumer trust levels. The analytical challenge is untangling what was already in motion before COVID from what the pandemic catalyzed, requiring you to look at pre-2020 baseline data alongside the 2020 to 2022 surge.
4. The Mobile Banking Gap: Does Smartphone-Based Banking Reduce Financial Exclusion Among Lower-Income Americans?
Sub-field: Financial Inclusion / Consumer FinTech
Difficulty: Beginner to Intermediate
Why it’s interesting: Proponents argue that mobile banking removes the physical and fee-related barriers that keep lower-income households out of the traditional banking system, but FDIC data consistently show that unbanked rates remain high in certain income brackets despite widespread smartphone ownership.
Suggested research question: Does access to mobile banking apps meaningfully reduce the proportion of unbanked households among Americans earning under $30,000 per year?
Key methods/data sources: FDIC National Survey of Unbanked and Underbanked Households, Pew Research Center smartphone ownership data, Google Scholar for fintech inclusion literature, Excel-based data analysis.
Good fit for: Students curious about inequality, public policy, or the gap between what technology promises and what it delivers.
The FDIC’s annual unbanked survey is your primary resource here, and it contains granular data, including but not limited to breakdowns by income, race, education level, and the reasons people give for not having bank accounts. This research focuses on the paradox that smartphone ownership among lower-income Americans is high, neobanks have removed many traditional barriers to account-opening, and yet unbanked rates haven’t declined as sharply as expected. Your paper needs to grapple with what’s keeping people out of the banking system, whether it’s documentation requirements, distrust of financial institutions, inconsistent income that makes account maintenance stressful, or something else entirely. Reading both the FDIC data and qualitative research from consumer finance scholars will give you a fuller picture than either source alone.
5. Bitcoin as a Store of Value: A Decade-Long Comparison with Gold
Sub-field: Cryptocurrency / Asset Markets
Difficulty: Intermediate
Why it’s interesting: Bitcoin’s proponents have argued since its early days that it functions as a hedge against inflation similar to gold, but its extreme price volatility raises real questions about whether it actually behaves that way in practice, a debate that intensified during the 2022 crypto crash.
Suggested research question: Over the past ten years, how closely has Bitcoin’s price movement correlated with gold during periods of high inflation or economic uncertainty?
Key methods/data sources: Historical Bitcoin price data from CoinGecko or Yahoo Finance, gold price data from the World Gold Council, CPI data from the Bureau of Labor Statistics, correlation analysis using Python or Excel.
Good fit for: Students interested in investing or macroeconomics who are comfortable working with numerical data.
Gold is considered a store of value largely because it holds purchasing power during inflationary periods and tends to move independently of stock markets during crises. Bitcoin’s advocates claim it shares those properties, but the actual price data tells a more complicated story. You can pull Bitcoin’s historical price from CoinGecko and gold’s from the World Gold Council, then overlay them against CPI inflation data from the Bureau of Labor Statistics. You’ll use correlation analysis to calculate how closely Bitcoin and gold move together during specific economic events like the 2020 market crash, the 2021 to 2022 inflation surge, and the 2022 crypto collapse. The goal is to let the numbers show whether Bitcoin actually functions as a hedge in practice.
6. The Environmental Cost of Bitcoin Mining: Is Proof-of-Stake a Meaningful Solution?
Sub-field: Cryptocurrency / Sustainability
Difficulty: Intermediate
Why it’s interesting: Bitcoin’s annual energy consumption rivals that of entire countries, and Ethereum’s 2022 switch from proof-of-work to proof-of-stake was claimed to reduce its energy use by over 99%, yet critics argue the comparison is misleading and that the broader crypto sector’s footprint remains understated.
Suggested research question: How does the per-transaction energy cost of proof-of-work cryptocurrencies compare to proof-of-stake alternatives, and is the difference large enough to matter environmentally?
Key methods/data sources: Cambridge Centre for Alternative Finance Bitcoin Electricity Consumption Index, Ethereum Foundation post-merge reports, academic papers on blockchain energy use via Google Scholar, data visualization tools.
Good fit for: Students interested in both technology and environmental science who want to engage with a contested empirical debate.
Bitcoin’s energy consumption is well-documented; the Cambridge Centre for Alternative Finance maintains a real-time Bitcoin Electricity Consumption Index that you can mine for historical data. Ethereum’s shift from proof-of-work to proof-of-stake in 2022 (known as the Merge) gives a before-and-after comparison to work with, since the Ethereum Foundation published energy use data on both sides of that transition. Your research will involve reading those primary documents, finding independent academic analyses of the claims, and interrogating whether comparing per-transaction energy use between Bitcoin and Ethereum actually makes sense, given that they serve different purposes at different scales.
