City Bank valuation report: fair value BDT 6.9–39.3 (Nov 2026)

City Bank in one page: the price assumes a 19% ROE indefinitely; we see 13%

Updated 3-Nov-2026: replaced the Q3 2026 figures from the DSE post with the Q3 2026 statements, added the Q3 income, cost, provision and tax changes behind the lower Q3 profit and the split of the Q3 rise in book value, and added the nine-month and Q3 figures to the model. Fair value and forecasts are unchanged.

Updated 2-Nov-2026: added the Q3 2026 results posted on the DSE (EPS and NAV only; the statements are not yet in our database); the business model and industry forces; the off-balance-sheet repo that funded the bond book; the three lending-rate regimes with City's loan yield and cost of deposits next to Bangladesh Bank and system rates; interest expense by deposit type; held-to-maturity and held-for-trading bonds; write-offs, interest suspense and special rescheduling; the reported operating cash flow; tax, capital instruments, ownership, ratings and five years of DSE events. Fair value, forecasts and prices are unchanged.

We put fair value for City Bank PLC (DSE: CITYBANK) at BDT 7–39 a share, base 18 (probability-weighted 20.7), against a price of 29.8 on 29-Oct-2026. The latest filing is the nine months to 30-Sep-2026 [20]: standalone Q3 2026 EPS was BDT 1.93 (Q3 2025: 2.31), nine-month EPS 4.89 (4.04) and NAV 40.99 a share [s20 p.9, p.14]. Model update note: our forecasts are built on H1 2026 and are not re-run on Q3; the Q3 figures are in the model's Half-yearly and 'Q3 2026' sheets.

City Bank earned BDT 13.1bn standalone in FY2025, up 20%, and ROE was 24.4% [s1 p.103, p.109]. The mix behind that profit changed completely in two years. Net interest income fell from BDT 14.7bn in FY2024 to 2.8bn in FY2025 while investment income rose from 16.6bn to 35.6bn [1 p.109]. City took the deposits it raised, added repo borrowing that cost BDT 14.2bn of interest in FY2025 [1 p.360], and bought government bonds at 2024-25's high yields. Net of the funding it uses, treasury contributed about BDT 22.2bn, 46% of FY2025 operating income, and lending about BDT 16.3bn. FY2025 also included BDT 6.1bn of gains on bonds [1 p.361]; without them ROE was 16.5%.

Our forecast keeps much of the bond income, because 153bn of the 280bn government portfolio is 10- to 20-year paper [2 p.12], but lets bills and shorter bonds roll off into today's lower yields. Net profit is BDT 10.2bn in FY2026F (EPS 5.85), then BDT 9.8bn and BDT 10.0bn; ROE falls from 15.6% to 12.0%. For the long run we anchor on City's record: its FY2021-25 ROE with bond gains stripped has a median of 16.5% [1 p.103]. Our forward normalised ROE (bond yield at the 10-year T-bond less 0.5pp, half of that fall passed to funding cost, through-the-cycle credit cost) is 10.1%. We use the average, 13.3%, below our 17.7% cost of equity; in our model value then comes out below book [8].

The price implies a sustainable ROE of about 19.1%: close to the FY2025 level with bond gains stripped, and above anything City earned before the 2024-25 rate spike. Asset quality is better than the sector: gross NPL was 2.50% in December 2025 and provisions exceeded the Bangladesh Bank requirement by BDT 3.4bn [s1 p.166, p.343]. The stock rose 78% in two years against 3% for the DSEX [3]. Uncertainty is High: two inputs nobody can observe, how fast bond income fades and how far funding costs follow, drive most of the value.

City Bank is a first-generation private bank with 3.3% of system deposits and Bangladesh's American Express franchise

City Bank is one of Bangladesh's first-generation private commercial banks. It runs 124 branches, 81 sub-branches, 247 SME units, 419 agent outlets, 549 ATMs and 6,219 staff [1 p.103], with conventional, Islamic window and offshore banking units. Corporate and SME lending is the bulk of the book, but the brand that sets City apart is retail cards: it is the sole American Express issuer and acquirer in Bangladesh [1 p.72], with 580,000+ American Express cardholders at the end of 2025, up from 506,000 in 2023 [1 p.74].

