$7.79
$25.44
bear $7.79
base $16.35
bull $25.44
margin of safety $7.75
market price* $23.40
Peoples Financial Corporation is the holding company for The Peoples Bank, based in downtown Biloxi, Mississippi, with 17 branches in Harrison, Hancock, Jackson and Stone Counties on the Gulf Coast. At the end of 2025 it had $727 million of assets, $604 million of deposits and $100.7 million of shareholders' equity, about $21.80 a share. At June 30, 2026 assets were $771 million, deposits $610 million and equity $101.7 million ($22.03 a share). WHAT THE SHARES HAVE ACTUALLY EARNED. Reported earnings per share were $0.03 (2016), $0.54, $0.13, $0.34, a loss of $0.52 (2020, as restated), $1.84 (2021), $1.91, $1.96, $4.66 (2024) and $0.85 (2025). Several of these overstate the bank's normal earning power. 2021 included a $5.7 million release of loan-loss reserves set up in 2020; 2022 included a $2.4 million tax benefit; and 2024 included a one-time $13.3 million tax benefit from reversing a reserve against deferred tax assets in the third quarter. Before that benefit, 2024 pre-tax income was $8.4 million, or about $1.46 a share after a 19% tax charge (the rate the bank actually bore in 2023 and 2025). From 2014 through 2020 the bank lost a net $11.9 million in total, with loan-loss provisions of $7.4 million in 2014, $2.6 million in 2015 and $6.0 million in 2020. Net interest income, the bank's main revenue, peaked at $26.5 million in 2023 and fell to $23.5 million in 2024 and $20.4 million in 2025, as the Federal Reserve cut rates by 100 basis points in 2024 and 75 in 2025 (per the company) and as deposits left. THE EARNINGS LEVEL WE USE. We use 2025's $0.85 a share (pre-tax income $4.8 million) as the current level, not the 2023-2024 results, because the deposit base and rate conditions behind them have since changed. The first half of 2026 supports that level: net income was $2.34 million ($0.51 a share) against $2.55 million in the first half of 2025; the twelve months to June 30, 2026 come to $0.80 a share and the first half annualized to $1.01. The second quarter was weaker than the first ($0.90 million against $1.45 million; net interest income $4.92 million against $5.77 million), and we could not tell from the available data what drove that sequential drop. DEPOSITS, FUNDING AND CAPITAL. Approximately half of deposits are fully FDIC insured, and the vast majority of the uninsured balance is public funds (county and municipal) secured by pledged bonds; the bank also names casino-industry customers whose balances move, and it competes in bids for public funds throughout the year. In 2025 deposits fell $116.3 million (16%) to $604.4 million after the loss of several public-fund accounts, including a $62.3 million drop in non-interest-bearing demand deposits. The bank met that outflow mainly by drawing down cash, which fell $88.2 million, not by borrowing: advances from the Federal Home Loan Bank were only $0.8 million at year-end. In 2026 deposits held steady ($609.8 million at June 30) while assets grew $43.9 million and Federal Home Loan Bank advances rose to $38.5 million; at year-end the bank had a further $121.4 million available on its Federal Home Loan Bank line and $28 million of other contingency lines. We found no financial covenants described on the borrowings in the annual report. The bank's community bank leverage ratio was 15.76% at the end of 2025 against the 9% level required under that framework, and dividends from the bank need prior approval of the Mississippi Department of Banking and Consumer Finance. COSTS AND THE BOND PORTFOLIO. Non-interest expense has stayed between $21.5 million and $23.2 million every year since 2016 and was $22.7 million in 2025, about 82 cents for each dollar of net interest and non-interest income, which is high for a bank. The bank holds $414 million of bonds at cost earning about 3.0%, and they are carried $33.8 million (8.2%) below cost: $25.1 million on bonds marked to market, which has already reduced equity by $28.5 million after tax ($6.17 a share, a loss that would reverse if the bonds are held to maturity), and $8.6 million on bonds carried at cost that is not reflected in book value ($1.87 a share before tax, which would put book value near $19.93). Loans were $263.7 million at the end of 2025, up $30.1 million (13%) in the year but after a decade of shrinking; loans yield about 6.0% against 3.0% on the bonds. Credit quality is clean today: nonaccrual loans were $0.5 million (0.20% of loans) and the allowance was $2.9 million (1.1%). BASE CASE, $16.35 A SHARE. Starting from $0.85, we credit one thing: bonds that mature within five years ($183 million at cost, 44% of the portfolio, by their contractual dates) are reinvested at current yields. The $33.8 million markdown implies replacement yields roughly 1.5 to 2 points above the 3.0% the bonds earn (our estimate); we credit only 1.0 point, worth $1.8 million before tax or $0.32 a share after a 19% tax, phased in evenly over five years. We charge against it the cost creep of the last three years: expenses grew 1.3% a year from $21.9 million in 2022 to $22.7 million in 2025 while non-interest income grew 1.2% a year, a net squeeze of about $0.2 million a year before tax ($0.035 a share). Earnings per share therefore go from $0.88 in year one to about $1.00 in year five and stay flat. Last