7. What Drives Cryptocurrency Adoption in Developing Economies? Evidence from Nigeria, Argentina, and El Salvador
Sub-field: Cryptocurrency / Financial Inclusion
Difficulty: Intermediate
Why it’s interesting: These three countries have some of the highest crypto adoption rates in the world, despite very different contexts: Argentina’s use is tied to hyperinflation and currency controls, Nigeria’s to remittances and a large youth population, and El Salvador’s to a government mandate, which makes them a useful comparative lens.
Suggested research question: To what extent do currency instability, remittance costs, and regulatory environment explain differences in crypto adoption rates across developing economies?
Key methods/data sources: Chainalysis Global Crypto Adoption Index, World Bank remittance data, IMF country reports, academic literature via Google Scholar, and case study methodology.
Good fit for: Students interested in international economics, development, or policy who prefer reading and analysis over data crunching.
Each of these three countries has a high crypto adoption rate, but the underlying driver is different in each case. Argentina’s use is tied to inflation because its citizens are converting pesos into stablecoins to preserve value when the official currency is losing purchasing power rapidly. Nigeria’s adoption is tied heavily to remittances and a young, tech-literate population working around currency controls. El Salvador is the only country in the world to make Bitcoin legal tender, which is a government-imposed experiment rather than an organic one. The Chainalysis Global Crypto Adoption Index ranks these countries and gives you a quantitative starting point, but your paper should explore the institutional and economic context behind each number. This is a case-study-driven research project that rewards reading IMF country reports and local journalism alongside the fintech data.
8. Blockchain in Cross-Border Trade Finance: How Distributed Ledgers Are Reducing Settlement Delays
Sub-field: Blockchain / Trade Finance
Difficulty: Intermediate to Advanced
Why it’s interesting: Cross-border trade finance has historically relied on paper-based documentation processes that take days and are prone to fraud. Pilot programs using blockchain at institutions like HSBC and JPMorgan have claimed to cut settlement times from days to hours, but independent evaluation of these claims is limited.
Suggested research question: What measurable improvements in settlement time and fraud reduction have blockchain-based trade finance platforms demonstrated compared to traditional letter-of-credit processes?
Key methods/data sources: Published case studies from HSBC, Contour, and Marco Polo network, World Trade Organization reports, academic papers on distributed ledger trade applications via Google Scholar, and qualitative analysis of pilot program outcomes.
Good fit for: Students interested in international business, logistics, or enterprise technology with some comfort level reading financial and technical documentation.
Traditional trade finance relies on letters of credit, largely paper-based documents that move between importers, exporters, and their banks in a process that can take five to ten days and involves significant risk of document fraud. Blockchain pilots from banks like HSBC and platforms like Contour have claimed to compress that timeline to hours by replacing physical documents with immutable digital records shared across parties in real time. Your research will involve reading those published pilot outcomes to understand what was measured, under what conditions, and how the results compare to the baseline. This is an area where the gap between promotional press releases and independently verified results is wide, and part of your contribution as a researcher is making that distinction. The World Trade Organization and the International Chamber of Commerce both publish reports on trade finance digitization that give you an institutional framing beyond what the tech companies are claiming.
9. Alternative Credit Scoring: How AI-Based Models Differ from FICO in Predicting Loan Default
Sub-field: AI in Finance / Credit Markets
Difficulty: Intermediate
Why it’s interesting: Traditional FICO scores exclude roughly 26 million American adults who lack sufficient credit history, and AI-based models that draw on alternative data: utility payments, rent history, cash flow, promise to fill that gap, though their accuracy and fairness remain actively debated.
Suggested research question: Do machine learning credit scoring models that incorporate alternative data outperform FICO scores in predicting loan default among borrowers with thin credit files?
Key methods/data sources: Academic papers on alternative credit scoring via Google Scholar, CFPB reports on credit invisibility, published research from lenders like Upstart, and conceptual analysis if real loan data is inaccessible.
Good fit for: Students with an interest in data science, fairness in technology, or consumer finance.
About 26 million Americans are considered credit invisible, meaning they lack enough traditional credit history for a FICO score to be calculated. Lenders like Upstart have published claims that their machine learning models, which incorporate factors like education, employment history, and cash flow patterns, approve more borrowers from this group while maintaining comparable default rates. Your research task will be to read both the company’s published claims and the independent academic and regulatory assessments of those claims, such as the CFPB’s published evaluation of Upstart. The methodological question is how you even define “better”: a model that approves more people is only an improvement if those approvals don’t come with higher default rates, and assessing that trade-off is where the analysis gets substantive.