Business model, from the FY2025 filings. What it sells: loans (BDT 506bn at end-2025), deposit and payment accounts, trade services and cards [1 p.178]. How it earns: in FY2025 interest on loans was BDT 52.0bn, coupons on bills and bonds 29.3bn, gains on government securities 6.1bn, and commission, exchange and brokerage 9.1bn, against interest paid on deposits and borrowings of 51.9bn [s1 p.360–361]. Loan rates are set by the bank since Bangladesh Bank withdrew the SMART formula on 8-May-2024 [15 p.1]. To whom: industrial borrowers took 50% of loans, CMSME 18% and consumers (incl. housing and cards) 18% [1 p.178]; 37 borrowers or groups each owed more than 10% of the bank's capital, with BDT 133.9bn funded and 57.9bn non-funded outstanding [1 p.341]. Inputs: deposits of BDT 632bn, of which fixed deposits 360bn [2 p.7], and the branch, ATM and agent network above. Funding: deposits; borrowings of BDT 66.3bn, including IFC 6.1bn, the OPEC Fund 3.7bn and Bangladesh Bank refinance lines 29.7bn [s1 p.346–348]; repo with Bangladesh Bank and banks, BDT 138.3bn outstanding at end-2025 and off the balance sheet under Bangladesh Bank rules [s1 p.316, p.387]; bonds of 11.1bn; and equity of 61.1bn [s1 p.345, s2 p.7]. Control: directors and sponsors held 30.55% of the shares at end-2025, institutions 19.45% and foreign holders 5.44% [1 p.93]; no director holds more than 4.95% (IFC's nominee) [1 p.369].

Industry forces, as the annual report describes them. Rivalry: 'around 61 commercial banks operating over 10,000 branches', with competition on 'deposit and lending rates', digital platforms, SME lending and service [1 p.97]; City held 3.3% of system deposits in June 2026 [4]. Suppliers (funding): the report treats depositors as suppliers and says 'banks in Bangladesh generally have ample liquidity in the system, reducing supplier power' [1 p.98]. Buyers: 'customers in Bangladesh enjoy high switching power', and corporate clients 'negotiate strongly' in syndicated lending, trade finance and cash management [1 p.98]. Entry barriers: minimum capital, a Bangladesh Bank licence, compliance cost and the need for a deposit base [1 p.98]. Substitutes: mobile financial services, microfinance and NBFIs [1 p.98]. In its SWOT the bank lists 'margin sensitivity to interest rate cycles' and 'domestic concentration' as weaknesses [1 p.101].

At 30-Jun-2026 City held deposits of BDT 730bn and loans of BDT 548bn [2 p.7], about 3.3% of all scheduled-bank deposits and 3.0% of loans (BDT 2,210 and 1,809 thousand crore, June 2026) [4]. Industrial credit is half the loan book (BDT 254bn), CMSME 90bn, consumer loans 55bn, housing 20bn and credit cards 18bn [1 p.178]. The bank's 2.7mn customers made 34.4mn transactions on the Citytouch app in 2025, up from 21.7mn in 2023 [1 p.74]. The branch network has been widened mainly through sub-branches (81 in 2025 against 11 in 2021) and ATMs (549 against 338), while agent outlets were cut from 1,183 to 419 [1 p.103].

Subsidiaries are small and none is significant. City Brokerage (FY2025 profit BDT 63mn), City Bank Capital Resources (61mn), CBL Money Transfer in Malaysia (15mn), City Hong Kong (44mn) and the new City Credit Bureau together earned BDT 184mn, 1.4% of group profit of 13.2bn; the bank is 98.6% of group profit [s1 p.368, p.44]. None passes the 10% test on revenue, profit, assets or value [1 p.368]; we value the standalone bank, keep the subsidiaries at their carrying value inside its book and do not use a sum of the parts.

