Franklin Financial Services Corporation (FRAF): The Trust Franchise Repricing

The holding company, organized in the early nineteen eighties and listed on the Nasdaq Capital Market, conducts substantially all of its business through Farmers and Merchants Trust Company of Chambersburg, a Pennsylvania-chartered bank founded in nineteen zero six that operates twenty-three offices across Franklin, Cumberland, Dauphin, Fulton and Huntingdon counties in Pennsylvania and Washington County in Maryland. The FDIC Summary of Deposits shows the bank holding the large majority of its home county deposit base, a share of about one-third that no challenger can match without buying it, and the same table shows the bank near half of the deposit market in Fulton County. The strategic consequence of that position is pricing power on the liability side, and the average cost of all deposits fell to roughly one and a half percent in the most recent quarter from near two percent a year earlier.

The second pillar is wealth management, a trust and brokerage franchise that held assets under management of well over a billion at mid-2026, up from near a billion and a half a year earlier, and generated a meaningful quarterly fee line. The trust franchise matters for two reasons that show up in the financial statements. First, it is the source of the relationship: a client who leaves assets at the bank after an estate settlement tends to keep the checking account, the home loan and the credit line, which is why a trust-heavy community bank can hold deposit beta lower than a peer of similar size. Second, fee income at a little over one-fifth of revenue is sensitive to the interest rate environment only in a thin way, which cushions the income statement when the yield curve does what it did in the early part of the decade.

The franchise sits in a rural Pennsylvania market with the demographic and competitive structure of rural Pennsylvania: the bank is the largest institution headquartered in its home county, but the primary trade area overlaps with regional banks, out-of-state entrants and a network of credit unions that all compete for the same deposit base. The market also carries the single biggest structural weakness in the company's own disclosure, which is the office real estate line in the commercial real estate portfolio. The bank's regulatory non-owner-occupied commercial real estate concentration ratio stood at well over three times Tier 1 capital plus the allowance at the start of the year, and that number is the reason the two credits that went on nonaccrual in the most recent quarter should be read as a trend line, not an anomaly.

The deposit moat is the franchise. A one-third county deposit share in a market of a few billion means the bank can price deposits on relationship and convenience rather than on rate, and the math of that position shows up directly in net interest margin. The average yield on interest-earning assets in the most recent quarter sat just above five percent against a cost of funds under two percent, and the spread between those two numbers is wider than what any regional bank paying over two percent for deposits can achieve in the same rate environment. The moat is durable so long as the bank continues to be the institution of record in its home county, which is a function of a century of branch presence, the trust franchise, and the fact that the next closest challenger does not come close to matching the county deposit share.

The trust and wealth management product line is the second moat and the one that is harder to replicate. The franchise manages the bulk of its assets under management in bank-held trust accounts with a smaller sleeve at third-party brokers, and the fee structure is tiered by assets, which means the revenue line scales with the balance sheet of the client base rather than with the rate environment. The moat here is the credential: a Pennsylvania-chartered trust company with a hundred years of estate administration is not something a de novo or a fintech can substitute for, and the estate management fee line, a modest single-digit percentage of the total wealth fee, understates the cross-sell value of the relationship.

The loan book is concentrated in the ways that are normal for a bank of this size and in one way that is not. Commercial real estate averaged the largest share of gross loans, well past half the book, and the municipal and public sector line sits as the single largest commercial sub-sector. The concentration is the price of the margin, because the bank can underwrite rural Pennsylvania commercial properties better than any outsider and price them accordingly. The office real estate sub-line is the part of the book where that advantage is being tested, and the two nonaccruals that drove the recent provision are both in that sub-line. The technology stack is not a moat and is not a meaningful cost center; the franchise competes on relationship, not on platform.

The most recent quarter produced net income in the mid single digits of millions, up roughly twelve percent year over year, with diluted earnings per share near a dollar and a half against just over a dollar in the prior-year quarter. The engine of the increase is net interest income, which grew to just under twenty million from the mid-teens, driven by the deposit cost reset rather than by volume. The average cost of funds fell nearly forty basis points and the deposit cost fell a similar amount, and those two lines explain most of the roughly two million dollar increase in tax-equivalent net interest income. The loan yield was essentially flat near six percent, which means the margin expansion is a liability-side story, and that distinction matters for how long the expansion lasts.

The first half produced net income near thirteen and a quarter million, up a third from the year-ago period, with diluted earnings per share near three. The efficiency ratio improved to near sixty-one percent from the high sixties, and the improvement came from both sides: noninterest income grew at a high single-digit rate while noninterest expense grew only at a low single-digit rate, with the largest expense line, salaries, up barely three percent. The dividend line is the other part of the story: the board declared a quarterly dividend up three percent over the year-ago quarter, and the payout ratio sits in the low twenties, which is low enough to support further increases without straining the balance sheet.