10. Bias in Algorithmic Lending: Do Automated Credit Systems Produce Racially Disparate Outcomes?
Sub-field: AI in Finance / FinTech Ethics
Difficulty: Intermediate
Why it’s interesting: A 2019 UC Berkeley study found that fintech mortgage lenders discriminated 40% less than face-to-face lenders — but other research has documented that training AI on historical loan data can encode past discrimination into future decisions, making this one of the sharpest contested questions in applied AI ethics.
Suggested research question: What mechanisms allow algorithmic lending models trained on historical data to reproduce racially disparate loan approval rates even when race is not an explicit input variable?
Key methods/data sources: Home Mortgage Disclosure Act (HMDA) data, academic literature on algorithmic fairness via Google Scholar, CFPB enforcement actions, policy analysis from the Brookings Institution.
Good fit for: Students interested in civil rights, ethics, or technology policy who prefer writing and analysis over quantitative methods.
A key paper to start with is the 2019 UC Berkeley study by Bartlett et al., which used Home Mortgage Disclosure Act data to compare lending outcomes between algorithmic and face-to-face lenders. They found that algorithmic lenders discriminated less in approval decisions but still charged Black and Hispanic borrowers more than comparable white borrowers, complicating the optimistic narrative around AI lending. Your paper will need to engage with exploring proxy input variables that AI lenders factor in, such as neighborhood, zip code, school attended, that may correlate with race, even when race itself is excluded. Reading the original Bartlett paper alongside responses from legal scholars and AI fairness researchers gives you the range of perspectives you need to write something analytically rigorous.
11. Machine Learning in Bank Fraud Detection: How Effective Are Automated Systems and at What Cost?
Sub-field: AI in Finance / Cybersecurity
Difficulty: Intermediate
Why it’s interesting: Banks now process millions of transactions per second, making human-led fraud review impossible — yet the machine learning models used to flag suspicious activity produce false positives that can freeze accounts for innocent customers, creating a real trade-off between security and usability that banks rarely discuss publicly.
Suggested research question: How do supervised machine learning fraud detection models balance minimizing false negatives (missed fraud) against minimizing false positives (wrongly flagged transactions), and what does that trade-off cost consumers?
Key methods/data sources: Academic papers on fraud detection algorithms via Google Scholar, publicly available fraud detection datasets on Kaggle (e.g., the IEEE-CIS Fraud Detection dataset), CFPB complaint data, and introductory Python if the student wants to experiment with a model.
Good fit for: Students interested in applied machine learning or cybersecurity who want to combine technical reading with policy analysis.
Banks use supervised machine learning models trained on historical transaction data to flag suspicious activity in real time by training them on which combinations of transaction size, location, timing, and merchant category tend to precede confirmed fraud. The metric that makes this research interesting is the false positive rate, i.e., how often the model flags a legitimate transaction as fraudulent, and what happens to the customer when it does? The CFPB complaint database lets you look at real consumer reports of accounts being frozen or transactions declined, which gives you a qualitative window into what the statistical trade-off looks like on the ground. Kaggle hosts the IEEE-CIS Fraud Detection dataset, which you can experiment with in Python if you want to understand how to run a model yourself, though the literature review alone is substantial enough to anchor a strong paper.
12. Robo-Advisors vs. Human Financial Advisors: Who Delivers Better Outcomes for Small Investors?
Sub-field: AI in Finance / Wealth Management
Difficulty: Beginner to Intermediate
Why it’s interesting: Robo-advisors like Betterment and Wealthfront promise low-cost, algorithmically managed portfolios accessible to investors with as little as $1, but whether they actually outperform human advisors after accounting for fees, tax efficiency, and behavioral coaching during market downturns is still debated in academic literature.
Suggested research question: For investors with under $50,000 in assets, do robo-advisor-managed portfolios produce better net-of-fee returns than human-advisor-managed portfolios over five-year periods?
Key methods/data sources: Academic papers via Google Scholar, published performance reports from Betterment and Wealthfront, SEC filings for registered investment advisors, and Morningstar data where accessible.
Good fit for: Students interested in personal finance and investing who want to engage with a practical, relatable question.
