Leading AI lending platform Upstart Holdings (UPST) — a company striving to build partnerships with top-tier banks — reported better-than-expected Q3 earnings on November 4, posting EPS of $0.23 versus the $0.08 expected by analysts. However, the stock tumbled following the earnings release as investors shifted their attention to the company’s weak Q4 guidance.
Meet Your ETF AI Analyst
- Discover how TipRanks' ETF AI Analyst can help you make smarter investment decisions
- Explore ETFs TipRanks' users love and see what insights the ETF AI Analyst reveals about the ones you follow.
Adding to investor concern, CEO Dave Girouard acknowledged that Upstart’s risk assessment model had overreacted to macroeconomic signals during the quarter, resulting in a decline in loan approvals. While this overreaction warrants careful monitoring, I believe Upstart remains well-positioned to lead the U.S. AI-driven lending market over the long term. In my view, this pullback represents a buying opportunity for long-term investors, and I remain Bullish on Upstart’s future prospects.
The Model Overreaction Needs Immediate Fixes
One of the key reasons behind my bullish outlook on Upstart is my belief that the company’s AI-driven risk assessment model will ultimately outperform traditional FICO-based credit evaluation methods. For my investment thesis to hold, however, Upstart’s model must consistently deliver the accuracy and reliability it promises.
As management revealed during the Q3 earnings call, Upstart’s AI model began tightening its credit box in late Q2, continuing through August before the company moved to normalize the model in pursuit of improved outcomes. This tightening had a significant impact on Q3 performance.
The loan conversion rate dropped from 23.9% in Q2 to 20.6% in Q3, despite loan application demand hitting a three-year high, which limited the company’s ability to capitalize on strong demand. Loan originations came in at $2.9 billion, down slightly from $3 billion in Q2 and below expectations of $3.3 billion. Consequently, Q3 revenue missed analyst estimates by roughly $2.6 million, even though it still reflected a robust 71% year-over-year increase.
Addressing the model’s overreaction, CEO Dave Girouard noted that overreacting to macroeconomic signals is preferable to underreacting, as the latter could lead to lending to uncreditworthy borrowers — a mistake that could prove costly in a downturn, as seen during the sharp rate hikes of 2022. While this rationale makes sense, banks and lending partners naturally prefer AI models that deliver precise and stable credit assessments.
To address the issue, Upstart has been normalizing its model since September to correct the overreaction. Moreover, with a long-term focus, the company is rolling out enhancements designed to reduce month-to-month volatility in model calibration by half, which should result in greater stability, predictability, and overall performance of its AI risk assessment system.
Lackluster Q4 Guidance Stems From Several Negative Developments
While I remain bullish on Upstart’s long-term potential, the company’s disappointing Q4 guidance stands out as a notable short-term headwind that could keep the stock under pressure in the near future.
Upstart guided for Q4 revenue of $288 million, implying quarter-over-quarter growth of only about 4%, compared to analyst expectations of $304–$307 million. The company’s adjusted EBITDA forecast of $63 million also came in below analyst estimates of $67 million.
In my view, a key driver behind this soft guidance is a slower-than-anticipated recovery in loan conversion rates. Although Upstart has been working to normalize its AI risk assessment model, the Q4 outlook suggests that these adjustments are taking longer than expected to translate into improved performance. Additionally, Upstart appears poised to scale back its use of balance sheet funding in Q4 to remain resilient in the face of potential economic uncertainty.
For reference, the company deployed $449 million from its balance sheet in Q3 to meet strong consumer credit demand — a level that is not sustainable given Upstart’s business model as an intermediary rather than a direct lender. Acting like a bank would expose the company to heightened credit default risk, which contradicts its core strategy. As Upstart reduces its reliance on internal funding, loan originations will likely decline in Q4, a factor management appears to have incorporated into its cautious guidance for the quarter.
Upstart’s Long-Term Outlook Remains Intact
Although Upstart is facing some challenges, I believe its long-term growth opportunities remain intact. In Q3, Upstart’s new markets, including small-dollar loans, auto financing, and home loans, accounted for 12% of total loan originations, indicating that the company’s diversification efforts are yielding positive results. Given that many of these new markets have a significantly larger addressable market opportunity than the personal loan market, these diversification efforts are likely to trigger an acceleration in revenue growth over the next few years.
From a technological perspective, Upstart has already demonstrated that its risk assessment model, trained on more than 90 million repayment events, delivers better outcomes compared to the FICO model used by banks.
With every passing quarter, Upstart’s AI models are being trained on an increasing number of credit events and repayment events, which should ultimately improve the overall accuracy and functionality of its model. The company seems well-positioned to dominate the AI lending market for many years to come, which should help it attract more borrowers and lenders to its platform. This flywheel effect is expected to boost revenue and earnings growth over the next few years.
Is UPST a Good Stock to Buy Now?
On Wall Street, UPST stock carries a Moderate Buy consensus rating based on six Buy, five Hold, and two Sell ratings over the past three months. UPST’s average stock price target of $56.45 implies more than 45% upside potential over the next twelve months..
I agree with the consensus analyst view, as Upstart has successfully executed its turnaround strategy to return to profitability, even while interest rates remain elevated. At a forward P/E multiple of 23x, Upstart is valued like a company nearing maturity, when in reality, an acceleration in growth can be expected in 2026 after recovering from the fallout of the AI model overreaction.
Upstart’s Post-Earnings Selloff Looks Overdone
Upstart’s stock plunged after its Q3 earnings release, driven largely by the company’s disappointing Q4 guidance and the AI model’s overreaction to macroeconomic signals during the quarter. While the performance of the AI risk assessment model clearly requires prompt correction—a process the company has already initiated—I believe this setback does little to alter Upstart’s long-term growth story. The company remains well-positioned to capitalize on its diversification initiatives and to solidify its standing as a technological leader in the U.S. AI-driven lending space.