The credit line is where the quarter broke. Nonaccrual loans doubled from the year-end level, and the nonperforming loan ratio moved up to just over one percent of gross loans from about half a percent. The provision for credit losses on loans came in near a quarter and a half million in the quarter against barely seven hundred thousand a year earlier, and the increase was driven by specific reserves on two collateral-dependent credits of about a million. The allowance for credit losses was near one and a third percent of loans at quarter end, up from just over one and a quarter percent at year-end, which is a meaningful build but not yet a signal that management sees a portfolio-wide problem. The net charge-off line was a recovery for the first half, which means the credit losses are still sitting in the allowance rather than hitting the income statement as actual losses.

The earnings trajectory for the second half depends on the deposit cost structure continuing to hold. The cost of deposits at mid-2026, just above one and a half percent, is the single most important number in the model, and it is a number that can move quickly if the competitive dynamics in the home county shift or if the rate environment forces the bank to re-price its time deposit base. The bank carries two hundred million of Federal Home Loan Bank advances on the balance sheet at over four percent, and the ability to roll those advances into cheaper deposits is what keeps the blended cost of funds under two percent rather than at the rate the advances actually cost. The margin of about three and a half percent in the most recent quarter is therefore partially a function of funding mix, and a shift in that mix is the most direct threat to the earnings line.

The commercial real estate credit line is the second execution variable, and the two nonaccruals are the leading indicator. The first is a construction loan on a mixed-use commercial project in the high single digits of millions that matured and went on nonaccrual in the most recent quarter, and the bank committed to advance up to a quarter million more in funding to fully enclose the property and protect the collateral, with a substantial portion already advanced. The second is a loan in the mid single digits of millions secured by a group of commercial office buildings that is a purchased participation where the bank is not the lead lender, and that loan went on nonaccrual in the quarter with a specific reserve based on a recent appraisal. The combined exposure is a bit under two percent of gross loans, and the specific reserves on the two credits total near two million, which means the bank has reserved roughly an eleventh of the combined balance. The question that matters for the thesis is whether these two credits are the last two to go on nonaccrual or the first two of a line that keeps producing them.

The wealth management line is the third execution variable and the one with the least visible risk. Assets under management are growing, the fee line is up at a mid-single-digit rate year to date, and the trust franchise is a structural advantage that does not depend on the rate environment. The risk in the wealth line is client attrition and the aging of the trust client base, which is a slow variable that shows up in the fee line over several years rather than in a single quarter. The capital return line is the fourth variable: the share repurchase plan authorized in the winter saw no open-market buys in the first half, with only a small tranche purchased outside the plan to fund the dividend reinvestment plan, and that inaction at a premium to book is a data point that suggests the board is not yet treating the share price as cheap enough to accelerate buybacks.

The office real estate credit line is the largest single risk in the franchise, and the two nonaccruals are the evidence that the line is live. The office building loan is a purchased participation, which means the bank's loss is bounded by its participation share but the recovery depends on the lead lender's workout process, and the specific reserve reflects a recent appraisal that the bank has accepted as the recovery basis. The construction loan is the more active risk: the bank is advancing additional funding to protect the collateral, which is the right call in the short term because an unfinished building loses value faster than a completed one, but it also means the bank's exposure is growing as the funding commitment is drawn down, and the specific reserve is based on a discounted as-is appraisal that requires a refresh as the project progresses. The combined exposure in these two credits is a little over one percent of total shareholders' equity, which is a number that is manageable if it stays contained and a number that is not manageable if it is the first tranche of a broader office line stress.

The deposit funding risk is the second downside scenario and the one that would hit the income statement fastest. The cost of deposits is supported by a large noninterest-bearing checking base and a transaction account base of over a billion, and both lines are sensitive to the competitive environment in a county where the bank's share is high enough to be visible to every challenger. A shift of one percentage point in the average cost of time deposits would cost a meaningful annualized amount of interest expense, and the same shift across the full transaction account base would cost more than a million, which is roughly one-eighth of the quarterly net income line. The Federal Home Loan Bank advances provide a floor under the funding structure, but the bank is already at a substantial level of those advances and the cost of rolling them at current rates is a margin headwind, not a margin tailwind.

The third downside scenario is a continued expansion of the nonaccrual line beyond the two current credits. The watch list, which includes loans rated below pass, stood in the low fifties of millions at quarter end, down from the high fifties at year-end, which is a positive number but one that does not eliminate the risk that additional credits migrate into nonaccrual status in the second half. The allowance, at near one and a third percent of loans, is adequate for the current portfolio but would need to be built significantly if the nonaccrual line moves toward two percent of loans, and a build of that magnitude would consume a meaningful portion of the quarterly earnings line. The negative mark-to-market securities balance of over twenty million is a separate risk that does not hit earnings but does reduce tangible book value and compress the multiple the market is paying, and a further rise in rates would deepen that hole without any offsetting income statement benefit.