System deposits grew 10.7% and loans 4.3% in FY2025-26

Bangladesh's 61 scheduled banks held BDT 2,210 thousand crore of deposits in June 2026, up 10.7% in a year, while loans grew 4.3% to BDT 1,809 thousand crore [s4, Table-102]. City's annual report says that 'with private sector credit demand declining to multi-decade lows, surplus liquidity was increasingly deployed into sovereign' securities, and puts private credit growth at 6.10% in December 2025, 'the lowest in at least two decades' [1 p.122]. The weighted-average deposit rate eased only from 6.31% to 6.21% and the lending rate from 12.33% to 12.13% [4].

Overdue loans were 34.3% of outstanding loans in June 2026 (BDT 619,507 crore of 1,808,738 crore) [s4, Table-101]. Overdue is a wider measure than classified NPL and is not like for like with City's 2.50% [4]. Bangladesh Bank's banking-system classified-loan ratio (all scheduled banks, classified over outstanding loans, quarter-end) was 17.8% in September 2024, 31.2% in December 2025, 32.7% in March 2026 and 33.2% in June 2026 (Banking Sector Update, Table 4); BB's quarterly Financial Stability Assessment gives different figures for some quarters (35.73% against 36.3% for September 2025) [22]. It is a system figure on BB's basis and is not like for like with City's 2.50% at end-2025, taken from its own filings. City's deposits grew faster than the system's (22.8% in FY2025 [1 p.108]); the filings do not attribute that growth to transfers from other banks.

Rates are the second driver. The 10-year treasury bond cut off at 9.27% on 25-Oct-2026, the 91-day bill at 8.56% and the 364-day bill at 8.79%; overnight call money averaged 9.24% on 29-Oct-2026 [5]. Banks that bought 10- to 20-year bonds in 2024-25 at higher yields still earn those coupons; new money earns about 9%. Our forecast assumes policy and market rates drift lower through 2027-28; that path is our assumption, not a published Bangladesh Bank projection.

Deposits grew 15.6% in H1 2026; CASA fell to 40.9%

City's deposits rose from BDT 514bn to 632bn in FY2025 (+22.8%) and to BDT 730bn by June 2026, another 15.6% in six months [s1 p.108, s2 p.7]. Fixed deposits did most of the work: they rose from BDT 360bn to 423bn in H1 2026, while current accounts rose from 145bn to 164bn and savings from 119bn to 134bn [2 p.7]. The CASA share fell from 51.2% in 2022 to 43.0% in 2025 [1 p.102], and current plus savings was 40.9% of deposits in June 2026. Cost of deposits rose from 2.6% in 2021 to 5.5% in 2025 [1 p.103]. The Corporate and Institutional Liability Division reports deposits up 28% to BDT 55.4bn in 2025 [1 p.119], and deposits from other banks fell from BDT 14.1bn to 0.05bn [1 p.348]. Depositor concentration (top-20 depositors, government share) is not disclosed (searched note 15 and the Pillar 3 liquidity pages [s1 p.348–349, p.160–184]).

FY2025 interest expense of BDT 51.9bn included BDT 32.2bn on deposits and 14.2bn on repo borrowing, up from 5.4bn in FY2024 [1 p.360]: City borrowed against bonds to fund more bonds. With the repo cost included, our measured cost of funds was 8.0% in FY2025 against the bank's own 6.22% [s1 p.360, p.400]. In H1 2026 interest expense fell to an annualised 6.6% of funding as repo was run down. We forecast cost of funds of 6.5% for H2 2026, 6.3% in FY2027F and 6.1% in FY2028F.

