> But one thing strikes me as odd about this whole arrangement. The better a business performs, the quicker it is able to repay the loan, and the higher its effective APR becomes! It's essentially a prepayment penalty in disguise. So you'd better make sure the loan doesn't help your business too much, or you'll end up getting hosed by the loan fee.

Yes, this is right -- if your business suddenly starts doing a lot better, your effective APR will be higher.

What we heard overwhelmingly from customers, though, is that the downside risk of credit obligations they can't meet (liquidity problems are asymmetrically damaging!) substantially outweighs the theoretical "risk" of a higher effective APR caused by significant outperformance in the business. (Stated differently, we're taking the risk of your business underperforming, in return for you paying us back somewhat faster -- but still at a capped rate -- if things go better than you expect.)

I think the model we have makes more sense for most businesses. But there's no dogma; we'll certainly revisit this over time if we find that a lot of customers seek a different risk profile.

I hope to one day open some businesses and be cursed with high performance and pay a penalty to stripe.

I wonder if the correlation might be age, since it seems to me that younger people try to avoid bullshit and seek out authenticity (giving or receiving) moreso than older generations, which were embedded in more traditional business models across all industries. (I could be completely wrong about this being related to age or generation, though; totally guessing. Authenticity does seem to be a universally growing trend.)

I thought the comment was apt.

E.g. You charge a 10% upfront fee for the loan, and repay at a 10% rate, from the perspective of the company they’re actually paying 12.9% to Stripe during repayment.

If you could pay just the 10% or even if you just got a .5% reduction to 12.4% overall going to Stripe during the repayment period, it becomes more of a win-win and a nice customer loyalty type bonus.

(I find myself guilty of fanboying over certain musicians and founders sometimes, too, but keep in mind people on the receiving end of it sometimes feel automatically uncomfortable, even if you're saying things they like to hear.)

Stripe reduce their margin slightly on payments - but it makes for a much more compelling sell to potential borrowers.

Let me go back 12 months, pretend I took a loan of $X amount at Y date and simulate the repayment based on a customers actual historical data.

Interesting. From this perspective, Stripe is actually getting into the factoring business: https://en.wikipedia.org/wiki/Factoring_(finance)

As the borrower this loan is significantly de-risked since there aren't any dreaded monthly payments you must make.

Here's my attempt at a quick interest calculator for Stripe Capitals loans: https://docs.google.com/spreadsheets/d/1RH9PpJ9kdB7X4xcZvqb7...

https://connect.calcapp.net/?app=atzggn

Also, you don't have to duplicate the spreadsheet to make it editable, you can just use it straight away.

(Disclaimer: I created it with Calcapp, which is a SaaS product I built.)

I don't agree with how you say that 10% is the Quoted Interest Rate, because Stripe isn't doing that. They are saying they have a fixed fee. "Interest rate" means it compounds.

Ahh you caught the mistake with the interest periods! Thanks!

How does that make any sense?

But you're probably right in this case - Stripe likely include their fee in the monthly payments - updated the sheet.

Appreciate the feedback!

In short, you’re charging the highest interest rate to those whose businesses do better than you expected and the lowest rate to business who underperform your projections.

As such, you’ve created a financial incentive to underperform, and are entering this business of lending money by literally penalizing the least risky borrowers and rewarding the riskiest borrowers.

When you incentivize something you get more of it.

Your borrowers know their business better than you, and you’ve provided them with a mechanism to exploit that asymmetry in information.

You sort of have to assume that your incentives will be effective: the customers who have the most reason to believe they are growing faster than baseline will avoid your product and those who are pessimistic will embrace it, greatly increasing your default rate.

if you have reasons to believe that you can grow faster than "the baseline", it makes sense to use that evidence to borrow from a traditional bank. If you are undertaking a risky business move, the bank will not want to lend you the loan, and this model stripe has is going to be the next best thing.

Having a stripe loan would make me, as a successful business owner, make that move much sooner.

All that gives pretty good dataset for building predictive models that tell how the business will do. These models could be actually better at predicting the future than individual business owner (on large population). Stripe has information about the past performance of many similar businesses while the owner just knows about his.