year's earnings fell from $1.46 (2024 after a normal tax charge; $4.66 as reported) to $0.85, so year one at +3% is above the latest actual change. The dated reason: the six months to June 30, 2026 (reported August 14, 2026) already earned $0.51 a share, a $1.01 annual pace, and deposits were stable, so $0.88 sits below that pace. Left out because we cannot point to evidence that it is in hand: further loan growth (one year after a decade of shrinkage), new deposit wins, margin gains from rate moves, cost cuts, any sale of the bank, and bonds maturing after five years or paying down early. Beyond year ten we keep earnings flat, resting on a structural franchise (a $604 million deposit base and branch network in place) rather than any growth. BEAR CASE, $7.79. Two named problems arrive and then a credit cycle. First, public-fund deposits keep leaving: another $60 million (9.9% of deposits, about half of 2025's loss) is funded by running down bonds earning about 3.0% while the deposits lost cost an average 1.63% (the 2025 rate on interest-bearing liabilities), a net $0.8 million before tax or $0.14 a share, and maturing bonds go to fund outflows rather than being reinvested at higher yields. Second, expenses keep creeping up as above, so earnings drift from $0.67 in year one to $0.53 by year five. Third, in years four to six the bank goes through a credit stretch like its own 2014-2016 record, when it booked $7.4 million, $2.6 million and $0.6 million of loan-loss provisions: we use $10.55 million over three years, 4.0% of today's $263.7 million loan book, which costs $8.5 million after tax ($1.85 a share) and leaves earnings at about minus $0.05, minus $0.09 and minus $0.09 a share in those years before returning to $0.53, flat thereafter. Equity would fall about $8.5 million from $100.7 million; given the 15.76% leverage ratio we assume no share issuance (a judgement). This case is worth well under book value because the bank earns about 4% on its equity, and a bank earning that little is worth less than the capital it holds. BULL CASE, $25.44. The bank's own levers work: it starts from the first-half 2026 pace of $1.01, holds expenses flat as it has for a decade (no creep), and earns the full yield gap implied by the bond markdown (about 1.8 points, our estimate) on the $183 million maturing within five years, worth $0.58 a share after tax. Earnings reach $1.59 by year five and stay there. DEBT. As a bank, deposits and borrowings are its raw material and their interest is already deducted in the earnings above, so no separate debt is subtracted in any case. Each value is the value of earnings after funding costs: $16.35 base, $7.79 bear, $25.44 bull, with $38.5 million of Federal Home Loan Bank advances at June 30, 2026 already paid for out of those earnings. MARGIN-OF-SAFETY PRICE, $7.75. How sure are we of the earnings level? Moderately: 2025 ($0.85), the trailing twelve months ($0.80) and the first-half 2026 pace ($1.01) bracket it, but the deposit base proved unstable in 2025 and the second quarter of 2026 slipped. The worst plausible outcome is the bear case: more public-fund loss followed by a credit stretch. The structure survives it: capital is far above requirement, borrowings are small, and the bank has contingency funding lines, though it depends on public-funds depositors and on a state regulator's approval to pay dividends. Outcomes we do not model include a major storm on the Gulf Coast and regulatory restrictions on dividends. If we are wrong: earnings stuck at $0.85 forever would be worth about $14.17 a share, and earnings drifting down to $0.55 about $9.17. The price assumes a five-year holding period. The bank paid $0.36 a share in 2025 (down from $0.44 in 2024 and $0.53 in 2023), covered 2.4 times by 2025 earnings, and $0.18 more in the first half of 2026. At $7.75 the shares would yield 11% on 2025 earnings and 4.6% in dividends and cost 36% of book value. A 20% annual return over five years means $7.75 must become $19.28; after about $1.80 of dividends, the shares must be worth about $17.50, which is roughly 71% of a book value of about $24.70 that we would expect by then if the bank earns as in the base case and pays $0.36 a year (before counting any recovery of the bond markdown). That is why the margin-of-safety price is this low: the bank's own earnings history gives no reason to credit improvement that has not yet appeared. CHANGES FROM OUR EARLIER VIEW (dated September 3, 2026; the earlier values were built differently and are not comparable). We now include the June 2026 quarter, which shows stable deposits and rising Federal Home Loan Bank borrowings. On a closer reading we also corrected two points: the 2025 deposit outflow was met mostly by drawing down cash, with borrowings only starting to rise in 2026, and the 2021-2022 earnings were boosted by a reserve release and a tax benefit rather than only by higher rates. We also now credit the bond repricing explicitly while charging for expense growth, and the downside includes a full credit stretch. WHAT COULD MAKE THIS WRONG. Deposits resuming their decline faster than assumed, the second quarter's weakness continuing, or expenses rising faster than 1.3% a year would pull the base case toward the bear case; a stable deposit base with bonds repricing faster than we credit would move it toward the bull case.