One upfront challenge with this topic is that robo-advisors serve a different kind of investor than a full-service human advisor, so you’ll need to narrow the comparison to small-balance investors for whom both options are realistically available. Betterment and Wealthfront publish performance reports and methodology documents that explain how their portfolios are constructed, rebalanced, and tax-loss harvested. Your research will involve analyzing those documents and then comparing the fee structures and reported returns against academic data on actively managed versus passive human-advised portfolios for similar investor profiles. An oft-overlooked variable is behavioral coaching, how human advisors often prevent clients from selling during market downturns, and whether robo-advisors can replicate that function through automated nudges.
13. The GameStop Short Squeeze: What January 2021 Revealed About Retail Investor Power and Market Structure
Sub-field: Market Structure / Retail Investing
Difficulty: Beginner to Intermediate
Why it’s interesting: The GameStop event was the first time retail investors coordinating on Reddit demonstrably forced institutional short-sellers to cover losing positions at massive cost, raising genuine questions about whether existing market rules were designed for a world where millions of small investors can move in coordination.
Suggested research question: To what extent did the January 2021 GameStop short squeeze expose structural vulnerabilities in equity markets that short-selling regulations had failed to anticipate?
Key methods/data sources: SEC’s official report on the GameStop episode (published January 2021), academic papers on short-selling mechanics via Google Scholar, news archives from the Financial Times and Wall Street Journal, case study methodology.
Good fit for: Students interested in markets, finance, or media who enjoy case-study-based research.
The SEC published a detailed staff report on the GameStop episode in October 2021, and it’s a primary source you should read closely as a prerequisite to this research. The report documents the sequence of events, the mechanics of the short squeeze, and the role of payment for order flow. The practice where brokerages route customer trades to market makers in exchange for compensation, which became controversial when Robinhood restricted GameStop trading at a critical moment. Your paper can use this as a case study in how regulatory frameworks designed for a pre-social-media market interact with coordinated retail behavior at scale. It shouldn’t focus on the legality of the incident, but what it revealed about gaps between how markets are assumed to function in regulatory thinking versus how they actually function when millions of small investors move together.
14. Algorithmic Trading and Flash Crashes: How High-Frequency Trading Contributed to Sudden Market Disruptions
Sub-field: Market Structure / Algorithmic Finance
Difficulty: Intermediate to Advanced
Why it’s interesting: The 2010 Flash Crash saw the Dow Jones drop nearly 1,000 points in minutes before partially recovering. Regulators later attributed it largely to the interaction of high-frequency trading algorithms, raising the unsettling possibility that automated systems can destabilize markets faster than any human can intervene.
Suggested research question: What conditions – in terms of market liquidity, order-book depth, and algorithmic trading volume – make equity markets most vulnerable to flash crashes?
Key methods/data sources: CFTC-SEC joint report on the 2010 Flash Crash, academic papers on high-frequency trading via Google Scholar, historical market data via Yahoo Finance or WRDS (if accessible through a school or library), case study comparison.
Good fit for: Students with an interest in how financial markets actually work mechanically, particularly those comfortable reading technical regulatory documents.
The 2010 Flash Crash is the anchor case for this topic, along with the CFTC-SEC joint report published five months afterward. The report traces how a single large sell order triggered a cascade through HFT algorithms that were simultaneously withdrawing liquidity from the market, amplifying the price drop rather than absorbing it the way traditional market-makers would have. Later research has identified other mini flash crashes, like in the Treasury markets in 2014, in currency markets in 2019, that show this wasn’t a one-time anomaly. Your paper should engage with what specifically makes algorithmic interaction dangerous, such as the speed differential between algorithmic and human reaction times, the way liquidity can evaporate almost instantaneously, and whether circuit breakers implemented after 2010 have actually resolved the underlying structural problem or just raised the threshold at which it triggers.
15. FinFluencers: How Financial Creators on TikTok and YouTube Are Shaping Investment Decisions Among Young Adults
Sub-field: Retail Investing / Financial Literacy
Difficulty: Beginner
Why it’s interesting: A 2022 CFA Institute survey found that nearly a third of retail investors under 25 cite social media as their primary source of investment information, yet there is almost no licensing or disclosure requirement governing what these creators can say, creating a significant consumer protection gap.
Suggested research question: What proportion of investment-related content posted by top financial influencers on TikTok contains claims that are misleading, unverified, or in conflict with established financial guidance?
Key methods/data sources: Original content analysis of a sample of FinTok videos, CFA Institute and FINRA reports on social media and investing, academic papers on financial literacy via Google Scholar, and basic qualitative coding methods.