Also need to remember that business owner is not just optimizing for low APR. He should be still making money through the sales. So hopefully in total there’s still the financial incentive to have as good sales as possible.

What is crazy, is how eerily similar Patrick's response is with the Payday Industry's response about why Payday Loans are good for consumers.

>the downside risk of credit obligations they can't meet ... substantially outweighs the theoretical "risk" of a higher effective APR.

"The $15 cost of a $100 payday loan also pales in comparison with the lost income when a car is out of commission and a job lost. Good payday lenders clearly disclose their loan terms and conditions, including the dollar amount of any fees and the APR."[1] Payday loans are clearly bad for the consumer, even as much as the industry tries to defend it.

[1]https://www.americanbanker.com/opinion/why-payday-loans-are-...

The better analogy would be a fixed-cost loan with variable installments based on your income- a much, much friendlier loan structure for consumers.

What makes payday loans unaffordable is their structure, more than their cost. In California, a typical payday loan goes like this:

1- You write me a $300 check and date it for two weeks from now (when you get paid)

2- I give you $250 in cash

3- Two weeks later, I cash the check

If you had to borrow $250, what are the chances that you have $300 left over on your next paycheck? Zero. So really, it's:

3^- You come back to the store and say "don't cash that check, I'll get hit with a $25 overdraft fee."

4- I say "ok, give me $50 and I'll move your due date back 2 weeks."

5- You say "phew, thank goodness!"

6- Two weeks later, goto 3*

So the one-time payment is what makes it horrible. Even if they charged 0% APR and all you had due was $250, you'd still be hosed. An installment loan, though, where you pay $50 every two weeks for N months is clearly better, as proven by step 4 here.

I don’t think this is the case. Payday loans compound interest, don’t they? You end up owing more money the longer you take to pay them off. With this, you only ever have to pay back the advance and fixed fee no matter how long it takes.

SBA loans have better rates, but require more work. I particularly like the American Express model, where they extend credit to pay a merchant invoice directly. You repay the credit in 30/60/90 days and you pay 12% (which used to be 9%).

California has light regulation in the form of disclosures similar to those for consumer credit. CA is also the only state that requires any licensing but the license is for the company, not for any of the brokers, and is super easy to obtain.

But I worry that it's vulnerable to bad people.

Got a hefty personal bill coming up, and you're pretty confident your SaaS business will be dead within 6 months but Stripe has no way of knowing based on current numbers?

Thinking of getting a divorce (startup or marital)... maybe time to grab the largest advance possible before your soon-to-be ex(-co-founder) realises?

The problem is that even if only 1% of Stripe's customers might match these kinds of profiles, that same 1% might be taking 20% of the Stripe Capital advances.

You worry too much for them, that's why they charge a hefty interest fee. A certain % of non-payments or partial payments are built in.

> Decent fraud model

Yeah, I would think so.

In a regular loan, the bank makes you bend over backwards to prove you’re risk free, and then charges you interest to give you the loan anyway, and you’re on the hook to pay back a constant amount whether you’re under or over performing, at risk of going bankrupt.

* What does the entry level look like? This appears to be a decent way to launch a hobby into a business, but what is the income you're looking for? Does my app need to gross 1k/mo before it's an option?

* What are the term limits? Or what if the app is shutdown, or underperforms?

This appears to be a good way for people to launch hobby projects into full businesses, I was just wondering if that was indeed the case.

You could charge a flat interest rate but require the customer to pay you back a minimum of x% of their stripe volume per month.

Allow the customer to pay a higher percent of volume if they choose, and reward them for paying you back by lowering the effective rate.

I somewhat agree with the point the GP made, the fee structure is very much an early repayment penalty in disguise.

I mean, if you have several payment methods on your website, one resource to improve your APR could be to make more promiment other payment methods for the duration of the loan.

For businesses that expect a slow payback, the APR isn’t nearly that much.

I’d use it to reserve EC2 instances, for example. I get a 40% cost reduction on AWS over 3 years, and I’d expect to take that long to pay back the capital. The APR would be fine in that case, and I make 20% more than I would otherwise.

