$18.85
$68.76
bear $18.85
base $51.64
bull $68.76
margin of safety $22.00
market price* $31.19
EVERTEC is a payment-processing and financial-technology company based in Puerto Rico. It runs the ATH debit network and the ATH Móvil payment app, provides merchant card acquiring in Puerto Rico and the Caribbean, and provides technology outsourcing for Banco Popular, Puerto Rico's largest bank. It also sells payment processing and core-banking software to banks across Latin America. Brazil is now its largest market there, through Sinqia (bought in late 2023), 75% of Tecnobank (bought in October 2025 for about $150 million) and Dimensa (bought from TOTVS on April 30, 2026 for about $199 million). 2025 revenue was $931.8 million. First-half 2026 revenue was $522.7 million, up 14%. Latin America rose 42% on the acquisitions, the Puerto Rico payments business rose 7% and merchant acquiring 6%. Business Solutions, mostly work for Popular, fell 9%. The Popular relationship. Popular produced about 29% of 2025 revenue. That share will likely be smaller in 2026 as the Brazilian purchases add revenue. The relationship rests on three separate agreements with different end dates: - The master services agreement covers Popular's core banking, card processing and other technology work. It runs to September 30, 2028. - The ATH network agreement runs to 2030. - The merchant-acquiring agreement, under which Banco Popular sponsors EVERTEC and refers merchants to it, runs to 2035. Since the 2022 renegotiation, Popular is no longer required to use EVERTEC exclusively. It pays guaranteed annual minimums only through September 2028, and it has received a 10% discount on certain services since October 2025. In the first half of 2026, Business Solutions revenue fell about $12 million, mainly because of that discount, while its profit fell only about $4 million. Whether the services agreement is renewed, and on what terms, is the largest single uncertainty in this company. It is not, however, all-or-nothing. EVERTEC runs systems at the heart of Popular's operations, which would take years to replace, and we see no sign in the filings that Popular has begun to do so. The 2022 renegotiation shows the more likely risk is a lower price, not a lost customer. For scale: even if EVERTEC lost the entire agreement in 2028 and could cut none of the costs behind it, our base case would fall to roughly $32 per share. It would not fall to zero, because the debit network, merchant acquiring and Latin American businesses would remain. How we measure earnings. We start from operating cash flow, remove working-capital swings, deduct stock compensation as a real cost, and deduct only the maintenance part of capital spending (roughly $80 million a year, mostly capitalized software). On that basis, owner earnings after interest were $1.35 per diluted share in 2016, $1.80 in 2020, $2.33 in 2021, $1.82 in 2022, $1.18 in 2023, $2.49 in 2024 and $2.65 in 2025. Amortization of acquired intangibles is large, about $65 million in the first half of 2026. We add it back as a sunk acquisition cost and instead carry the wear-out of acquired customer lists in our growth assumptions. Because EVERTEC carries substantial debt, we value the business before its lenders and then subtract what it owes. Adding back cash interest paid ($72 million in 2024, $64 million in 2025), the business produced about $234-235 million a year of cash in both 2024 and 2025. We measure from 2024 because the Sinqia purchase changed the business. To that level we make two adjustments: - We add about $17 million for a full year of Tecnobank and Dimensa. EVERTEC owns both and paid for them with borrowed money, which we subtract below. This is our estimate from the rise in Latin America segment earnings, since the company does not disclose their profits separately. We hold it below what the second quarter alone would suggest until more quarters with Dimensa are reported. - We subtract about $6 million for the full-year effect of the Popular discount. That gives a starting level of about $245 million a year. First-half 2026 profit was also reduced by items we do not treat as recurring: about $9 million of acquisition and refinancing fees, a tax charge on moving cash out of a foreign subsidiary to help pay for Dimensa, and a write-off of an equity investment. What it owes. At June 30, 2026, debt was $1.29 billion: - a $392 million Term A loan and a $35 million revolving credit line, both due December 2027 - an $858 million Term B loan due 2030, enlarged in November 2025 and May 2026 to fund the acquisitions - about $9 million of notes and deferred purchase payments We also subtract $78 million for the outside owners of partly owned subsidiaries, including the 25% of Tecnobank that EVERTEC does not own, and about $6 million for the BB Chain stake bought in July 2026. The total is about $1.38 billion. The loans float at a benchmark rate plus 2.00-2.25%, close to our 6% rate, so we make no adjustment for their cost. We treat the $261 million of cash as working cash for settlement and operations across the region; moving it out of foreign subsidiaries triggers tax, as the 2026 charge showed. Interest runs about $80 million a year. Growth