$37.44
$75.23
bear $37.44
base $65.51
bull $75.23
margin of safety $32.00
market price* $63.36
Citizens Financial Group is one of the larger US regional banks: about $234 billion of assets at mid-2026, $186 billion of deposits and $143 billion of loans (year-end 2025), served through more than 1,000 branches in 14 states and the District of Columbia, plus a growing private bank and private wealth business. Because it is a lender, deposits and other funding are the raw material of the business and interest is an operating cost, so we measure owner earnings as net income available to common shareholders per diluted share and do not net any debt against the value. On that basis owner earnings per share were $1.97 (2016), $3.25 (2017), $3.52 (2018), $3.81 (2019), $2.22 (2020, pandemic reserve build), $5.16 (2021, flattered by reserve releases and a mortgage boom), $4.10 (2022), $3.13 (2023), $3.03 (2024) and $3.86 (2025). For 2017 through 2025 the median is $3.52; years that mix a deposit-cost squeeze (2023-24) with a reserve build and a reserve release (2020, 2021) are all inside that window on purpose, so it contains a real cycle. Operating cash flow was not used as the starting point because bank cash flow swings with loan sales and trading flows (for example $111 million in 2020 and $4.1 billion in 2022); in 2024 and 2025 net income plus non-cash charges tied to operating cash flow within about 1 percent. What has changed since our earlier analysis: the first half of 2026 confirmed the recovery. Net interest income was $3.19 billion, up 13 percent from $2.83 billion a year earlier, fee income was up 10 percent to $1.26 billion, and expenses rose only 5 percent to $2.77 billion. Net income to common holders was $1.04 billion versus $0.74 billion in the first half of 2025, or about $2.42 per share. Counting the second half of 2025, that is four consecutive quarters of at least about $1.08 per share, with the latest, $1.30, the highest. The share count is down to about 427 million. Our earlier figures (roughly $28 bear, $51 base, $68 bull) were built on a different basis and are not comparable with the figures below. Base case, $65.51 per share. We treat $3.86, the 2025 result, as the through-cycle level, because four corroborating quarters sit at or above it. 2025 owner earnings grew 27 percent over 2024; for year 1 we assume $4.30, about 11 percent above 2025, below that trailing rate and equal to the weakest of the last four quarters annualized. The dated reasons are the year-over-year net interest income gains in the first two quarters of 2026 and expenses growing at less than half the pace of revenue. We hold $4.30 for three years and then fall back to $3.86 permanently, which stands in for the cycles every bank goes through (earnings fell 42 percent in 2020 and 41 percent from 2021 to 2024). We credit no growth in loans, share gains or private bank build-out: loans have not grown on a sustained basis (average loans fell in 2024) and the private bank is new. Buybacks are not credited separately because the full earnings stream is already treated as distributable. Business worth: $65.51; owes: nothing is subtracted (deposits fund the loans); difference: $65.51. Bear case, $37.44. This is a full credit cycle of the 2008-2011 kind, which cost US banks cumulative loan losses of about 7 to 8 percent of loans over four years, with annual charge-off rates of roughly 2.5 to 3 percent at the 2009-2010 peak. We build it as provisions of $2.6 billion, $4.0 billion, $3.4 billion and $1.6 billion over four years (about $11.6 billion, or 8 percent of $143 billion of loans, versus a 2025 charge-off bill of $684 million) against pre-provision profit of $2.7 billion a year (between the 2024 level of $2.6 billion and the 2025 level of $2.9 billion). After tax and preferred dividends that is owner earnings of about minus $0.14, minus $2.73, minus $1.62 and plus $1.71 per share, a cumulative loss of about $5 billion in the three worst years against $24 billion of common equity, so we model no capital raise, but we do not credit dividends or buybacks through the downturn. Afterwards earnings settle at $3.03 a share, the actual 2024 trough, permanently, because deposit pricing and margins stay as squeezed as they were then. The bear value is about the same as tangible book value of $38.07 per share at year-end 2025. We think a franchise that earned a 9 to 10 percent return on tangible equity even in its worst recent year, with $186 billion of deposits and 1,000 branches, is worth about its capital even after a deep cycle; if it earned much less than that for years, it would not be. Business worth: $37.44; owes: nothing subtracted; difference: $37.44. Bull case, $75.23. The business's own levers execute: first-half 2026 earnings annualize to about $4.87 per share, we credit $4.85 for five years (net interest income and fee growth continue while costs rise about half as fast) and then $4.40 permanently to allow for cycles. Business worth $75.23; owes nothing; difference $75.23. We tried to break the case with specific facts. Commercial real estate is the sore spot: general office loans are only about $2.4 billion, roughly 96 percent of commercial real estate loans were current at year-end 2025, and nonaccrual loans fell during 2025. Even a 25 percent loss on all general office loans would cost about $0.6 billion, or $1.40 per share, once. The second challenge is whether first-half strength is a one-off: it was driven by net interest income up 13 percent in both the first and second quarters, which is not a single quarter print, and the bank's own risk disclosure shows it relies on receive-fixed swaps to hedge an asset-sensitive balance sheet, so falling short-term rates would dent earnings, which the bear case covers. Long-term borrowings rose to $15.2 billion at mid-2026 from $11.2 billion at year-end 2025; these fund the bank and are not subtracted. We could not reconcile the rise in borrowings to any single event from the filings and note it as unexplained. Margin-of-safety price, $32.00. We are fairly sure the earnings level is at least $3.86 a share because it is supported by four quarters of higher earnings, but we are less sure about the next cycle, which is not visible in any filing and which this bank has gone through twice in nine years in some form. The worst plausible outcome is a credit cycle that is worse than 2008-2011 in commercial real estate and commercial and industrial loans together, or a sharp fall in rates that compresses the margin at the same time; the bank has $24 billion of common equity, $2.1 billion of preferred stock ahead of it, and 10 to 11 percent regulatory capital, so it would survive, but shareholders would see several years of lost earnings and a dividend cut. If we are wrong on the margin and the bank earns $3.03 a share permanently, or credit costs run 1.5 times our bear case, the shares are worth about $30 to $37. A price of $32 is roughly 84 percent of year-end 2025 tangible book value, a price at which the bank's own hard capital covers the shares even in the cycle in our bear case, with a dividend of $1.84 a year (about 5.8 percent) plus about $1.4 billion a year of buybacks over the holding period. The price assumes a holding period of three years or more for the earnings to be reflected. From that price, the base case implies about $5.50 of dividends and a business earning $4.30 a share with tangible book value that keeps growing, which together imply about 20 percent a year; the bear case implies roughly breaking even.