That's a great idea!

Just be honest, you don't want to deal with the regulatory issues that surround offering a fair compounded interest rate.

Or, people who have a good history of selling with Stripe, but, lack the credit rating, history or collateral to use other lenders.

It's a very carefully and cleverly thought out payment model, which leads me to believe they knew exactly which consumer base they were going after.

It seems novel for folks outside the industry but businesses are bombarded by MCA adverts on the daily and have been for years.

We do think this is a significantly improved overall user experience—it works straight out of your Stripe account. It’s automatic, based on your sales. The cost of the loan is a single fixed fee that adjusts to the loan amount paid over the course of the loan—there is no interest rate or additional fees. The effective APR is dependent on how long it takes to repay the loan.

Wow, that's a whopping 10-15% p.a. interest rate. Time to mint the money has come for Stripe, yeah?

In the early stages of an internet-enabled, distribution-bound company, 1 dollar reinvested (say in ad spend) can net you 1.5-2 dollars in increased sales. So even at a quite high APR, it's still a net win for many companies to take that debt on.

Obviously you'd rather take on debt at a lower APR, but if your risk of default is high, nobody is going to give you a low APR.

I didn't read the exact terms, but based on the landing page it seems like your downside is capped, which is a big thing. I.e. if you suddenly stop making sales, you don't repay them . (not sure if there's a clause that makes you repay if you never make sales for a couple of years after taking out the loan)

You would need to consult an appropriate scholar to be sure, but it could be to be compatible with Islamic finance, which prohibits charging interest. Some banks do mortgages structured around fees rather than interest too.

Some Jews won't push a button in an elevator on the Sabbath, but it's fine if its programmed to stop at every floor.

18 miles of translucent wire stretches around NYC called an eruv:

On the Sabbath, which is viewed as a day of rest, observant Jewish people aren't allowed to carry anything—books, groceries, even children—in public places (doing so is considered "work"). The eruv encircles much of Manhattan, acting as a symbolic boundary that turns the very public streets of the city into a private space, much like one's own home. This allows people to freely communicate and socialize on the Sabbath—and carry whatever they please—without having to worry about breaking Jewish law.

Loopholes that violate the spirit of the law but technically fit the letter of it make sense in doing your taxes, but it seems extremely risky to take that approach with a god. "Haha guys, very clever, you got me" doesn't seem like the end result for a being known for smiting.

Regarding the loopholes, well, in Rabbinic Judaism the notion is that there simply can't be any loopholes there. Unlike laws written by humans, the word of God is considered axiomatically perfect and flawless - if you find a "mistake", that can't be a mistake, that's fundamentally impossible, God does not make mistakes. There simply can't be an accidental hole or a mistaken omission. The letter of the law is considered literally divinely perfect, every nuance or "loophole" is considered as intentionally placed there by an all-knowing being that knows all the future consequences of that nuance or "loophole". If God wanted the law to be slightly different, He obviously could and would have made it different, but He did not, so it should be interpreted exactly as written.

Can I (setting aside the obvious secular punishments for a moment) murder someone with a bomb activated via one of these switches?

In the case of murder, the result and intent is prohibited, so no matter how you get there, that's a sin.

In the case of having a fire on Shabbat, there's nothing wrong with wanting a fire or having a fire, the result is not prohibited or a moral wrong; but in that particular time and place you are prohibited to light a fire.

A hundred years ago the standard solution was to simply have someone else (non-jewish) do it, e.g. local gentile boys used to spend saturday mornings doing minor chores in jewish households for some tips. Nowadays it is done with technology. Murder is inherently wrong, so having someone else do your murdering for you is also wrong, but these actions are merely prohibited for you personally so having someone else (or something else!) do it in your stead is considered just fine.

That sounds like a bad incentive. If a shabbos goy decided to convert he would have to stop being paid, which means that you would be paying the shabbos goy to not convert.

Judaism is for the Jewish people, and while some individuals may join (e.g. in marriage) the Jewish people, from the perspective of Judaism it's perfectly okay if all the non-Jewish peoples (i.e. most people in the world) stay non-Jewish forever and their relationship with God is different.