Good fit for: Students interested in media, social behavior, or consumer protection who prefer hands-on, original research without heavy quantitative methods.
This is probably the most methodologically accessible topic on the list, since you can watch, code, and analyze a sample of financial content yourself. A content analysis involves selecting a defined sample (say, the top 20 finance accounts on TikTok by follower count), pulling a random sample of their videos, and classifying each video against a simple rubric, such as whether or not it includes a risk disclosure, if its claims are verifiable, and if the creator recommends a specific security without a license. The CFA Institute has published survey data on how young investors source financial information, providing quantitative grounding for this topic. Comparing the actual content against FINRA’s guidance for investment advice communication gives you a regulatory benchmark to evaluate against.
16. Tokenizing Real-World Assets: What Fractional Ownership Through Blockchain Could Mean for Everyday Investors
Sub-field: Blockchain / Capital Markets
Difficulty: Intermediate
Why it’s interesting: Major institutions, including BlackRock and JPMorgan, have begun experimenting with tokenizing real estate, bonds, and private equity on blockchain, which could allow retail investors to hold fractional shares of assets that were previously accessible only to institutions, though legal and liquidity questions remain largely unresolved.
Suggested research question: What regulatory and technical barriers currently prevent tokenized real-world asset platforms from scaling to retail investors in the United States?
Key methods/data sources: SEC statements on tokenization, academic papers on asset tokenization via Google Scholar, industry white papers from Securitize or Backed Finance, and policy analysis methodology.
Good fit for: Students interested in investing, law, or blockchain technology who want to research an emerging and underexplored area.
Asset tokenization converts ownership rights in a physical or financial asset like real estate, private equity, or fine art into digital tokens on a blockchain, which can then be traded in smaller fractions. BlackRock’s BUIDL fund, launched in 2024 on Ethereum, is a live example of a major institution doing this with U.S. Treasury securities, and it makes for a useful case study. Your research will focus on what the regulatory and legal barriers to retail access are, such as the securities law in the U.S. that requires most private investment tokens to be sold only to accredited investors, which excludes the majority of people the technology is theoretically meant to help. Reading SEC statements on digital asset securities and comparing them to how platforms like Securitize describe their compliance approach gives you the gap between promise and reality that makes for substantive analysis.
17. Neobanks and the Unbanked: How Digital-Only Banks Are Reaching — or Missing — America’s Most Financially Excluded Households
Sub-field: Financial Inclusion / Consumer FinTech
Difficulty: Beginner to Intermediate
Why it’s interesting: Chime, Varo, and similar neobanks have marketed themselves as solutions to financial exclusion by eliminating minimum balance requirements and overdraft fees, but FDIC data continues to show high unbanked rates among the very groups these companies claim to serve, raising questions about whether product design alone can overcome structural exclusion.
Suggested research question: How do the account features and eligibility requirements of major U.S. neobanks compare to what unbanked households identify as the primary barriers to opening a bank account?
Key methods/data sources: FDIC unbanked survey, neobank terms of service and fee disclosures, CFPB consumer complaint database, and academic literature on digital financial inclusion via Google Scholar.
Good fit for: Students interested in inequality, social policy, or the gap between what companies advertise and what they deliver.
The FDIC’s 2023 unbanked survey found that the most commonly cited reasons for not having a bank account included distrust of banks, unpredictable fees, and minimum balance requirements, which are all things neobanks explicitly advertise as solved. Your research task will be to check whether neobank product design actually addresses the reasons people give for being unbanked, and then to look at what the FDIC data shows about whether unbanked rates have meaningfully shifted in the populations neobanks claim to serve. The gap between what’s technically available and what people actually adopt is a recurring theme in financial inclusion research, and the CFPB complaint database gives you a ground-level view of where neobanks fall short for the customers who do sign up. Chime’s 2021 controversy around account closures, which majorly affected low-income customers who relied on it as their only financial account, is a useful case study in the limits of a private company serving a public-interest function.
18. Are FinTech Remittance Apps Actually Making It Cheaper to Send Money Across Borders?
Sub-field: Payments / Financial Inclusion
Difficulty: Beginner to Intermediate
Why it’s interesting: The World Bank’s target is to reduce the global average remittance cost to under 3% of the transfer amount, yet as of 2024, the global average still hovers around 6%, despite widespread claims from apps like Wise, Remitly, and Revolut that they have disrupted the space.
Suggested research question: For the ten highest-volume remittance corridors in the world, how do the total transfer costs of fintech remittance platforms compare to traditional money transfer operators like Western Union?
