Mandy from accounts and borrowed time
Shopify Credit has a moneyball kinda vibe
Picture a merchant. Call her the kind of founder Shopify built its brand on: the rebel, the arm-the-little-guy hero of a thousand keynotes. She sells homeware, or skincare, or something your aunt bought at Christmas and genuinely liked. In early 2025 her landed costs went sideways. Tariffs due at the border before a single unit sold. Freight that never quite came back down. An inventory buffer she’s carried since the great supply chain trauma taught everyone to over-order. Her margin got thinner and her cash cycle got longer, and one Tuesday, on the worst cash-flow day of the month, a button appeared in her dashboard offering money. No application. No bank manager. No personal credit check, because the platform already knows her sales better than she does. She pressed it. Of course she pressed it. It was designed to be pressed. UX gotta be crazy not to push it.
This is a story about what happens after the button.
Note 6 in the Earnings Call - always a doozy
Shopify Capital originated $4.2 billion in loans and merchant cash advances in 2025, up $1.2 billion on the year before, and another $1.4 billion in the first quarter of 2026 alone, nearly double the prior year. The gross book now stands around $2.3 billion. The market shrugs at this, and it’s half right to: against Shopify’s valuation the book is a rounding error, the duration is short, the fee income is lovely. I thought it was a more of a story, turns out it is just a different one.
But buried in note 6 of the quarterly filings is a table almost nobody reads, and it might be the most honest dataset in ecommerce. The delinquency table. At the end of 2025, only 91.9% of Capital deals were current, down from 93.7% a year earlier. Sharper still: of the loans written in 2025 that still carry a balance, 10.8% were more than six months behind their original schedule by the end of March. One in nine.
Here’s why that number means something different than it would at a bank. Shopify Capital repayments are taken automatically, as a percentage of every day’s sales, at the till the lender also operates. You cannot miss a payment. You can only sell less than the model forecast. Which means that delinquency table isn’t really a credit report at all. That was a major learning along the way- take note Hendrikk. It’s a demand census: a quarterly, audited, machine-collected record of how many small merchants had a worse year than the algorithm predicted. One in nine of the class of 2025 is running six months behind their own expected life.
And the class of 2025 is precisely the vintage that borrowed into the squeeze. The tariff-at-the-border, landed-cost-inflation, buffer-stock vintage. I’ll be honest about the limits here: Shopify doesn’t disclose what merchants use the money for, so nobody can prove the split between ambition-borrowing and survival-borrowing. But when originations nearly double in the same eighteen months that import costs went vertical, you’re allowed to notice the shape of the thing. I was one of those merchants. It is not one thing, it is for everything. And they are laissez a faire about it’s reclaim too.
Weather system/Naming convention
It got one last week, in fact, from an unlikely poet: the CEO of FedEx. Raj Subramaniam told Fortune he’s watching combinations of trade flows he hasn’t seen in decades. US imports falling while exports rise. Intra-regional traffic climbing, Latin America inbound surging, Southeast Asia and India accelerating, all at once. He described it as the supply chain moving from one equilibrium to another, and coined a word for the in-between: reglobalization. MAGA.
Courier’s coinage belongs is a story about a candle shop’s overdraft. Transitions between equilibria are the most expensive place a small merchant can live. New suppliers mean new payment terms and no negotiating history. Re-routed lanes mean unfamiliar lead times, which mean fatter safety stock. Duties land at new borders on new schedules. Every single adjustment is cash out before revenue in. FedEx sees reglobalization as package flows on a dashboard in Memphis.
Note 6 is the same transition, seen as debt.Mandy is not happy.
And while the map redraws itself, the toll on every road went up. Fresh numbers this week from the TD Cowen/AFS Freight Index: ground parcel cost per package hit a record 42.4% above its 2018 baseline in Q2, with 2026 tracking as the most expensive shipping year ever recorded. The fuel surcharge alone rose 40% year on year. Hormuz oil shock, delivered to your door. Savour the mechanism: FedEx and UPS have re-tuned their fuel fee formulas so often that the elevated surcharge revenue persists even when diesel falls. At $4.50 diesel, ground surcharges now run around 25%, versus 22.5% a year ago. That’s not a pass-through. That’s a ratchet with a memory. Friction, monetised, by the same company whose CEO is narrating the storm. Sell the weather report and the umbrellas, why not.
Now stack the merchant’s year in order. Tariffs as cash at the border. Record cost per parcel, ratcheted. Inventory bought against that inflated base. All of it financed on 18-month, sales-linked credit, against a trade transition that Subramaniam himself frames as years long. A multi-year reordering of world trade, funded at the bottom of the stack with sub-two-year money. The enterprise next door finances the same transition with a five-year revolver and a treasury team. The rebel finances it with a button. One in nine of the class of 2025 running behind isn’t a mystery, and it isn’t necessarily bad merchants or even bad underwriting. It’s two clocks that don’t agree. Sorry for all the nasty banking language. I once watched the Wolf of Wall St and think I know it all. Oh, and I was in Corporate Finance for a bit. My loan book was interesting.
