| Course | FI310 FinTech Principles and Concepts |
|---|---|
| Unit | Unit 4 |
| Paper type | Digital lending case study |
| Length | About 1,150 words, 7 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | BS in Finance |
| Updated | October 2026 |
Free sample paper for FI310 Unit 4
Eight Minutes or Three Weeks: A Unit 4 Digital Lending Case Study of Who an Automated Model Approves
[Student Name]
Purdue University Global
FI310: FinTech Principles and Concepts
Unit 4 Digital Lending Case Study
[Instructor Name]
[Date]
The online lender, the community bank, the applicants and every figure are composite; the articles and book cited are real.
Background: The Online Lender
The lender offers small-business loans of $10,000 to $150,000, repaid by automatic weekly debits over six to eighteen months. An applicant links the business's bank accounts, and the lender's model pulls twelve months of transactions. It scores deposit consistency, the size and frequency of balance troughs, existing debt payments visible in the account, and the owner's personal credit score as one input among several. Most decisions arrive in about eight minutes, and approved funds land the next business day. Prices are higher than a bank's: an annualized rate of 28 to 45 percent once fees are counted.
Background: The Community Bank
The bank's small-business loans start at $50,000 and require two years of tax returns, financial statements, a personal guarantee and often collateral. A loan officer meets the owner, visits the business and writes a credit memo for a committee. Decisions take two to four weeks. Rates run 9 to 12 percent.
Applicant 1: The Food Truck
A food truck owner with three years in business wants $40,000 for a second truck. Deposits are steady from March to October and thin in winter, when balances fall near zero twice. Her personal credit score is 640 after a medical debt from three years ago.
The online model scores her troughs harshly and approves only $18,000. The community bank's officer, reading her tax returns and talking to her, sees a profitable seasonal business with a clear plan, but the bank's minimum is $50,000 and her score is below its usual cutoff. She is declined by the bank and partly approved online at a high price.
Applicant 2: The Accounting Practice
A two-partner accounting practice wants $480,000 to buy its office building. Its deposits are large and steady, with a deep balance every April after tax season. The online model approves its $150,000 maximum in minutes, but the term is eighteen months, far too short to finance a building. The bank approves $480,000 on a twenty-year mortgage at 9.5 percent. The bank's slower process produces the right product; the fast one produces the wrong one.
Applicant 3: The New Landscaping Firm
A landscaping firm founded nine months ago wants $25,000 for equipment. Its owner has a strong personal score but only nine months of deposits. The online model, which needs twelve months, declines automatically. The bank's officer is interested, but the bank requires two years of returns and declines too. Neither process can see a new business well.
The automated model sees cash flow that a loan officer must ask for, and the loan officer sees purpose and context that a model cannot read, so the food truck is underserved by both, the accounting practice is overserved by the model and served well by the bank, and the young firm falls through both.
Comparing the Two Processes Side by Side
The two processes differ in what they look at, how long they take and what they cost. The online model reads twelve months of bank transactions and a credit score, never meets the owner and never asks what the money is for. Its speed comes from that narrowness: a decision based on data that is already digital can be made by a program in minutes. The community bank reads tax returns and financial statements, which describe profit rather than cash, and adds a conversation and a site visit, which describe purpose and character. Its slowness comes from that breadth.
Neither set of information is complete. Bank statements show money moving in and out but not whether the business is profitable after owner draws and taxes. Tax returns show profit but lag by a year and can be shaped by depreciation choices. A conversation reveals purpose but also invites the loan officer's own biases. A lender that combined the two, using cash-flow data to screen quickly and a short conversation to understand purpose, could serve all three applicants better than either does today.
The costs differ as well. The online lender's rate of 28 to 45 percent reflects its higher losses, its funding costs and its speed. For the food truck, borrowing $18,000 at 35 percent for a year costs about $3,500 in interest and fees; the same amount at a bank's 11 percent would cost under $1,100. Speed is worth something, but borrowers often do not see the full price because it is stated as a weekly debit rather than an annual rate.
What the Evidence Says
Research on fintech credit scoring has found that digital information, such as the data an online lender collects, predicts default as well as or better than traditional credit scores and can complement them, reaching borrowers with thin credit files (Berg et al., 2020). Cash-flow underwriting is one form of that. But speed and data do not fix product design: a short, high-cost loan is the wrong instrument for long-lived assets, and the model has no way to say so. Technology-based lenders in mortgage markets processed applications faster without higher defaults, but they did not change which loans made sense for which borrowers (Fuster et al., 2019).
Who Is Turned Away, and Why It Matters
The applicants the model declines or underserves share a pattern: seasonality, short histories and past credit problems that do not reflect current business strength. Some of these are exactly the borrowers a loan officer might approve after a conversation. Because automated models are trained on past repayment, they can reproduce past patterns, and a lender should test whether its model treats similar businesses differently for reasons unrelated to repayment.
What a Better Process Might Look Like
A practical improvement for the online lender would be to recognize seasonal patterns rather than penalize them. A food truck whose balance falls every January is not a worse risk than one with flat deposits if its summer deposits reliably cover the gap, and twelve months of data are enough to see that pattern. For the bank, a smaller-loan product with standardized documents, a credit score cutoff that considers explained past problems and a decision within a week would capture borrowers like the food truck that it now sends to expensive alternatives.
