ORRF — Orrstown Financial Services, Inc. · 2026-09-10 · Verdict: WATCH · Conviction 3
Price $42.05 (screen row, universe_v2.csv, screen build 2026-09-10; no live price) · Mkt cap $827.1M (19,669,626 shares at 30 June 2026) · EV and EV/EBIT n/m for a deposit-funded bank · P/E 10.2x FY2025 GAAP, 9.5x annualised H1 2026 · P/TBV 1.57x · ADV $3.5M Sources read: 10-K 2026-03-12 (Items 1, 1A, 7), 10-Q 2026-08-06, DEF 14A 2026-03-27, 8-Ks 2026-04-22, 2026-05-05, 2026-05-06, 2026-05-29, 2026-07-22, Form 4s (12m), Q2 2026 earnings release EX-99.1 (no transcript or prepared remarks was filed).
Desk stats - Revenue trend. Net interest income $199.8M in 2025 against $155.3M, up 28.7%, fees $52.3M against $37.4M (10-K Item 7), almost all from owning Codorus Valley twelve months instead of six after the 1 July 2024 merger. Q2 2026 reverses it: $48.8M against $49.5M, minus 1.4% (8-K 2026-07-22). Core reverses again: excluding loan accretion and adding back the accelerated note mark, H1 core spread income was $90.9M against $86.8M, up 4.8%, on loans up 4.6% to $4.113B (10-Q). - Normalised after-tax profit. H1 2026 pre-tax $52.165M. Add back $1.6M accelerated amortisation of the fair value mark on notes redeemed 30 June 2026. Subtract $2.4M of Q1 life insurance death benefits, one-off and tax-free (life insurance income $5.137M in H1 2026 against $2.600M). Subtract $1.7M accelerated loan accretion (10-Q). That is $49.7M for the half, $99.3M annualised; at the desk's 25% rather than the structural 20.7% (8-K 2026-07-22, which also names a one-off $1.6M benefit from purchased federal tax credits), normalised after-tax profit is about $74.5M, or $78.5M at the actual rate, against FY2025 GAAP net income of $80.9M. - EV / normalised after-tax profit. Not meaningful for a bank whose liabilities are its raw material. The equivalent is price / normalised after-tax profit of 11.1x at 25% tax, 10.5x actual, 12.7x fully burned down for accretion per section 3. - Financial-lane equivalents (no EV/EBIT). P/TBV 1.57x on tangible book of $26.71 a share, from $25.21 (10-Q non-GAAP). ROTE 13.96% in Q2 and 14.76% in Q1 as reported, 14.6% normalised on $508.6M average tangible common equity. Tangible book compounded 11.9% annualised over the half. CET1 12.0%, Tier 1 leverage 10.1%, tangible common equity 9.5% of tangible assets (8-K 2026-07-22). - Is the growth sustainable? The 2025 growth was bought and is finished. What continues is organic and cash-backed: 4.6% loan growth, 19.9% noninterest-bearing deposits, wealth fees up 7.7% to $11.5M on $3.1B under management (10-K Item 1). No numeric guidance was given, only that the redemption "is expected to enhance future performance" (8-K 2026-07-22, CEO). - What the screen got wrong. Net income growth of 266.7% is merger arithmetic: 2024's $22.1M carried $22.7M merger expenses, a $15.5M provision on acquired non-PCD loans, $4.8M for an executive retirement and $0.5M legal settlement, and the company's own adjusted 2024 figure is $56.1M, making real growth 47.8% on a half year of Codorus (10-K Item 7). The revenue field of $37.1M for 2025 is a bad XBRL tag against roughly $252M actual, which is why the row reads "net income exceeds revenue"; a tagging artifact, not an earnings-quality flag. Dividend yield of 2.5% is stale against the declared $0.30 quarterly, 2.85% here. Above all the screen cannot see $21.5M of purchase accounting accretion inside 2025 interest income (10-K Item 7, note 6 to the net interest income table), which flatters both the P/E numerator and the ROTE it ranks on.