(Regular readers will spot the sibling essay here: while the carriers harvest the surcharge, Amazon Shipping is waiving surcharges and undercutting rates to lift exactly these small-merchant relationships. The rails play I wrote about yesterday, arriving right on cue.)
Stock on pallet 4D - location 496BAD
For the cash advance product, the structure is not a loan in the way your da understands the word. Shopify purchases a portion of the merchant’s future receivables at a discount. Read that again. The stock in the back room, the pallets she paid tariffs on, the buffer she built: in a meaningful legal sense, it has already been partly sold before a customer ever touches it. Not to a customer. To the platform.
So ask the question that matters: who controls that inventory? Formally, she does. Functionally, the clock does. The maximum term is 18 months, and if sales run slow enough that she’s off pace, she can find herself in default of a “flexible” product. So what does a founder do when the vintage runs behind? She discounts. She runs the sale she didn’t plan, in the season she didn’t choose, at the margin she can’t afford, because the fastest way to feed a percentage-of-sales repayment is to make the sales number bigger even as the profit number shrinks. The markdown calendar is now an amortisation schedule. Brand equity, that thing she spent seven years building: that’s what’s actually being remitted daily.
And when she’s repaid 65%? A fresh offer appears. Often larger. The treadmill has a loyalty programme. How’s your profitability planning guys.
One more structural detail. Not my forte but, these loans are originated through a bank partner and purchased by Shopify, and some are sold on, in full, to third-party investors, with Shopify booking a gain on sale. Somewhere, right now, the future sales of a candle shop are being traded between institutions that will never smell the candles.
Rebels a go go
Replatforming is creeping back into the conversation this year. But look at where. It’s an enterprise conversation. Enterprises have treasury teams, revolving credit facilities, banking relationships that live outside any commerce platform. Their working capital is portable, so their platform choice is a genuine choice.
The rebel merchant’s working capital now lives inside the platform. Underwritten by her Shopify sales data. Disbursed into her Shopify balance. Repaid through her Shopify checkout. Migrating mid-loan isn’t a migration; the repayment mechanism is the platform, so leaving is a default event. Nobody locked the door. She financed it, one convenient tap at a time. The cohort most celebrated for independence is the one accumulating the least practical freedom to move, and the friction (regular readers know I think friction is the real currency of this industry) has been perfectly monetised. Merle Travis wrote it in 1946 and Tennessee Ernie Ford made it famous: I owe my soul to the company store. The company store now has an API.
Tale wags dog. Dog not happy.
Atlantic Angle - pond life
If you’re reading this in Ireland, the UK or the EU and feeling smug: don’t. Capital rolled into Germany, the Netherlands, Ireland and Spain only within the last year and a bit. Which means the European book is a young vintage. Irish and German SMEs took this money in the last fourteen-odd months and haven’t yet carried it through a full demand cycle. The delinquency story you’ve just read is a North American story so far. The European edition is not cancelled; it’s scheduled.
And note what didn’t happen when it arrived. In Europe, becoming a bank takes a licence, supervision, years of a regulator breathing on your neck. Becoming the company store took a rollout schedule. A merchant cash advance is a purchase of receivables, not a loan, and business-to-business finance sits lightly regulated on both sides of the Atlantic, so this product crossed the ocean without a single authority treating its arrival as a banking event. It arrived as a dashboard button. Buttons don’t hold press conferences.
I want to be balanced but not OTT, interesting without being pass remarkable: the rebel merchant isn’t being ignored by this system. She’s being fought over. Carriers are tightening discounts for their big accounts while dangling incentives at small and medium shippers. Amazon is waiving surcharges to win her parcels. Shopify pre-approved her before she asked. Everyone wants the rebel’s flow: her orders, her data, her remittances. She just keeps the risk. Being the most courted customer in commerce and the least protected one turns out to be the same position.
And none of this makes Shopify a villain, which is exactly what makes it worth writing about. Shopify in this story isn’t a villain; it’s weather. The weather system finally has a name too, coined by a courier CEO watching the map redraw itself in real time. The product is genuinely convenient, the underwriting is genuinely clever, and every individual tap of that button was genuinely rational. That’s how you end up with one in nine of a vintage running six months behind, a stockroom that’s partly pre-sold to the till, and a generation of independent merchants whose independence is now a term sheet.
Reglobalization, invoiced from the bottom up.
So, merchants, one question before the next offer appears at 65%: is your markdown calendar yours, or your lender’s?
That’s your read. Mind your buttons.