Upstart Q1 2026 Preview: All Systems Gu?
- Henry Dierkes
- 20 minutes ago
- 5 min read
Upstart heads into Q1 2026 earnings looking a lot stronger than the stock price suggests.
That’s the short version.
The longer version is a little messier, because this is still Upstart. Volatile stock. Complicated business. Private credit headlines. Margin pressure. Macro worries. A management team that has not always made life easy for investors.
But strip away the noise and the setup is pretty clear: Upstart is trying to prove that AI can do in credit what Amazon and Costco did in retail. Use a structural advantage, pass some of the economics back to customers and partners, then let the flywheel build.
That’s the bet.
What is Upstart really trying to build?
Upstart is trying to become the AI underwriting layer for credit.
Not just personal loans. Not just non-prime borrowers. The company now has a much broader product set across personal loans, prime, auto, HELOC, small dollar loans, and the newly launched Cash Line revolving credit product.
That matters because the old knock on Upstart was simple: “Cool model, but it’s too tied to non-prime personal loans and too dependent on loose funding markets.”
Fair criticism.
But the business in 2026 doesn’t look like the business from 2021 or 2022. Back then, Upstart had no committed capital. Today, well over half of platform capital is committed. That changes the durability of the model in a meaningful way.
The bull case is that Upstart’s AI can separate risk better than traditional FICO-based underwriting. If that works, the company can approve more good borrowers, offer better rates, improve credit investor returns, and scale originations without blowing up credit performance.
Simple idea. Extremely hard to pull off.
What did Q4 prove?
Q4 2025 was a strong quarter.
Revenue came in at $296 million, up 35% year over year. GAAP net income was $18.6 million. Adjusted EBITDA was $63.7 million, good for a 21.5% margin.
The bigger point, though, was the balance sheet.
Upstart reduced loans held on its balance sheet by $245 million quarter over quarter, including a 25.7% reduction in R&D loans. That’s important because Upstart’s long-term thesis depends on being a marketplace, not a direct lender stuffing loans onto its own books.
Also notable: 70% of Auto and HELOC funding in Q4 came from banks, credit unions, and private capital partners. Eleven different capital partners funded those loans, with another 13 signed for the coming year.
That’s not nothing.
It suggests the newer products are starting to find external funding support, which is exactly what investors needed to see.
What changed after Q4?
A lot, actually.
Since Q4 earnings, Upstart launched Cash Line, repurchased $100 million of stock at an average price of $31.31, announced a $333 million auto asset sale to Bayview, signed a $200 million auto forward-flow agreement with Wafra, announced a $1 billion forward-flow deal with Eltura Ventures and Aperture Investors, upsized a Centerbridge agreement to $1.2 billion, and extended/upsized a Fortress agreement to $1.25 billion.
That’s a mouthful.
The takeaway: capital partners appear to still be showing up.
That matters even more because private credit has been getting dragged through the headlines lately. If private credit funding cracked, Upstart would feel it. But so far, the actual company-specific evidence does not show that stress.
Could that change? Absolutely.
But as of this preview, the funding announcements point in the other direction.
Why does the macro matter so much?
For Upstart, the macro isn’t just background noise. It feeds directly into loan pricing.
That’s where the Upstart Macro Index, or UMI, comes in. UMI measures how macro conditions are affecting expected defaults for Upstart-powered unsecured personal loans.
In Q1 2026, UMI averaged 1.39, down slightly from 1.40 in Q4 2025. March came in at 1.42, which sits inside management’s longer-term guidance assumption of 1.40 to 1.50.
So the macro was not a surprise headwind this quarter.
That’s a big deal because Q3 2025 was partially disrupted by model tightening tied to elevated macro risk signals. This time around, the data looks more stable.
Not perfect. Stable.
For a company like Upstart, stable is good.
What does the origination data say?
Upstart’s new monthly origination disclosure is a major improvement.
Investors no longer have to wait for earnings day to guess what happened to originations. By the time the quarter is reported, they already know the monthly origination picture.
For Q1 2026, total originations were $3.443 billion, up 7.8% quarter over quarter and 61.3% year over year.
April also started Q2 on solid footing, with $1.274 billion in originations, up 51% year over year. Originations per day reached $48.2 million, up 5% month over month and 51% year over year.
That does not mean revenue automatically follows one-for-one. Take rate matters. Margins matter. Product mix matters.
But the origination trend is constructive.
What is Henry watching for in Q1 earnings?
Three things matter most.
First, revenue. Henry’s UpstartIQ models land in a tight range, with one estimate around $301 million and the other around $303 million. That would imply roughly 41% to 43% year-over-year revenue growth.
Second, GAAP profitability. If Upstart can keep showing profitability while growing originations, that strengthens the “growth, profits, and credit performance” thesis.
Third, the balance sheet. This is probably the biggest one. Investors need to see continued progress reducing exposure, especially R&D loans tied to newer products.
There’s also the Paul Gu factor.
Paul is now CEO. Dave Girouard handed him the keys. The company has changed its guidance framework, added monthly origination reporting, and appears to be taking investor communication more seriously. That matters because Upstart’s communication has been, let’s say, a little adventure in the past.
Maybe the Gu era brings a cleaner message.
Investors would welcome that.
What are the risks?
The biggest risk is margin compression.
Upstart is intentionally passing more of its AI-driven underwriting advantage to borrowers and credit investors. Strategically, that makes sense. It can strengthen the flywheel.
But if the company gives up too much margin too quickly, investors may punish the stock.
The second risk is funding. Private credit headlines are not irrelevant. Upstart depends on outside capital partners. If that funding environment tightens, the business could feel pressure fast.
The third risk is macro. UMI looks fine right now, but credit can change quickly. If macro risk rises again, Upstart’s models may tighten, borrower terms may get worse, conversion may drop, and revenue may come under pressure.
That’s the game.
The Rebellionaire TL;DR
Upstart enters Q1 2026 earnings as a stronger, more diversified, better-funded company than it was during the messy 2022–2023 period.
The market still doesn’t fully trust it. Fair. Upstart earned some of that skepticism.
But the company has made real progress: more committed capital, broader products, stronger capital partner activity, better transparency, a new CEO, and monthly origination disclosures that reduce some of the earnings-day guesswork.
Henry’s base expectation is roughly $303 million in revenue, GAAP profitability, and continued balance sheet reduction.
The whole story comes down to one question:
Can Upstart use AI-driven risk separation to grow originations, stay profitable, and keep credit performance strong?
If yes, this gets interesting fast.
If no, the market’s skepticism was right.
That’s why Q1 matters.
Check out Henry's full report here:
Resources
[1] Henry Dierkes, “Upstart Q1 2026 Earnings Preview,” Rebellionaire, May 4, 2026.