check. Cash flow before interest rose only about 0.6% in 2025. Owner earnings per share rose 6.4%, mostly because of buybacks. Our base case therefore assumes no growth in the first year from the $245 million level. Base case, about $51.64 per share. Cash flow holds at $245 million in the first year and then grows 2% a year. We credit that growth to the ATH network and Puerto Rico merchant acquiring, which have grown for many years, net of erosion in Popular work. EVERTEC's revenue has risen every year for a decade. When the services agreement comes up in September 2028, we assume a price concession of about $12 million a year, similar to the 2022 renegotiation. After year ten we hold cash flow flat. The debit network, the merchant agreement to 2035 and the installed banking software are durable, but we credit them with no further growth. We give no credit to future acquisitions or buybacks. - The business is worth about $4.49 billion. - It owes about $1.38 billion. - That leaves about $3.11 billion, or $51.64 per share on about 60.3 million diluted shares. Bear case, about $18.85 per share. The case's assumptions: - Tecnobank and Dimensa earn about $5 million a year less than we estimate. - Nothing grows until 2028. - Popular then moves or reprices a large part of its outsourced work, cutting about $40 million a year from 2029. - Cash flow then declines 2% a year through year ten and 3% a year after that. This reflects three named pressures: the attrition of acquired Latin American customers that the company already reports, a weaker Brazilian real, and a further concession when the ATH agreement with Popular comes up in 2030. - The Term A loan and revolving line due December 2027 are refinanced at a spread one point higher, adding about $14 million to what is owed. The business is worth about $2.53 billion and owes about $1.39 billion, leaving about $1.14 billion, or $18.85 per share. Even in this case cash flow stays above $170 million a year against about $80 million of interest, so the debt remains serviceable. Bull case, about $68.76 per share. The services agreement renews on today's terms, and Tecnobank and Dimensa earn at the top of our estimate. The Puerto Rico payments and merchant businesses keep growing close to their recent pace, lifting cash flow 4.5% a year for five years and 3% for the next four, then flat. The business is worth about $5.52 billion and owes about $1.38 billion, leaving about $4.15 billion, or $68.76 per share. How we set the margin-of-safety price of $22. The range between the cases is wide: the bear case is barely more than a third of the base case. Most of that gap turns on a decision Popular will make within about two years. We therefore set the price so the downside is protected rather than relying on the middle outcome: - From $22, the bear case implies roughly breaking even over three years, once the cash the business generates in the meantime is counted. - The base case implies more than 20% a year even if it took five years for the share price to reflect the business's value. - If Popular renews on reasonable terms, the bear case's largest risk disappears and that value is likely to be recognized sooner. The margin-of-safety price gives no credit for that renewal in advance. The price also leaves room for two risks we could not measure: the pending tax grant described below, and the estimated earnings of the new Brazilian businesses. What changed since our previous estimate. Our values are close to before: base $51.64 versus $50.82, bear $18.85 versus $19.55, bull $68.76 versus $65.73. The base and bull values rose mainly because buybacks reduced the share count to about 59.8 million at June 30, 2026. The bear value moved slightly lower after we rebuilt that case year by year on the updated figures. The main change is the margin-of-safety price, cut from $29 to $22. From $29, the bear case would have implied a loss of about a third, and with the Popular decision due in 2028, that risk is too large for a 20% target. What would make this wrong. The terms of the Popular renewal in 2028 are the largest uncertainty. EVERTEC's main Puerto Rico tax grant also matters. It taxes most of its Puerto Rico income at 4% instead of the normal corporate rate of up to 37.5%. The grant expired in November 2024, and EVERTEC applied to renew it under the current incentives law. The company expects approval during 2026, retroactive to the expiration, but approval had not been reported as of mid-2026. Renewal is the normal outcome and we have assumed it, but if it were refused, the extra tax could cost a sum comparable to a large part of the Popular work. EVERTEC also keeps buying businesses with borrowed money, and its new $150 million buyback authorization could raise debt faster than earnings. The earnings of Tecnobank and Dimensa are our estimate, not a disclosed figure. Brazil's currency and economy, Puerto Rico's economy and possible minimum-tax changes also matter. We could not fully explain why operating cash flow in 2025 absorbed $56 million of working capital. Over ten years these swings have largely netted out, so we treat it as timing, but a lasting build-up in receivables would lower our starting level.