HomeTrust Bancshares, Inc. HTB

Key data

The business

HomeTrust Bank sits in Asheville, was chartered in 1926, and runs about $4.4B of assets through 41 branches across North Carolina, upstate South Carolina, east Tennessee, and southwest Virginia. It converted out of a thrift years ago and now lends commercial: commercial real estate, construction, and business loans, alongside a retail book of home equity lines, indirect auto paper booked through car dealers, SBA loans, and equipment leases. It is deposit-funded, and the moat is the ordinary one for a bank this size, depositor inertia in small markets a national bank does not bother to fight for. Funding is a tailwind again: interest paid out fell from $19.4M to $15.9M over the last four quarters. The live tension is the August 17 agreement to acquire Blue Ridge Bankshares, a Virginia bank working through regulatory remediation and a shrunken fintech-banking book. That deal pushes HomeTrust north into Virginia and is the swing factor in everything below. The concrete thing on the balance sheet is a used-car loan written through a dealership, or an SBA loan to a mountain-town contractor.

The numbers

The five quarters show a bank holding its top line and slowly giving back the bottom.

Net interest income has parked at $44 to $45M for five straight quarters while net income slid from $17.2M to $15.6M, down about 9% from a year ago. Diluted EPS held near a dollar only because the share count kept shrinking. The give-back is on the cost side, where operating expense ran about 10% higher than a year ago.

The compounding is real but modest. HomeTrust moved its fiscal close from June to December in 2023, so the older rows end in June; on a like basis diluted earnings grew from $2.23 to $3.81 over about four years, near 14% a year, and the bank earns roughly 11.5% on tangible common equity and 10.9% on all equity. The dividend takes only 14% of earnings, so most of that return retains into book value, which now stands at $34.19 a share of hard equity. That is the machine: a low-double-digit return quietly building the book. What the market may be under-weighting sits in the next table.

Read the first column down. Borrowers ninety days behind or worse climbed from about a quarter of a percent of loans in mid-2023 to 1.29% by March 2026, more than a fivefold rise, while the cushion set aside against loans fell from 1.27% to 1.11%. Charge-offs, the loans actually written off, stayed low near 0.2%, and loans 30 to 89 days late jumped to 0.69% of assets in the March quarter from 0.17%. Either management is right that these bad loans are well covered by the real estate behind them, or the reserve is lagging the problem. That gap is the whole question.

Management

Insiders have been sellers, not buyers: seven open-market sales worth $3.4M and zero purchases over the past twelve months, with CEO Hunter Westbrook selling about $2.3M across two May transactions. Those carried the Form 4 exercise-and-sell codes rather than reading as discretionary conviction cuts, which softens the signal, though a founder-level bank rarely sells into a story it fully believes. Capital allocation is steadier: the $14.9M of stock bought back in 2025 came at an average $37.99, above today's tangible book of $34.19 and so mildly book-dilutive, while the big $43.7M repurchase back in 2022 landed at $26.47, well below book and well-timed. The light payout leaves room to keep compounding equity. Pay detail is not in this filing set.

How it fails or surprises you

Credit keeps migrating. Noncurrent loans have risen almost every quarter for two years to 1.29% while reserves fell to 1.11%. If the next report shows noncurrent through 1.5% and charge-offs breaking above 0.35%, the low-loss story ends and provisions eat the earnings that hold the multiple. The print: noncurrent-to-loans and net charge-offs in the Q3 2026 10-Q.

The reserve is the fact the read explains least. A prudent bank builds reserves as bad loans climb; this one let reserves fall from 1.27% to 1.11% into a fivefold rise in noncurrent loans. That proves the bull wrong if collateral marks turn soft. The tell is another quarter of reserves declining against rising noncurrent.

Blue Ridge bought cheap and cleaned (right tail). Blue Ridge traded below its own tangible book on its troubles; if HomeTrust's mark proves conservative and the Virginia book behaves, it buys a franchise below book and re-rates. The right-tail print is the S-4's pro forma tangible book per share and day-one CECL mark, which still had not been filed as of this writing, leaving exactly whether HomeTrust is paying up or stealing it unquantified.

Closing thoughts

A specific print settles this, so name it: the S-4 for the Blue Ridge deal, carrying pro forma tangible book per share and the day-one CECL mark on the acquired loans. That one document turns the merger from a story into a number, tells you whether you are buying a Virginia franchise below book or paying to inherit someone else's credit problem, and it lands on top of HomeTrust's own reserve-versus-noncurrent gap. An ambiguous print, a modest mark and a vague remediation timeline, leaves you holding a fair-value bank at 1.4 times book with rising past-dues and no cushion, and the honest move then is to wait for two clean credit quarters before adding. The fatter tail is the downside, because the falling reserve says the bank has not fully faced its own credit, and avoiding that permanent loss is worth more than catching the merger upside.

The bet, said again, is that HomeTrust keeps turning mountain-town deposits into loans that get paid back, and can absorb a bigger troubled bank in Virginia without importing its problems. What breaks it is already in motion: loans going bad faster than the reserve admits. The one pair of numbers that tells you first is noncurrent loans against reserves, 1.29% versus 1.11% today, and if that spread keeps widening next quarter the compounding story is on hold no matter what the merger deck promises.