Key methods/data sources: World Bank Remittance Prices Worldwide database, platform fee calculators (Wise, Remitly, Western Union), academic papers on remittance markets via Google Scholar, comparative spreadsheet analysis.
Good fit for: Students interested in global development, immigration, or consumer finance who want a clear, data-driven research question.
The World Bank maintains a publicly accessible Remittance Prices Worldwide database with quarterly cost data for over 365 country corridors, and it’s the cleanest dataset available for this kind of comparison. You can use it to pull the average total cost of sending $200 along specific corridors and compare those figures against what Wise, Remitly, and Western Union charge for the same transfer at the same moment. The research question prompts you to look at how fintech competition has driven costs down sharply on high-volume corridors like the U.S. to Mexico, but lower-volume corridors in sub-Saharan Africa or the Pacific remain expensive even where fintech apps nominally operate. Your paper should distinguish between where competition has worked and where it hasn’t, and engage with the structural reasons for that difference.
19. Peer-to-Peer Lending Platforms: Do They Actually Serve Borrowers Traditional Banks Turn Away?
Sub-field: Alternative Finance / Credit Markets
Difficulty: Intermediate
Why it’s interesting: Peer-to-peer lending platforms like LendingClub launched with the explicit goal of providing credit to underserved borrowers at lower rates than traditional banks, but by the mid-2010s, LendingClub had shifted its funding model to institutional investors and tightened its credit criteria, calling into question whether P2P ever delivered on its inclusion promise.
Suggested research question: How have the borrower credit profiles and loan approval rates on major U.S. P2P lending platforms changed between 2010 and 2023, and what does the shift suggest about whether P2P has maintained its original financial inclusion mission?
Key methods/data sources: LendingClub publicly released loan data on Kaggle, academic papers on P2P lending via Google Scholar, CFPB lending data, and trend analysis using Excel or Python.
Good fit for: Students comfortable with datasets who want to tell a story about how a technology’s real-world impact diverged from its stated goals.
LendingClub released years of anonymized loan-level data publicly, and it’s one of the most useful datasets available for a high school finance researcher. You can download it from Kaggle and use Excel or Python to analyze how borrower credit scores, loan purposes, and default rates have changed over time. The academic literature documents a shift that happened across P2P platforms around 2015 to 2016, where platforms that started as marketplaces for individual lenders increasingly began facilitating loans funded by hedge funds and banks, which came with pressure to tighten credit standards. That shift is the core of your argument about whether P2P delivered on its inclusion promise, and you’ll need to apply multiple data science methodologies to comb through the data.
20. M-Pesa and the Case for Mobile Money: What Kenya’s Experience Reveals About FinTech and Financial Access in Low-Infrastructure Regions
Sub-field: Financial Inclusion / Mobile Money
Difficulty: Beginner to Intermediate
Why it’s interesting: M-Pesa launched in Kenya in 2007 and within a decade had accounts linked to over 96% of Kenyan households, making it one of the most studied examples of technology-driven financial inclusion in the world, and one that many other countries have tried and failed to replicate.
Suggested research question: What specific features of Kenya’s regulatory environment, mobile network infrastructure, and informal economy made M-Pesa successful, and why have similar models struggled to achieve comparable results in other sub-Saharan African markets?
Key methods/data sources: Academic papers on M-Pesa’s economic impact (including the widely cited Suri and Jack 2016 study in Science), World Bank financial inclusion data, Central Bank of Kenya reports, comparative case study methodology.
Good fit for: Students interested in international development, Africa, or how context shapes whether a technology succeeds or fails.
Tavneet Suri and William Jack’s 2016 paper in Science, which found that M-Pesa access lifted 2% of Kenyan households out of poverty by enabling consumption smoothing, is one of the most cited empirical studies in development economics. However, the paper’s methodology uses a quasi-experimental design based on geographic variation in M-Pesa agent availability, and subsequent researchers have debated both the causal identification strategy and the generalizability of the findings. Your paper can use Kenya as a benchmark to ask why similar mobile money deployments in Tanzania, Ghana, and elsewhere haven’t replicated the same results, which will require you to read Central Bank reports and IMF working papers from those countries alongside the Kenyan literature. The answer usually involves a combination of regulatory design, agent network density, interoperability between providers, and whether the system is integrated with existing informal financial practices.
