Bid Cap
Company library Banks

Company report

BayFirst Financial Corp. BAFN

Three-pass checked

The bet you're really making is that BayFirst, a small Florida bank, has finally cleaned out the pile of tiny government-backed small-business loans that nearly wiped it out, and that the plain community bank left over is worth more than you pay. You're betting the fresh money it just raised covers the losses still buried in that old loan book. Right now it is ugly: a $32.7 million loss last quarter, and what the company says each share is worth on paper fell from $14.22 to $4.82 in three months. You pay about 1.4 times that written-down number, with no real earnings yet to point at.

Key data

Price$6.80
52-week range$4.26 – $11.34
P/E (trailing / forward)not meaningful, net loss
Price / tangible book1.41x

BAFN · price with moving averages

Daily · 6MWeekly · 3Y
$4$8$12$16$20 Oct '23May '24Dec '24Jul '25Feb '26Oct '26 Bid Cap
EMAs82140

Source: market data.

The business

BayFirst Financial Corp is the holding company for BayFirst National Bank, a roughly $1.1 billion community bank in St. Petersburg, Florida, formerly First Home Bank. For years its engine was not local lending, it was a national machine that churned out small SBA 7(a) loans, government-backed small-business loans by the thousand, sold the guaranteed 75% for an upfront gain, and kept the servicing. That gain-on-sale factory drove noninterest income to $58.9 million in 2024 and made the bank look bigger than its balance sheet.

In the fourth quarter of 2025, management shut the SBA 7(a) business. What is left is two things bolted together: a plain Tampa Bay deposit-and-lending bank, and a $142 million runoff pile of unguaranteed SBA loans, spread across more than 7,000 tiny credits, that keeps charging off. The moat is a local deposit franchise, and it is a thin one. Everything in the numbers below is the collision between the ordinary bank you would want to own and the loan book you would not.

The numbers

The quarters show a business whose old income source has flipped negative while credit costs ate the equity. Net interest income has settled near $9.4 million, but the gain-on-sale line that used to sit on top of it is gone, and fee income actually turned negative on writedowns.

Quarter endedNet interest incomeNet incomeDiluted EPS
2025-06-30$12.1M-$1.9M-$0.54
2025-09-30$11.3M-$19.1M-$4.71
2025-12-31$11.2M-$2.7M-$0.75
2026-03-31$9.4M-$5.9M-$1.54
2026-06-30$9.4M-$32.7M-$8.05

The June quarter is the reckoning: a $29.0 million provision and a $38.4 million total adjustment across the SBA book took the loss to $32.7 million. Reserves now stand at $45.1 million. Net charge-offs ran 2.08% of average loans, and net interest margin held at 3.48%.

Fiscal yearNet interest incomeNet income
2021$36.5M$24.6M
2022$30.0M-$0.3M
2023$36.4M$5.7M
2024$37.1M$10.2M
2025$45.2M-$24.6M
2026, 1H to June$18.9M-$38.6M

The compounding story is not slow, it is broken and restarted. Tangible book value per share went $20.95 a year ago, $14.22 at March, $4.82 now, a 77% destruction in twelve months. A capital raise lifted total equity to $115.9 million from $75.6 million, but even that fresh money did not prevent book from breaking to the downside. So this is not a cheap compounder on sale. It is a reset bank you are paying 1.4 times reset book to own.

The bank-level credit trend, from the FDIC call reports, is the whole argument, and it is still going the wrong way:

QuarterCharge-offs/loansNoncurrent/loansReserves/loans
2024-061.45%1.38%1.37%
2024-121.31%1.79%1.45%
2025-061.85%2.10%1.51%
2025-121.68%2.69%2.28%
2026-031.86%2.46%2.22%

Noncurrent loans nearly doubled while reserves built to catch up, and the June 10-Q pushed charge-offs to 2.08%. Reserves are keeping pace on paper, but they have been chasing, not leading. The variant this memo holds: the market prices survival at 1.4 times fresh book, but the question no reserve ratio has yet answered is whether $45.1 million is enough against a $142 million unguaranteed book that is still deteriorating, and only the next provision line settles it.