The stock trades in the low sixties per share, a market capitalization near three hundred million on a little under four and a half million shares outstanding, and a premium of roughly a half again over book value. The trailing twelve-month diluted earnings per share near five puts the price at a little over thirteen times trailing earnings, and the year-to-date multiple on earnings near three is near ten and a half, which is the number the bank itself uses in its performance table. The annualized dividend of a little over a dollar and a third yields just over two percent at the current price, and the payout ratio in the low twenties means there is substantial room for the dividend to grow without any change in the earnings line. Tangible book value, after deducting the goodwill from the bank's last acquisition, puts the market price at well over one and a half times tangible book, and that is the multiple that matters most for a bank whose unrealized securities loss balance is dragging reported book value down.

The bear case values the stock at book value, roughly forty per share, on the assumption that the commercial real estate credit line continues to stress and the deposit cost advantage erodes. That scenario requires the nonaccrual line to move toward two percent of loans or higher, the deposit cost to rise by a full percentage point, and the wealth management growth to stall, and the combination would compress the return on equity from near fifteen percent toward ten percent and the multiple from a half again over book toward par in a single re-rating. The probability of that full bear case playing out is low, but the partial version, in which the credit line adds another two or three million of specific reserves and the deposit cost rises half a point, would put pressure on the current price and is the scenario that the thirteen times trailing multiple does not fully discount.

The base case holds the deposit cost near one and a half percent, the margin between three and a half and three and three-quarters percent, and the earnings line in the low twenties of millions for the full year, which supports a return on equity near fourteen percent and book value growth of roughly six or seven percent per year after dividends. At those numbers, the multiple is fair to modestly rich, and the two percent dividend yield plus the book value growth produces a total return of roughly eight percent per year, which is the number a patient holder should expect in the base case. The bull case, which requires the commercial real estate line to resolve with the current reserves, the deposit cost to hold, and the assets under management to keep growing at a double-digit rate, supports a return on equity in the mid-teens or higher and a multiple of a bit more than one and a half times book, which would put the stock in the mid-seventies on current book value plus one year of earnings retention.

The multiple analysis for the bull and bear cases rests on the same underlying book value and the same dividend line, and the difference between the two is entirely a function of the return on equity that the market assigns to the franchise. At a fifteen percent return on equity with a payout in the low twenties, the stock is a capital return story, and the two percent yield plus the buyback plan that the board has not yet activated are the two levers that determine whether the current price is a floor or a ceiling over the next two years. The authorized repurchase plan of one and a half hundred thousand shares, which remains fully unexecuted in the first half, is the single most actionable variable in the valuation framework, and a meaningful acceleration of that plan at the current multiple would be a signal that the board itself considers the share price to be within the zone where it is cheap to buy back the shares.

The franchise is real and the deposit moat is the strongest part of it. A one-third county deposit share, a cost of deposits near one and a half percent, and a trust franchise holding well over a billion in assets under management are not numbers that can be built quickly or cheaply by any competitor, and the three and a half percent margin that those numbers produce is the reason the stock re-rated from near par a year ago to a half again over book today. The market has priced that re-rating correctly at the low end: thirteen times trailing earnings on a fifteen percent return on equity is not expensive by community bank standards, and the two percent dividend yield with a payout in the low twenties gives the shareholder a floor under the equity that does not depend on the multiple holding.

The counterargument is the credit line. The nonaccrual ratio near one percent of loans, the specific reserves added in a single quarter of about a million, and the commercial real estate concentration ratio at well over three times capital are not numbers that belong in the story of a bank being re-rated to a premium multiple, and the honest read is that the market has priced the deposit franchise at the top of its range while still carrying the full credit risk at face value. The two nonaccruals are contained, the specific reserves are in place, and the watch list is trending down, but the office real estate line in rural Pennsylvania is a line that can produce more nonaccruals in the coming year, and the premium multiple does not leave much room for a second or third credit to go on nonaccrual without the stock taking a meaningful step back.

The judgment on the stock at the current price is that it is fairly valued on the base case and slightly rich on the numbers the market is currently paying. The fifteen percent return on equity is the number that supports the multiple, and the deposit cost near one and a half percent is the number that supports the return on equity, and both of those numbers are holding at mid-year. The two percent dividend yield is the shareholder's protection against a multiple compression, and the unexecuted repurchase plan is the option the board holds to narrow the gap between book value and market price. The stock is not a value at the current multiple, and it is not a momentum trade at thirteen times trailing earnings, and the correct framing is a quality compounder being paid a quality multiple for a franchise doing exactly what it is supposed to do, with a credit line that requires a quarter of attention each time the nonaccrual ratio is reported.

The four variables to watch are the cost of deposits, the nonaccrual loan ratio, the commercial real estate specific reserve line, and the execution of the share repurchase plan. If the cost of deposits holds near one and a half percent or below, the nonaccrual ratio stabilizes at its current level or moves down, the specific reserves on the two current credits do not need to be built further, and the board begins to execute the repurchase plan at or near the current price, the present level is the floor and the mid-teens return on equity in the bull case is the destination. If any of those four variables moves in the other direction, the premium multiple compresses first, and the two percent dividend yield is the only part of the equity that does not move with the multiple. The franchise deserves the multiple it is paying, and the credit line is the reason the multiple should not be paid more.