21. Central Bank Digital Currencies: Comparing China’s Digital Yuan with the EU’s Digital Euro Project
Sub-field: Digital Currency / FinTech Policy
Difficulty: Intermediate
Why it’s interesting: China’s digital yuan is the most advanced CBDC pilot among major economies, with real-world testing in dozens of cities, but its design includes features, such as programmable expiry dates and government oversight of transactions, that raise privacy concerns that the EU’s digital euro project is explicitly trying to avoid.
Suggested research question: How do the design choices behind China’s digital yuan and the EU’s digital euro reflect each jurisdiction’s differing priorities around financial surveillance, monetary control, and consumer privacy?
Key methods/data sources: People’s Bank of China white papers on the digital yuan, European Central Bank publications on the digital euro project, academic papers on CBDC design via Google Scholar, and comparative policy analysis.
Good fit for: Students interested in international relations, monetary policy, or privacy and technology governance.
The People’s Bank of China has published white papers on the digital yuan, and the European Central Bank has published its own design documents for the digital euro, providing primarily material for this research. The digital yuan includes programmable features like expiry dates on government-issued funds and transaction monitoring capacity at the central bank level, while the EU’s design documents prioritize offline functionality and explicit privacy protections in response to European data governance norms. Your paper’s analytical core will be to unpack what these design choices say about each government’s relationship with its citizens, because a CBDC’s framing as a monetary policy tool, a surveillance instrument, or a financial inclusion mechanism changes what it looks like in practice. The Atlantic Council’s CBDC tracker gives you a comparative overview of all active projects, which helps you place these two in a global context.
22. Stablecoin Regulation: Why the U.S., EU, and UK Are Taking Different Approaches to the Same Problem
Sub-field: Cryptocurrency / FinTech Regulation
Difficulty: Intermediate
Why it’s interesting: The collapse of TerraUSD in 2022, which erased roughly $40 billion in value within days, pushed regulators in multiple jurisdictions to develop stablecoin frameworks, but the EU’s MiCA regulation, the UK’s Financial Services and Markets Act amendments, and proposed U.S. legislation take meaningfully different approaches to reserve requirements, issuer eligibility, and consumer protection.
Suggested research question: How do the reserve and disclosure requirements for stablecoin issuers under the EU’s MiCA regulation compare to proposed U.S. stablecoin legislation, and which framework offers stronger consumer protection?
Key methods/data sources: EU MiCA regulation text, U.S. Congressional stablecoin bill drafts, UK FCA consultation papers, academic and policy analysis from the Atlantic Council’s CBDC tracker, and comparative legal analysis methodology.
Good fit for: Students interested in law, policy, or how governments respond to financial innovation.
TerraUSD’s collapse in May 2022 is your origin point, because it’s what turned stablecoin regulation from a theoretical discussion into an urgent policy priority across multiple jurisdictions simultaneously. The EU responded first with MiCA, which came into force in 2024 and requires stablecoin issuers to hold 1:1 reserves in approved assets and caps daily transaction volumes for non-euro stablecoins. The UK took a narrower approach, bringing fiat-backed stablecoins into existing payment regulations rather than creating a dedicated framework. U.S. progress has been slower, with competing bills in Congress reflecting disagreements between banking regulators and crypto advocates about whether stablecoin issuers should be required to obtain bank charters. Reading the actual legislative texts and regulatory consultation documents lets you compare reserve requirements, redemption rights, and issuer eligibility criteria across all three frameworks precisely.
23. Open Banking in the EU vs. the United States: Why the Same Idea Looks So Different Across the Atlantic
Sub-field: FinTech Regulation / Banking
Difficulty: Intermediate
Why it’s interesting: The EU’s PSD2 directive mandated that banks grant licensed third parties access to customer account data with customer consent, creating a legal foundation for open banking, while the U.S. has relied on voluntary industry standards and market pressure, resulting in a patchwork system that the CFPB has only recently begun to formalize through rulemaking.
Suggested research question: What political and structural differences between the EU and U.S. financial systems explain why open banking has advanced through regulation in Europe but through market-driven data aggregation in the United States?
Key methods/data sources: EU PSD2 directive text, CFPB open banking rulemaking documents (Section 1033 of Dodd-Frank), academic papers on comparative open banking via Google Scholar, analysis from think tanks like the Brookings Institution and Open Banking Excellence.
Good fit for: Students interested in comparative policy, banking regulation, or transatlantic political economy.