The repo is off the balance sheet. Under Bangladesh Bank DOS circular letter 06/2010 a repo is 'accounted for as a normal sales transaction and the financial asset is derecognised in the seller's book'; only Assured Liquidity Support repo is booked as a loan [1 p.316]. City's repo with Bangladesh Bank averaged BDT 137.2bn a day in 2025 (61.5bn in 2024) and peaked at 238.0bn; repo with banks averaged 2.3bn (1.5bn) [s1 p.387; s13 p.463]. Dividing repo interest by those averages gives 10.2% in FY2025 and 8.6% in FY2024 (derived, sheet 'Repo and bond book' [8]). Thirteen repo agreements totalling BDT 138.3bn were outstanding at 31-Dec-2025, ten of them with Bangladesh Bank, maturing between 1-Jan-2026 and 28-Jun-2026 [1 p.387], while balance-sheet borrowings were 66.3bn [s1 p.346–348]. Our 8.0% cost of funds divides all interest expense, including repo interest, by deposits and on-balance-sheet borrowings; it is not a rate paid on any one liability. The H1 2026 interim gives no repo balance; it states that 'an increase in average repo borrowings contributed to the higher expense' [2 p.14]. The Q3 2026 balance is not yet in our database.

The bank's own cost measures. The five-year table prints 'cost of deposit based on average EOD balance' of 2.6%, 3.2%, 3.5%, 4.9% and 5.5% for 2021-25 [1 p.103], and the highlights print a 6.22% cost of fund for 2025 [1 p.400]. By type, FY2025 interest on fixed deposits was BDT 20.3bn (12.8bn in FY2024), Mudaraba term deposits 3.7bn (2.1bn), deposit schemes 3.2bn (2.1bn) and current deposits 3.5bn (2.8bn), while savings deposits cost 0.8bn (0.9bn) [1 p.360]. On borrowings, repo cost 14.2bn, local banks and NBFIs 1.7bn (0.2bn), Bangladesh Bank refinance 1.3bn and offshore borrowings 1.7bn [1 p.360].

Three lending-rate regimes frame these numbers. A 9% lending-rate cap applied from April 2020; from July 2023 the maximum rate was SMART (the six-month moving average of T-bill rates) plus a margin, under BRPD Circular No. 09 of 19-Jun-2023; and BRPD Circular No. 10 of 8-May-2024 withdrew SMART and left rates to the 'banker-customer relationship' [15 p.1]. The cap circular is not in our document set at 2-Nov-2026. City's printed yield on advances was 6.6% and 6.9% in 2021-22 under the cap, 8.0% in 2023, 9.9% in 2024 and 10.5% in 2025, and its cost of deposit 2.6%, 3.2%, 3.5%, 4.9% and 5.5%: a spread of 4.0pp, 3.7pp, 4.5pp, 5.0pp and 5.0pp [1 p.103]. The system's weighted-average lending rate went from 7.43% (June 2021) to 12.33% (June 2025) and its deposit rate from 4.14% to 6.31% [4]. The policy rate was 4.75% in December 2021 and 10.00% from December 2024 to June 2026 [16].

Key drivers of revenue, margin and funding, with their current levels: the policy rate (9.50%, cut from 10.00% by Bangladesh Bank's Monetary Policy Committee effective 30-Jul-2026 and held on 23-Sep-2026; 10.00% from October 2024 to 29-Jun-2026) [21]; T-bill and bond yields, which set the return on new bonds (10-year cut-off 9.27% on 25-Oct-2026 [5]; government bond yield 10.24% in June 2026 [16]); system lending and deposit rates (12.13% and 6.21%, June 2026) [4]; BDT/USD for exchange income (122.3 at end-2025, 123.0 at end-September 2026) [16]; system deposit and loan growth (10.7% and 4.3% in FY2025-26) [4]; system private-sector credit growth of +4.75% and bank deposit growth of +11.96% y/y in August 2026 (Bangladesh Bank Major Economic Indicators; series in model sheet 'BB credit-deposit growth') [23]; and CPI inflation (8.48% in December 2025, 8.34% in September 2026) [16]. A bank has no commodity input; history for each series is in the model sheets [8].

How far funding costs follow market yields is the key judgement. The evidence points both ways. On the way up (FY2023-25) the system deposit rate rose 1.28pp while the lending rate rose 4.99pp, a pass-through of about 26%; in the latest year both eased, deposits by 0.10pp against 0.20pp for loans, about 50% [4]. City's own cost of deposits rose 2.0pp while its loan yield rose 2.5pp over FY2023-25, about 80% [1 p.103]. We use 50% in the long-run normalisation and show 0%, 30% and 70% in the valuation section.

