Horizon Bancorp (HBNC) Q4 2025 earnings review

The Pain Trade Pays Off: Margins Explode Post-Restructuring

Horizon Bancorp's massive balance sheet restructuring in Q3—which resulted in a $222M loss—has immediately borne fruit in Q4. Net Interest Margin (NIM) skyrocketed to 4.29% from 3.52% the prior quarter, propelling the bank to a net income of $26.9M ($0.53 EPS). By dumping underwater securities and expensive funding, HBNC has successfully engineered a 'V-shaped' recovery in profitability, achieving a peer-leading ROA of 1.63%. While the earnings engine is fixed, the bank is now shrinking its deposit base (-4.4% QoQ) to optimize costs, a strategy that requires careful liquidity monitoring.

🐂 Bull Case

Restructuring Success

The thesis has played out perfectly: selling low-yield assets and paying down debt caused NIM to surge 77bps QoQ to 4.29%. With FY26 NIM guidance at 4.25-4.35%, this profitability level appears durable.

Commercial Loan Engine

Despite balance sheet noise, core commercial loans grew $76M (9.1% annualized) in Q4. The bank is successfully swapping low-yield auto loans for high-yield commercial credits.

🐻 Bear Case

Credit Normalization

Non-performing assets (NPAs) rose to $40.6M (0.63% of assets) from $35.7M in Q3 and $27.4M a year ago. While not critical, the trend is consistently upward.

Deposit Outflows

Total deposits fell $245M (-4.4%) sequentially. While management calls this 'intentional exit of high-cost funds,' shrinking the liability base limits future asset generation capacity.

⚖️ Verdict: 🟢

Bullish. The bank executed a painful rip-the-bandage-off restructuring in Q3 and delivered immediate, massive margin expansion in Q4. Earnings quality is high (ROA >1.6%), and FY26 guidance suggests the new profitability baseline is real.

Key Themes

DRIVERNEW🟢🟢

Margin Super-Cycle

Accelerating. Horizon's NIM expansion is the standout metric, jumping from 3.52% to 4.29% in a single quarter. This was driven by the full-quarter benefit of the Q3 balance sheet repositioning (selling underwater securities/loans) and paying down expensive FHLB advances. Management guides for stability at this elevated level (4.25-4.35%) in FY26, suggesting the bank has structurally shifted its earnings power.

DRIVER

Commercial vs. Consumer Mix Shift

Stable. The bank is aggressively remixing its loan book. Commercial loans grew 9.1% annualized ($75.8M) in Q4, focused on C&I and Real Estate. Conversely, Consumer loans fell 2% QoQ and 30% YoY as the bank lets the indirect auto portfolio runoff. This mix shift is accretive to yield but concentrates risk in commercial real estate/business sectors.

CONCERNNEW

Creeping Asset Quality Issues

Deteriorating. While management touts 'excellent' credit quality, the numbers show a slow negative drift. Non-performing assets (NPAs) rose to $40.6M (0.63% of assets) in Q4, up from $35.7M in Q3 and significantly up from $27.4M in 24Q4. Net charge-offs rose to 0.08% annualized. This is not a crisis, but the trend is undeniably moving the wrong direction.

THEME🔴

Deposit Base Optimization (Shrinkage)

Decelerating. Total deposits dropped $245.5M (4.4%) in Q4 alone. The bank attributes this to 'planned exit of high-cost, transactional deposits.' While this helps NIM (cost of interest-bearing liabilities fell 34 bps), shrinking the balance sheet reduces the dry powder available for future lending without releveraging via borrowings.

DRIVER

Operating Efficiency Reset

Improving. Non-interest expense fell to $40.6M in Q4 from an inflated $53.0M in Q3 (which included debt prepayment penalties). The Q4 efficiency ratio improved to 54.2%, a massive recovery from the restructuring-impacted metrics of 2025. FY26 guidance puts expenses in the mid-$160M range, implying quarterly stability around $40-$41M.

Other KPIs

Net Interest Income (25Q4)$63.5 million

Accelerating. Up 8.7% QoQ and 19.5% YoY. This is the primary revenue engine, fueled by the 77bp expansion in NIM.

Tangible Book Value Per Share (25Q4)$10.32

Recovering. TBV recovered $0.56 QoQ to $10.32, though it remains significantly down from $13.68 a year ago due to the heavy realized losses taken in Q3 to restructure the bond portfolio.

Return on Average Assets (25Q4)1.63%

Reversing (Positive). A sharp turnaround from the -12.07% in Q3 (due to one-time losses) and significantly higher than the 0.92% normalized run-rate seen in 24Q3. This level of profitability is elite for a community bank.

Guidance

FY26 Net Interest Margin (FTE)4.25% - 4.35%

Stable/High. Guidance implies maintaining the elevated Q4 level (4.29%). This suggests the bank does not expect significant compression from potential rate cuts, likely due to the removal of fixed-rate underwater assets.

FY26 Net Interest Income (Non-FTE)Low-teens growth

Accelerating. Implies substantial growth over FY25's annualized run rate, driven by the full-year effect of the wider margin.

FY26 Loans (HFI)Mid-single-digit growth

Stable. Consistent with the growth seen in 25Q4, driven by Commercial lending offsetting the continued runoff of the auto portfolio.

FY26 Non-Interest ExpenseMid-$160 million range

Stable. Suggests quarterly expenses of ~$40-41M, consistent with the clean Q4 number ($40.6M), implying no major cost creep.

Key Questions

Deposit Floor vs. Liquidity

Total deposits shrank by nearly $250M this quarter as you exited high-cost funds. At what absolute dollar level do you need to stabilize deposits to support the mid-single-digit loan growth guided for FY26 without tapping wholesale borrowings?

NPA Drift

Non-performing assets have drifted up for three consecutive quarters (from $27M to $40.6M). Is this idiosyncratic to a few credits, or are you seeing broader stress in specific commercial segments like office or trucking?

Capital Deployment Priorities

With ROA at 1.63% and the restructuring complete, capital generation will be significant in 2026. Does the current valuation make buybacks the primary use of cash, or are you looking at M&A opportunities in Michigan/Indiana?

Consumer Portfolio Floor

The consumer portfolio dropped another 2% this quarter. Do you have a target floor for this portfolio size, or should we model a continued runoff toward zero?