Simmons First National: The Dividend Is Solid; the "Upside" Is a Repricing, Not a Discount

Open the quote for Simmons First National (SFNC) and the screen looks broken. The price-to-earnings multiple is negative. The dividend payout ratio shows as a negative number, too — which, on a company, usually means it is paying shareholders out of a hole it has just lost money into. And yet the same bank has just raised its quarterly dividend to $0.215 a share, its fifteenth straight annual increase in a 117-year unbroken paying record. At $22.23, that works out to a roughly 3.9% yield. How can a stock that pays and keeps raising a dividend look, on the numbers, like it is losing money?

The short answer is that the scary multiple is a one-time event sitting in the trailing year, not the business. The longer answer matters, because it separates the part of this stock you can prove from the part that is priced on hope.

What actually happened in the third quarter of 2025

In the third quarter of 2025, Simmons reported a net loss of $562.8 million, or $4.00 per share. That single quarter is what flips the trailing earnings multiple negative. It was not an operating collapse. It was a deliberate, one-time transaction.

Simmons had spent years funding a big loan book with relatively expensive wholesale and public-fund deposits while its bond portfolio earned little — a situation where the bank was paying more to borrow than it earned on those assets. To fix that, it sold roughly $2.4 billion of low-yielding securities and used the proceeds, along with $327 million of newly raised common stock sold at $18.50 a share, to pay down the more expensive funding. Selling bonds at a loss is painful on the income statement — the after-tax loss on those sales was about $626 million — but it permanently removes the low-return assets and the high-cost liabilities that were dragging on the margin.

That is the whole trick, and it is worth understanding because it reframes the loss. A bank that books a large one-time loss to reposition its balance sheet is not a money-losing bank; it is a bank that traded a clean future margin for a messy past quarter. Management called the quarter "transformative," and the quarters that followed back up the intent.

The earnings engine underneath the loss

Strip out the one-time item and the pattern is a bank quietly compounding. Adjusted (non-GAAP) diluted earnings per share ran $0.46 in the third quarter of 2025, $0.54 in the fourth, $0.47 in the first quarter of 2026, and $0.50 in the second. That adds up to a trailing twelve months of roughly $2.04 in normalized earnings per share. The trailing GAAP number is far lower — and negative — only because the $562.8 million loss is in the window.

What the repositioning did is visible in the margin. The fully taxable-equivalent net interest margin, the core measure of how much a bank earns on each dollar of assets it funds, jumped from 3.06% a year earlier to 3.84% by the second quarter of 2026 — and has held steady. The cost of its deposits sat at 1.93%. Loans grew 6% year over year to about $18.1 billion. The revenue line, $248.6 million in the second quarter of 2026, is up 16% from a year ago. The engine is not broken; it is re-based on a higher margin than it had in the recent past.

That is the provable part of the story, and it is the part a dividend investor should lean on.

Is the dividend safe?

This is where the value question gets concrete, because the whole point of holding a bank for income is that the payout is real and coverable. The quarterly dividend is $0.215, an annualized $0.86 per share, and it has now been raised for the fifteenth consecutive year. Against the roughly $2.04 of normalized trailing earnings, that dividend represents about a 42% payout — comfortably covered, with the company retaining most of its earnings to grow the book.

The negative payout ratio on the screen is the same artifact as the negative multiple: it is dividing a dividend by a negative trailing net income. Measured against actual, ongoing earnings, the coverage is solid. Simmons also has a $175 million buyback program on the table (about $161 million still available after buying back 0.7 million shares in the second quarter of 2026), so it is returning capital both ways. If you are buying this for the income, the payout is not the risk.

Now the "upside" — and where the headline is doing the work for you

Here is where I would push back on the framing of "upside." The average analyst price target sits around $24.50, which from $22.23 is a modest double-digit gain. That is a re-rating on the back of the higher margin, not a bargain. To judge whether the price is cheap, you have to look at the right book.

There are two "books." Book value per share, the accounting figure, was $24.11 at the end of the second quarter of 2026 — and because the stock trades below it, SFNC shows up as "trading below book" (a price-to-book of about 0.92), a classic value screen. But a meaningful chunk of that book is goodwill from the 2022 Spirit of Texas acquisition: about $9.69 per share. Strip that out and tangible book value per share is $14.42. Against that, the stock is trading at about 1.5x. That is not a discount. Comparable regional banks trade at roughly 1.0x to 1.3x book. On tangible value, Simmons is actually at the expensive end of its peer group, not the cheap end.

The honest way to state the valuation: at about 10.9x normalized trailing earnings and 1.5x tangible book, Simmons is in line with — and on book, pricier than — its regional peers. The "below book" screen is flattered by acquisition goodwill. The upside investors are paying for is the assumption that the higher 3.8% margin persists and compounds at the mid-teens return on tangible common equity the bank has been running (16.1% adjusted in the fourth quarter of 2025), gradually rebuilding the tangible book the repositioning and the equity raise diluted. Note what the repositioning did to that book: tangible book per share fell from $16.97 to $14.42 over the last year. The equity raise bought a better margin, but it bought it at the cost of per-share book. The repair is a forward compounding story, not a below-book entry you can point to today.

The test that decides it

So the whole thesis rests on one financial gate, and it is not the dividend. It is whether the re-based margin holds. The 3.8% net interest margin depends on deposit costs staying low and loan yields staying up. If that holds, the normalized earnings engine keeps compounding, the dividend keeps growing at its historical pace, and the 1.5x tangible book is defensible. If credit or rates erode it, there is less cushion than a below-book price would suggest.

The credit line is worth watching, not alarming. Nonperforming loans rose to 0.92% of the loan book, driven largely by a single real-estate construction relationship; net charge-offs were 20 basis points for the quarter, and the provision actually exceeded charge-offs, which is the conservative direction. Coverage of nonperforming loans is 143%. None of that is alarming, but it is the number to track alongside the margin, because a spike there would pressure both the earnings engine and the compounding that the "upside" depends on.

What this is for a portfolio

This is an income anchor, not a deep-value compounding play. The dividend is provable and well covered — that part of the headline holds. The "upside" is real but modest, and it rests on a re-based margin compounding into a diluted book, not on buying tangible assets below price. If you want a 3.9% yield that has a 117-year record and a payout comfortably inside earnings, and you are comfortable that the stock is fairly valued to slightly rich rather than cheap, it fits. If your plan depends on a large valuation gap to book closing, this is not that stock — the gap is an accounting artifact, and the multiple says so.

The thing to watch each quarter is the same: is the net interest margin still near 3.8%, and is deposit cost still low? While the answer is yes, the dividend keeps compounding. When it is no, the 1.5x tangible book is where you will find out whether the price was a re-rating or a reach.