I am going to set up a switch that buys you a hamburger delivery. But it has a 1 in 1000 chance of stealing 1000000 dollars. Are you committing a sin when you push the switch because you are hungry?

Seems to me like they're targeting businesses that do not meet your assumption of high sales volume.

This is not true in my experience. A lot of people just wing it through life without doing or being able to do any calculations. Not that it excuses them.

I can see that. But put another way, it's a flat fee. No disguise.

I see issues with you being able to get better rates elsewhere, but by calling it a fixed fee instead of using an APR with fine print saying there's a prepayment penalty, I see the OPPOSITE of disguise.

Not everyone has a credit card with a $25,000 limit on it. That really is the point of factoring future receivables like this. Stripe has visibility into your income that a traditional lender does not have, and can use this data to make loans that traditional banks and credit card companies would not make. That has value to people with more limited credit options than you apparently have, and is well worth the fees that Stripe is charging to the businesses that need it.

Otherwise you could switch 99% of your payment processing off of Stripe and pay them a trickle of your sales over a decade.

For example, I wouldn't be surprised if they legally tie repayment to your actual business sales (or any sales conducted through a qualified payment processor, comparable to Stripe; some legal structure that catches this), not just your Stripe sales. In the pitch they only mention repayment directly coming out of your Stripe sales, because that simplifies the pitch. Otherwise it exposes them to an extremely obvious angle for being taken to the cleaners both by scammers and plain desperate business operators.

They lend a business $100,000. That business immediately switches 99% of their processing to someone else. Repayment occurs at a comical rate - forever. There's no chance they left themselves open to something so obvious, it'd never get past even a mediocre lawyer.

Edit: what you would do is build a requirement into the loan contract that you must pay out of all sales conducted by your business and or all sales handled by any payment processor (I'd go with all sales, if I were Stripe), not just sales that go through Stripe.

A slight variation of that, would be that Stripe is granted the legal right to audit your business first, in the event of a certain % drop in sales / repayment. The second step, if they catch you evading them in the audit, is that the contract stipulates they have the right to collect from all business sales. The point of this step-based process, is to avoid constantly throwing a serious flag at operators on modest drops in sales. So if sales - repayment - drops by a modest N%, they do nothing initially, perhaps send a friendly message to check in; if sales (supposedly) drop dramatically (either short-term or longer-term), they ramp up their approach accordingly including becoming more aggressive in talking to you and possibly including an audit (business review) request. Keep avoiding them, it becomes a serious legal matter, as it would with any major creditor.

In any large scale financing business like this, the creditor must be very well equipped to pursue legal remedies against the borrower. You can bet that Stripe has built in an angle to pursue attempts to evade them, in small ways and up to entirely ghosting them on the loan.

I assumed the primary value was tying debt service to revenues at a micro level.

I assumed the secondary value was that -- since Stripe had intimate knowledge of your store, revenues, cash flows, correlations to customers -- that they could better price your loan and give you a more competitive rate on the loan.

Comparatively, 20% APR isn't so bad.

50-60% would violate most if not all usury laws, unless you are a bank. Even in my state the usury is capped at 18%, so to even charge 20% you have to be a bank or its an illegal loan.

>Usury laws typically apply to consumers, not loans to businesses.

That may be a fair characterization, but its definitely a jurisdictional issue state to state. I know some States for sure exempt commercial loans, others exempt loans over certain amounts, and some exempt loans to certain businesses by industry (but not a outright exemption on commercial loans).

Obviously Stripe has a good public image, but unfortunately with SV unicorns like Uber, AirBnB, CoinBase (almost any cryptocoin/blockchain start up), its not unfair to say we should just assume business practice comply with the law.

It's simply a side effect of the non-typical structure of the loan.

The loan structure is a new offer for the market an aligns with the customer's business needs for the first time ever. Saying you can get another loan elsewhere is incorrect for a certain segment of businesses.

Talking of APR comes from a saver mindset on a fixed budget, which is what people do, not corporations. Corporations focus on generating money first, and optimizing last.