Management

The record is destruction, honestly disclosed. Over twelve months the board wiped out three-quarters of tangible book, eliminated the dividend, which had been $0.16 a year, and issued roughly four and a half times the share count in a rescue raise, taking shares from about 5.3 million to about 24 million. That was the right call for the bank and brutal for anyone who held it in. Insider activity confirms nobody stepped up on the open market: zero purchases in twelve months, and the only Form 4 sales were four nonmarket transfers by one director in July at reported zero value, plan status not disclosed. The one favorable read is that sophisticated outside capital funded the raise at these levels, which is a vote the bank survives, not that legacy holders were protected.

How it fails or surprises you

Reserve adequacy (downside). The $45.1 million allowance sits against a $142 million unguaranteed SBA book charging off at 2.08% and rising. One more quarter of $20 million-plus provisions takes tangible book below $4. The print that reveals it first: the Q3 provision line against that reserve.

Deposit flight (downside). Deposits are $989 million, down 9% in the quarter and 15% in a year, as broker and high-rate promotional money leaves. Noninterest-bearing accounts are only 11.8% of the base. If core deposits follow, funding costs climb and the 3.48% margin erodes. Watch Q3 total deposits and cost of deposits.

Clean runoff (right tail). If the $142 million book amortizes without fresh provisions and deposits stabilize, the recapitalized bank earns through. Even a modest 0.8% return on $1.1 billion of assets is about $9 million, near $0.38 per share on the new count, enough to re-rate a $6.80 stock the market cannot yet see through the SBA noise. The tell: the first clean profitable quarter.

Closing thoughts

Next quarter's provision expense settles it. If Q3 comes in near the $4 to $5 million charge-off run rate, the $45.1 million reserve was adequate, book steadies around $4.82, and BayFirst is a cheap-ish Florida turnaround at 1.4 times book. If Q3 carries another outsized provision, the reserve was short and book breaks below $4. The near-term left tail is fatter, because the credit series is still deteriorating rather than flattening, noncurrent loans up to 2.46% and charge-offs to 2.08%. What is at risk is another chunk of the freshly raised equity; what it is worth if runoff finishes clean is a return to real per-share earnings and a re-rate toward book. Call that judgment, not a probability.

The bet is still that BayFirst has cleaned out the tiny SBA loans that nearly killed it, and that the community bank underneath is worth more than 1.4 times its reset book. What breaks it is the $142 million runoff book and the $989 million of deposits. The one pair that tells you first is next quarter's provision against the $45.1 million reserve, and total deposits. A second straight quarter of $20 million-plus provisions would prove the reserve, and the thesis, wrong.

Methodology

Method: one-page read on a single name, filing-grounded, judged over a full credit cycle, not one quarter.

Numbers current to the 10-Q filed 2026-08-14 (period 2026-06-30); credit trend from FDIC quarterly call reports through 2026-03-31.

Q4 2025 (Dec 31) net interest income and net income derived as FY2025 annual figures less the sum of the first three fiscal quarters; EPS from the same calculation; filed annual XBRL carries -$24.6M net income for FY2025 and -$6.32 EPS.

Price-to-tangible-book derived from the $6.80 price and the filed $4.82 tangible book per share; trailing and forward P/E are not meaningful on a net loss.

Vendor market cap and share count in the feed are stale (pre-recapitalization) and were not used; per-share figures come from the filing.

Fact check: quarterly and annual dollar figures from as-filed XBRL, credit ratios from FDIC, tangible-book and provision detail from the 10-Q and Q2 investor call. Verified 2026-09-06.

Bid Cap

Daily ideas, a 390-name database, and a model long/short book from an investor who mostly covers financials. $70 a month or $700 a year.

Subscribe on Substack