1. What the business actually does
A $5.6B commercial bank in Harrisburg lending across south central Pennsylvania and the Baltimore corridor (10-K Item 1). Loans $4.113B, 75% commercial, funded by $4.620B of deposits, $920.9M noninterest-bearing. Fees are large for the size, $52.3M in 2025, led by an advisory and trust arm with $3.1B under management. It doubled itself in July 2024 by merging with Codorus Valley Bancorp for $233.4M, buying the York market. Nothing since, so 2026 is the first year showing what the combined bank earns.
2. Why it is mispriced, the edge case
Only half an edge exists, which caps the verdict at WATCH. The half that does: headline net interest income is falling, minus 1.4%, which reads like a merger story fading. It is not. Loan accretion fell from $11.5M in H1 2025 to $8.5M and Q2 also absorbed the $1.6M note mark, yet net interest income fell just $0.4M for the half. Core spread income grew about 5%, and the redemption removed $31.0M of debt costing 7.72% (8-K 2026-05-29). A screener reading headline revenue sees decay; the filings show decay being absorbed.
The missing half is any seller. The stock is 3.3% off its 52-week high and up 21.4% in six months (universe_v2.csv). Insiders are not buying: zero open-market purchases in twelve months against 5,532 shares sold for $234K, mostly an option exercise by the chief administrative officer and a director trim, both retaining far larger stakes (form4_last12m.csv).
3. Credit quality, margin, deposit mix
Credit is clean and improving, the part the market has right. Nonaccrual loans fell to $23.8M, 0.58% of loans, from 0.74%; classified loans to $49.2M, 1.2%, from $58.4M and 1.5% at year end; nonperforming assets 0.44% of assets; charge-offs annualise near 0.10% (8-K 2026-07-22; 10-Q). The allowance is 1.13% of loans and was released partly on improving qualitative factors, including removing the concentration overlay on non-owner-occupied commercial real estate (10-Q). Worth watching, because the concentration is not small: commercial real estate was 296% of total risk-based capital at year end, just inside the 300% guidance threshold and up 87% over three years, office 59% of capital, construction and land 41% (10-K Item 1A).
Margin decides the case. Taxable-equivalent net interest margin was 4.04% in 2025 but included $21.5M of loan accretion plus $3.2M on acquired securities, about $24.7M pre-tax or 12% of net interest income (10-K Item 7), so core margin was near 3.55%. Accretion decays slowly and predictably, $4.9M, $5.3M, $4.7M, $4.2M, $4.3M across the five quarters to Q2 2026 (8-K 2026-07-22), roughly $4M to $5M of pre-tax income leaving each year. I cannot size the remaining accretable balance from this bundle because the loan note is not included, an honest gap. Burning off all $6.8M of H1 scheduled accretion takes normalised annualised pre-tax from $99.3M to $85.7M, and an estimated $1.2M net saving from the redemption gives $86.9M, or $65.2M after 25% tax: 12.7x price, 12.8% ROTE. Against that, noninterest-bearing balances grew $49.9M in the half while time deposits shrank, and core margin held at 4.00% (8-K 2026-07-22).
4. Balance sheet and capital allocation
CET1 12.0%, Tier 1 leverage 10.1%, tangible common equity 9.5% of tangible assets from 9.2% in a quarter. The subordinated notes are gone, leaving about $10.3M of trust preferred debt assumed in the merger. Tangible book rose $1.50 in the half, from $43.0M of net income less $11.8M dividends and $2.6M other comprehensive loss (10-Q). The dividend is $1.20 annualised, a 27% payout. The buyback is authorised for 500,000 shares and essentially unused, 8,330 shares in all of 2025 (10-K Item 7); at 1.57x tangible book that restraint is correct. The live risk is the opposite: a bank with 12% CET1, a finished integration and a new lending-background chief executive is a bank that goes shopping.
