NewtekOne's Q2 EPS slipped to $0.47, but the bigger story is the balance-sheet shift that could reprice NEWT

Q2 results were acceptable, but the market stayed focused on the business-model shift

NewtekOne reported diluted EPS of $0.47, within its guidance range of $0.42 to $0.52, and the stock closed the regular session at $14.95, down 1.19%, and remained unchanged in after-hours trading. That uneventful reaction says more than the EPS line alone. Investors did not react to one quarter of profitability; they are waiting for a clearer picture of how the transition away from loan sales will affect earnings quality and durability.

Management also announced it would withdraw prior guidance and provide updated forecasts within 45-60 days as it shifts toward retaining more SBA 7(a) loans on its balance sheet rather than selling them into the secondary market. That is the key near-term variable: the market needs proof that holding more loans can build a richer income stream, not just a larger balance sheet.

The strategic shift is toward banking-style income, not faster loan-sale turnover

This quarter matters because the income statement shows a modest step back while the balance sheet moves toward a more integrated business model. NewtekOne serves the $1 million to $100 million revenue SMB slice across lending, payments, insurance, payroll, and banking, so the better lens here is whether the company can deepen wallet share and retain more long-tail income over time.

Lower EPS was partly offset by better operating leverage

basic and diluted earnings per share ("EPS") were $0.48 and $0.47 vs. $0.53 and $0.52, respectively, for the three months ended June 30, 2025. But the operating picture was not weak. The efficiency ratio improved to 58.4% from 60.3%, and the business grew assets 50% while expenses only increased 3.6%. That is closer to scaling than breakdown.

The return metrics softened as the asset base expanded. Return on average assets ("ROAA") was 1.97% for 2Q26 compared to 2.58% for 2Q25, and return on average tangible common equity ("ROTCE") was 15.7% for 2Q26 compared to 19.6% for 2Q25. At this stage, that looks like the first-order effect of retaining more assets rather than proof the model is broken.

Deposit growth is the clearest support for the new model

The more important change is funding. NewtekOne has grown non-affiliate deposits from $142 million to $2.2 billion over 14 quarters, while core consumer deposits grew $297 million, or 21% Q/Q, and $951 million, or 124% Y/Y. Net interest income at the bank also rose to $25 million in Q2 2026 from $16.2 million in Q2 2025, supporting the case that the company is building a more stable income base as it holds more SBA 7(a) loans and C&I lending onto the balance sheet.

Execution is also improving. Earlier this summer, NewtekOne launched automated insurance integration within its proprietary loan operating system, reinforcing the idea that the company is trying to layer products and technology under the same lending relationship rather than relying mainly on upfront loan-sale economics.

Valuation gap matters more than the quarter-to-quarter EPS debate

This is now more of a positioning story than an EPS debate. The recent bridge still looks intact: Q1 came in at EPS of $0.43, Q2 improved to diluted EPS of $0.47 inside the guide, and book value per common share was $12.64. That does not make NEWT a buy, but it does make the valuation gap worth watching.

What the next update needs to confirm

Management said it would provide updated forecasts within 45-60 days as it retains more loans. The next update matters because it should clarify whether the company can convert deposit funding and loan retention into steadier bank-style earnings.

Look for three signals: - More recurring income: continued growth in net interest income, servicing, and cross-sold products. - Better funding economics: continued deposit growth and a lower-cost deposit base supporting loan retention. - A credible earnings path: updated guidance that reflects a richer, more durable mix rather than a simple slowdown in gain-on-sale revenue.

What could break the thesis

The risks are straightforward. If holding more loans does not translate into durable income growth, or if credit conditions worsen, the market may view the shift as a slower, heavier business model rather than an upgrade. As insured deposits comprised 81% of deposits as of June 30, 2026 and the deposit base has grown quickly, execution risk is less about demand alone and more about maintaining discipline, funding stability, and asset quality as the model evolves.