Open banking is the principle that customers should be able to share their financial data with third-party apps like budgeting tools, loan comparators, or payment services without having to go through their bank’s proprietary interface. The EU made this a legal requirement through PSD2 in 2018, mandating that banks build standardized APIs giving licensed third parties access to account data with customer consent. In the U.S., the same outcome has been pursued mostly through screen-scraping, which banks have resisted aggressively. The CFPB finalized Section 1033 rulemaking in 2024 to create a U.S. open banking framework, and comparing that rule to PSD2 will be the methodological backbone of your paper. The question you should address is whether the difference in approach reflects regulatory philosophy, industry lobbying power, or the structural difference between a unified EU regulatory regime and fragmented U.S. federal and state banking law.
24. RegTech and Anti-Money Laundering: How Banks Are Using Automation to Meet Compliance Obligations
Sub-field: Regulatory Technology / Compliance
Difficulty: Intermediate to Advanced
Why it’s interesting: U.S. banks spend an estimated $25 billion annually on AML compliance: much of it on manual transaction review that produces high rates of false alerts, and RegTech firms argue that machine learning can dramatically reduce both the cost and the error rate, though evidence for this claim in peer-reviewed literature remains thin.
Suggested research question: What measurable improvements in suspicious activity report quality and false-positive rates have banks reported after implementing machine learning-based AML transaction monitoring systems?
Key methods/data sources: FinCEN SAR statistics, academic papers on RegTech and AML via Google Scholar, industry reports from firms like KPMG and Deloitte on AML technology, case study analysis of published bank implementations.
Good fit for: Students interested in financial regulation, compliance, or the practical limits of applying AI to rule-bound institutional processes.
U.S. banks file roughly 3.5 million suspicious activity reports per year, and law enforcement agencies have consistently said that a large proportion of those reports are low-quality or duplicative, generated by rules-based systems that flag anything that matches a simple threshold rather than systems capable of identifying suspicious patterns. RegTech companies pitch machine learning as the solution, claiming it can reduce false positives and improve the signal-to-noise ratio in SAR filings. Your research should engage with FinCEN’s published SAR statistics, which give aggregate data on filing volumes and categories, and then look for academic and regulatory assessments of whether ML-based AML systems have actually delivered on those claims. The analytical challenge is that most performance data is proprietary, so your paper will need to work with what’s available in regulatory testimony, academic pilots, and published case studies from vendors like NICE Actimize or Featurespace.
25. DeFi After Terra/LUNA: Can Decentralized Finance Survive Its Own Systemic Risks?
Sub-field: Decentralized Finance / Crypto Markets
Difficulty: Advanced
Why it’s interesting: The collapse of the Terra/LUNA ecosystem in May 2022 wiped out roughly $60 billion in value in under a week and triggered a cascade of failures across the DeFi sector, exposing the degree to which protocols marketed as decentralized and trust-free were, in practice, highly interconnected and concentrated, in ways that closely mirror risks that traditional finance regulation was built to prevent.
Suggested research question: What structural features of the Terra/LUNA ecosystem – including its algorithmic stablecoin design, yield incentive structure, and cross-protocol dependencies – caused a localized de-peg event to become a sector-wide liquidity crisis?
Key methods/data sources: On-chain data from DeFi analytics platforms (Dune Analytics, DefiLlama), post-mortem analyses from Chainalysis and academic researchers, SEC and CFTC enforcement documents related to Terra, academic papers on DeFi systemic risk via Google Scholar, and case study methodology.
Good fit for: Students with a strong background knowledge of crypto markets who are comfortable reading technical white papers and synthesizing across multiple sources.
The Terra/LUNA collapse is extensively documented in on-chain data, which is publicly accessible and quite granular. Platforms like Dune Analytics and DefiLlama let you trace exactly how liquidity moved out of Anchor Protocol, how the UST de-peg triggered a redemption spiral in the LUNA minting mechanism, and which other DeFi protocols held enough Terra exposure to be dragged down in the aftermath. Your paper should work through the mechanics of how an algorithmic stablecoin maintains its peg and why the specific design of TerraUSD made it vulnerable to a bank-run dynamic at scale. The broader research question is whether the risks exposed in May 2022 are fixable at the protocol level through better reserve design or circuit breakers, or whether they’re inherent to a system that removes the lender-of-last-resort function that central banks provide in traditional finance.
Horizon Academic Research Program
If you’re looking for a competitive mentored research program to conduct research in FinTech, consider applying to Horizon’s Research Seminars and Labs! This is a selective virtual research program that lets you engage in advanced research and develop a research paper on a subject of your choosing. Horizon has worked with 1000+ high school students so far and offers 600+ research specializations for you to choose from. You can find the application link here!
Image source: Horizon Academic Research Program