5. Management: said versus did
Thomas Quinn retired 1 June 2026 after running the merger, replaced by Adam Metz, 54, promoted from chief operating officer, and before that chief revenue and chief lending officer since joining in 2016 (8-K 2026-05-06). He was named successor in February 2025 with a 54.5% salary rise, so this was planned (DEF 14A). They did make the combined bank more profitable: 1.53% return on average assets and 13.96% on average equity in Q2 are top-quartile for a $5.6B community bank, with credit improving four quarters running. The blemish is that 2025 cash incentives paid at 150% to 162% of target in a year whose reported growth was mostly prior-year charges falling away (DEF 14A). Say-on-pay passed with 94.91%, and supplemental retirement accounts credit interest at the company's own return on average tangible equity, capped at 15%, which is real alignment.
6. Valuation
Bear, 25%: accretion outruns core growth, margin settles at 3.55%, charge-offs normalise from 0.10% to 0.35%, multiple compresses to 1.2x book. Tangible book still compounds 6% to about $29.5 by end-2027, giving $35 to $36, minus 14%. Base, 50%: core spread income grows 4% to 5% and offsets most of the annual accretion decay, normalised earnings hold near $4.00 a share, ROTE settles near 13.5%, tangible book compounds 9.7% on a 27% payout to $30.70. At an unchanged 1.55x, $47.5 plus $1.80 of dividends, plus 17% over eighteen months. Bull, 25%: deposit remix and the redemption hold core margin at 4.00%, wealth keeps compounding, and a scarce $5.6B Pennsylvania and Maryland franchise re-rates or is bought at 1.9x book, about $58, plus 38%. Probability weighted, about plus 14% over eighteen months, near 9.5% annualised. Reverse DCF in bank form: at 1.574x tangible book, a 10% cost of equity and 4% terminal growth, the current price implies a sustainable return on tangible equity of about 13.4%. Normalised with accretion is 14.6%, fully burned down 12.8%. The price sits almost exactly between them, which is the whole answer.
7. Catalysts and timeline
Q3 2026 results in late October are the first clean quarter with no notes and no merger costs, and should show reported margin stepping toward the 4.00% core level. Full-year results in March 2027 bring the loan note that would size the remaining accretable discount. A whole-bank acquisition is plausible on this capital and more likely to hurt than help at 1.57x book.
8. Pre-registered kill criteria
- Taxable-equivalent net interest margin excluding accelerated fair value marks below 3.70% for two consecutive quarters, from 4.00% in Q2 2026.
- Classified loans back above $60M, or nonaccrual above 0.85% of loans, for two consecutive quarters, from $49.2M and 0.58%.
- Tangible book per share growth below 7% annualised for two consecutive quarters, from 11.9% in H1 2026.
- A whole-bank acquisition taking tangible common equity below 8.5% of tangible assets, from 9.5%, returning the company to growth by dilution. Other risks: commercial real estate at 296% of risk-based capital with office at 59% leaves little headroom before heightened scrutiny; the allowance is being released on qualitative judgement while special mention loans crept up to $111.3M; a third-party vendor breach exposed customer data on 21 May 2026, with the company's own systems unaffected (8-K 2026-05-29).
9. Verdict and summary
WATCH, conviction 3. Orrstown is a good bank at a fair price, and both cases you would build from the headline numbers are wrong. The 266.7% earnings growth is merger arithmetic against a 2024 carrying $43.4M of charges, so this is no profit tripling. But the falling revenue line misleads equally: net interest income is down 1.4% only because purchase accounting accretion is decaying from $21.5M a year, while core spread income grew about 5% on 4.6% loan growth with credit improving four straight quarters. Normalise the one-offs and you get roughly $74.5M after tax, 11.1x the market cap, falling to about $65M and 12.7x once accretion fully burns off. At 1.57x tangible book the market implies a sustainable 13.4% return on tangible equity and the honest range from the filings is 12.8% to 14.6%, so the run-off is already in the price, three percent off the high after a 21% six-month run, with no insider buying and no forced seller in sight. Keep the file open because tangible book compounds near 12% on a 27% payout and 12% CET1, so a patient holder earns the ROTE without a re-rating. Do not buy today because you are being paid the return, not the discount, and the new chief executive has both the capital and the background to spend it on another deal.
Research for discussion, not investment advice. Positions and sizing are the reader's decision.