Recommendations
For borrowers, compare the total annualized cost, not the weekly payment, and match the loan's term to the asset's life. For the online lender, add a manual review path for seasonal businesses with strong peak-season deposits, and stop offering eighteen-month terms for real estate purchases. For the community bank, consider a smaller-loan product with a faster, data-assisted review, since its minimum and documentation requirements turn away viable firms (Chishti & Barberis, 2016).
References
Berg, T., Burg, V., Gombović, A., & Puri, M. (2020). On the rise of FinTechs: Credit scoring using digital footprints. The Review of Financial Studies, 33(7), 2845-2897. https://doi.org/10.1093/rfs/hhz099
Chishti, S., & Barberis, J. (2016). The FinTech book: The financial technology handbook for investors, entrepreneurs and visionaries. Wiley.
Fuster, A., Plosser, M., Schnabl, P., & Vickery, J. (2019). The role of technology in mortgage lending. The Review of Financial Studies, 32(5), 1854-1899. https://doi.org/10.1093/rfs/hhz018
What the FI310 Unit 4 assignment asks for
The fourth FI310 unit asks for a case study of an automated lending process. Describe the digital lender's data, model inputs, decision time and pricing, then do the same for a traditional lender as a benchmark. Run named applicants through both, with each applicant's history, request and outcome. Compare what information each process uses and what it cannot see. Bring in research on alternative data in credit scoring and on technology-based lenders, and be exact about the limits of its findings. Discuss which applicants are underserved and why that matters for fairness. Price the options in annual terms so borrowers' costs can be compared. Finish with recommendations for both lenders and for borrowers, and cite the studies in APA 7.
How the FI 310 Unit 4 example is put together
The sample's online lender reads deposits, balance troughs and debt payments, approves in about eight minutes and charges 28 to 45 percent annualized; the bank wants tax returns, a visit and a committee and charges 9 to 12 percent. The food truck, with a 640 score and winter troughs, asks for $40,000 and gets $18,000 online after the bank declines. The accounting practice wants $480,000 for its building and is offered the online maximum of $150,000 over eighteen months, while the bank writes a twenty-year mortgage. The landscaping firm falls through both. The highlighted sentence summarizes what each process sees and misses, and a cost section shows $18,000 at 35 percent costing about $3,500 against under $1,100 at a bank.
Where the points sit in the FI310 Unit 4 rubric
Lending case studies in FI310 earn points for describing each process precisely and testing it against concrete applicants. Graders look for the model's inputs and limits, a fair benchmark, outcomes that follow from each applicant's facts and research used for what it actually found. The rubric rewards spotting mismatches between loan term and purpose, attention to borrowers a model may exclude, and costs compared on an annual basis. Recommendations should address more than one party. Case studies that celebrate speed without testing who is served, rely on a single applicant or misstate research findings receive fewer points. Clear figures and cited studies make the comparison persuasive.
Common FI310 Unit 4 mistakes, and how to avoid them
Weak Unit 4 case studies describe an online lender's speed and stop there, never testing whom the model serves. Others compare a digital lender with no benchmark at all. Applicants are often generic, so outcomes cannot be traced to their facts. Research on alternative data is cited as proof that automated lending is fairer, which overstates what the studies found. Many papers quote weekly payments instead of annual cost, the very habit that hides a high price. If your scenario uses consumer loans, invoice lending or a buy-now-pay-later product, a writer can test that model against applicants built from your prompt.
Get an FI310 Unit 4 example written to your instructions
Share the lender and applicant details your Unit 4 case supplies, or describe a small business you know that borrowed recently, plus the FI310 rubric and instructions. The case study is built on those applicants, outcomes explained by process and offers priced on one annual basis. Nothing is charged for the first custom sample; it normally arrives in 24-48h. The paper above is an original model document written by our desk, not a submitted student paper and not an official Purdue University Global document.
More FI310 papers and BS in Finance samples
- FI310 Unit 1: Discussion Post on Mobile Deposit
- FI310 Unit 2: Mortgage Closing Process Comparison
- FI310 Unit 3: Phone Wallet Payments Analysis
- FI310 Unit 5: Platform Business Analysis
- FI310 Unit 6: Seminar Reflection
- FI310 Unit 7: Risk Assessment Memo
- FI310 Unit 8: Technology Adoption Brief
- FI310 Unit 9: Financial Inclusion Analysis
- FI310 Unit 10: Sector Outlook Report
FI310 Unit 4 questions, answered
Where can I find a free FI310 Unit 4 sample paper?
This page holds a full FI310 Unit 4 digital lending case study comparing an eight-minute online lender with a community bank across three applicants, with recommendations and APA 7 references.
Why does the online model penalize the food truck?
Because it scores balance troughs, and a seasonal business shows deep ones every winter. Twelve months of deposits would reveal the pattern, but the sample's model treats the dips as warning signs rather than as a predictable cycle covered by summer sales.
What is wrong with the accounting practice's online offer?
The term. An eighteen-month loan cannot sensibly finance a building that lasts decades, because the payments would be far too large for the practice's cash flow. The bank's twenty-year mortgage matches the asset, which is the point the sample uses it to make.
Does the case say automated lending is unfair?
No. It says each process sees something the other misses. The model reads cash flow quickly, the loan officer reads purpose and context, and the young landscaping firm is invisible to both. Fairness testing is recommended, not assumed to fail.
How should borrowers compare a fast loan with a bank loan?
By total annual cost and by whether the term matches what the money buys. A weekly debit can look small while the annual rate is several times a bank's, so the sample converts both to annual figures before comparing them.