Pagaya Posted a Record $45 Million Profit. Is AI Lending Finally Working?

On July 30, Pagaya Technologies reported a record $45 million of GAAP net income attributable to shareholders in the second quarter, giving investors a concrete profitability milestone after years of emphasis on network expansion. Pagaya Technologies Ltd. (NASDAQ:PGY) uses machine-learning models to help a network of lending partners and institutional investors originate and finance consumer credit. The result suggests the platform may be moving toward repeatable profitability.

Pagaya Posted a Record $45 Million Profit. Is AI Lending Finally Working?

Photo by Clay Banks on Unsplash

The latest results make that case credible. Second-quarter network volume rose 33% to $3.5 billion. Total revenue and other income increased 19% to $387 million, while adjusted EBITDA grew 43% to $124 million. Pagaya raised its full-year net-income guidance, and its annualized Auto network volume reached a record $4.8 billion. The institutional-investor network expanded to 174, providing a broader funding base for loans generated through partner channels.

Machine learning leaves the basic credit cycle intact. Pagaya depends on capital markets remaining open and on its models correctly pricing borrowers through changing economic conditions. Securitization structures can be difficult for outside investors to evaluate, and fast network-volume growth may look less attractive if delinquencies rise or funding spreads widen. The stock also carries volatility from a relatively concentrated shareholder base and shifting risk appetite.

Hedge funds became more numerous in Q2. Insider Monkey tracked 38 funds holding Pagaya Technologies Ltd., up from 33 in Q1. Millennium Management reported 2,286,864 shares as of June 30. Citadel disclosed 585,956 common shares as well as call and put positions, a mixed structure that resists a simple directional interpretation.

Short interest remained substantial. The August 14 settlement showed 11.38 million shares short, down 8.3% from 12.41 million on July 31, with 3.8 days to cover. Because vendors publish materially different float denominators, the dated share count is more reliable than a percentage. The position can amplify another earnings beat and also leaves the shares exposed if funding or credit conditions worsen.

Pagaya’s record profit is the best evidence yet that its AI-lending model can produce more than transaction volume today. The next test is resilience: investors need to see stable credit performance, recurring investor demand, and cash generation across a less forgiving financing environment. Securitization spreads, loan delinquencies, and fee revenue per dollar of network volume will matter more than raw originations. Those indicators can show whether model performance is creating durable economics or simply enabling faster balance-sheet turnover without adequate risk-adjusted returns for shareholders.

READ NEXT: NVIDIA (NVDA): What Foxconn and Super Micro Are Telling Us about the AI Boom  and Pony AI Is Scaling Robotaxis Fast—Can the Stock Reach BofA’s $17 Target?

Follow Insider Monkey on Google News.