10% APR loan doesn’t mean the lender is making 10%
I recently read a report from the WSJ about the “best low-interest personal loans in 2026.” It was published over a month ago. The article talked about the most competitive rates you could get when applying for a personal loan in the United States.
With this, you can:
Choose a personal loan with a low interest rate to reduce your total borrowing costs.
Compare three to five lenders to find suitable rates and terms.
Consider fees, which could erode potential savings.
Among all the listed projects in the article for comparison, the best APR that keeps coming up for loan capital starting from $1k is up to 6.53 ~ 40%.
In my mind, I’d thought for this particular loan interest that the lender’s gain on my $1,000 loan is around $65.30 to $400.
But I was completely wrong.
Though unsecured credit is one of the important segments in global finance, in fact, global fintech revenues surpassed half a trillion dollars in 2025, growing 22% year over year.
Among the sectors that broke out and contributed a larger % of these profits are payments, and the one trailing behind is lending, which consists of unsecured lending, overdrafts, secured lending, BNPL, among others.
So if the lender isn’t making the most of all the money, where does the money go? Who gets a % of the lender’s profit, and why do they need to?
The reality behind every loan that a lender gives, underneath that transaction is an enormous financial machine.
Someone has to provide the capital.
Someone has to determine whether the borrower can repay.
Someone has to verify their income.
Someone has to move the money.
Someone has to service the loan.
Someone has to make sure lender pay back or face consequences in case of nonpayment
And the lender needs to price in the possibility that some borrowers will never pay it back.
Every layer adds cost.
Each step the money your lender gives to you passes through before landing on your balance and afterward.
- Capital sourcing
The first critical question for lending money to customers isn’t always who you will lend to. It’s always the source of capital you want to lend out in the first place.
Imagine a neobank wants to originate at least $1,000 × 10,000 loans. That’s $10m.
That means they need $10m for the lending capital
Most projects provide this money from their own balance sheet, which might be a VC-raised fund or revenue they’ve accumulated over time.
In this situation, the protocol is taking the risk.
Warehouse financing: A lender can obtain a credit facility from a bank/institution.
The institution is the risk taker here, so they add some APR on the money for the project to pay back.
They might say, “We lend you $10m at 7% APR.”
For any loans they give out, it will almost always be above this since it has become the benchmark for the capital.
When they give you loans at 17%, there’s a default deduction of 7%, leaving the project with just 10%.
- Underwriting
Now we've got the capital; who do we lend to?
This is the decision-making machine behind the loans that lenders give out.
So if they’ve got 100k users and just want to lend to 10k > 10% of their userbase, underwriting is what shows the project who is really worthy of getting these loans.
With the probability of whether they can actually pay back
What’s next after this is credit scoring
So when the lender puts all your info on the model, it scans through your information, credit history, debts, fraud signals, probability of default, etc.
This way, the lender can assess properly who they are giving the loans to and if they can really pay back or not based on the available information.
3. Loan origination
Once they are done with the necessary risk assessment, it’s about time you get paid.
At that point, the project created the loan. The origination system automatically generates the necessary parameters for each loan it originated.
Principal, APR, term, loan agreement, etc.
4. Servicing
Once the loan exists, it just unlocks work for another segment: servicing.
It manages the loan for the period of time you’re owing. The system scans payment due dates, fees, payment status, etc., and keeps in touch with the borrower and reminds them to pay up.
It’s basically the operating system of the loan. A lender can originate a loan and then have another company handle servicing.
5. Compliance
Just because you’ve got capital doesn’t automatically mean you can do whatever you want and lend to anybody.
There’s regulation and rules guiding your operation.
Who you lend to.
Disclosure.
Consumer protection.
Anti-money-laundering requirements, etc.
Just like in the case of servicing, a company can employ a lawyer to make sure everything was done in compliance with local rules, and also worthy nothing this rules differ from region to region too.
6. Collection
Now let’s say you didn’t pay the loan back.
The lender doesn’t automatically assume they won’t get the loan back after they’ve tried every way to contact you with no point of paying them back.
The lender attempts to recover the debt according to applicable rules. They might be able to recover the full amount, but may have some % unrecovered.
Most of the time, projects again use third parties to do this for them.
This is just a piece of the iceberg that comes to mind. There’s still some elements like fraud protection, risk management, etc.
Every segment is costing a % of what a lender gives out, so every time they give out a loan, they need to share the profit evenly with partners and projects that help in the process, which in a way makes getting credit costly in traditional finance.
In fact, the New York Fed analyzed 550 million monthly credit-card accounts representing around 90% of the U.S. credit-card market and found that credit-card operations have exceptionally high operating expenses of roughly 4–5% of balances annually.
Stablecoins invent a new wheel for credit.
Stablecoin is by far one of the great use cases of crypto as they move money 24/7 and are pegged to the dollar 1:1.
Stablecoins move faster, settle faster, and can be deployed programmatically. The financial system can potentially operate with less friction.
Stablecoins are only one half of the equation. The other half is what happens to that capital once it is onchain.
This is where vaults become powerful.
Instead of capital sitting idle, investors can deposit stablecoins into an onchain lending strategy.
The vault can define:
where capital goes
what assets qualify
what borrowers qualify
how liquidity is managed
how yield is distributed
how risk parameters are enforced
This isn’t a theory on paper. Even Delphi’s report last month reported: “This category is already bigger than most people realize today. Curated lending vaults hold almost $9B, up more than 50% over the past year even as overall DeFi lending TVL fell 36%. Counting every kind of vault structure, S&P put total deposits near $131B in April, up from $24B three years earlier.”
Vaults solve the capital allocation problem but didn’t solve the big elephant in the room: knowing fully the customers you want to lend the money to.
That’s where we come in, @cr3dentials.
Instead of every project building this infrastructure from scratch, they can tap into our zkTLS-powered verification infrastructure specifically designed for credit and lending use cases.
With zkTLS, users can verify real-world income, cash flow, or reputation from digital sources like bank portals, gig platforms, creator dashboards, and payment processors without handing over logins or screenshots, and we selectively reveal data points.
The user’s actual financial details stay private, and no direct API integrations are needed from the data sources.
Neobanks can currently verify customer earnings from platforms like Upwork, Stripe, Uber, YouTube, and Shopify, with more integrations expected in the future.
And this is just a part of what we do. We also handle recourse, enforcement, and repayment for our clients, and scaling up the system as we move. The final goal is to become embedded finance for neobanks.
It’s even better: we are focusing on the niche individuals that make the internet tick, the traditional finance neglected here. They don’t need any credit score for creditworthiness. As long as they earn from somewhere, they are eligible, given they meet the terms and conditions of the lenders.
Sum up, with the lower capital cost with stablecoins and vaults, the cost of credit becomes can be cheaper.
lower cost of capital + lower cost of underwriting + lower operating costs = room for lower borrowing rates.
If you’re building unsecured credit using stablecoin rails,
We want to work with you.
DM us 🫵