Balance-sheet growth: we model deposits +5% in H2 2026 (FY2026F +21%), then +13% and +12%, above the system's 10.7%: our assumption that recent inflows continue at a slower pace [8]. The AD ratio was 71.3% in December 2025 against an 87% cap [1 p.102], liquidity coverage 231.5% and NSFR 104.6% [1 p.400], so funding does not limit loan growth.

Loans grew 8.2% in H1 2026; industrial credit is half the book and cards 3.5%

Loans grew 13.8% in FY2025 to BDT 506bn and 8.2% in H1 2026 to BDT 548bn [s1 p.108, s2 p.7], three to four times the system's 4.3% annual pace [4]. By sector the December 2025 book was industrial credit 50%, CMSME 18%, consumer finance 11%, service sector 5%, housing 4%, credit cards 3.5% and agriculture 2% [1 p.178]. Bills purchased and discounted, mostly export bills payable abroad, were BDT 47bn [1 p.343]. 54% of the book matures within a year [1 p.178], so loan yields reprice quickly when rates move.

We forecast loans +4% in H2 2026 (FY2026F +13%), then +11% a year: above the system's 4.3% and below City's own FY2025 pace of 13.8%, our assumption given its AD ratio of 71.3% [1 p.102] and system loan growth [4] (model [8]). Loan yield follows rates down, from 10.5% annualised in H1 2026 to 10.0% by FY2028F; the system lending rate was 12.13% in June 2026 [4].

Retail is where City grows fastest by count: new personal loans rose from 40,336 to 52,604 and auto loans from 1,822 to 3,398 in FY2025, while new SME loans fell from 6,102 to 5,331 [1 p.228]. No loan of BDT 500 crore or more was restructured in 2025 [1 p.343]. 37 borrowers or groups each owed more than 10% of capital at end-2025 (31 a year earlier), with funded loans of BDT 133.9bn, 26% of the loan book, and no classified loans among them [1 p.341]; the bank's large-loan exposure was 37.9% in 2025 against 34.8% in 2024 [1 p.107]. Loans to companies in which directors have an interest were nil during 2025 [1 p.342]. Rescheduled-loan balances are not disclosed in AR2023-25 (searched the loans notes and Pillar 3 tables [s1 p.164–178, p.339–343; s13; s14]).

Net of funding, treasury earned 46% of FY2025 operating income; it fades as bonds roll off

Net interest margin, loan yield and cost of funds. City's reported NIM was 5.0% in FY2025 [1 p.103], but on our standard measure net interest income was 0.4% of average loans and investments [8]: the cost of all deposits and repo sits in interest expense while bond coupons sit in investment income [s1 p.360–361]. The meaningful margin is core NIM (net interest plus investment income over loans and investments): 6.0% in FY2024 and 6.0% in FY2025. Loan yield was 11.5% and cost of funds 8.0% in FY2025, a spread of 3.5%.

Interest on loans rose from BDT 41.8bn to 52.0bn in FY2025 (+24%) on loan growth of 13.8% [s1 p.360, p.108], while interest transferred to suspense rose from 4.2bn to 7.3bn and interest waived from 0.8bn to 1.5bn [1 p.353]. Our lending NII (net interest income plus the funding cost we charge to investments) fell from BDT 20.7bn to 16.3bn [8]; that figure depends on our allocation, and the filings do not split the change into loan yield, deposit repricing or suspense. For H1 2026 the bank attributes higher interest income to 'a higher average volume of loans and advances, and an improvement in yield on advances' [2 p.14].

Treasury on a like-for-like basis. Gross investment income was BDT 35.6bn in FY2025: 29.3bn of coupons on bills and bonds and 6.1bn of gains on government securities [1 p.361]. Charging the investment book its share of funding cost (cost of funds times average investments) leaves a net treasury contribution of BDT 22.2bn, 46.3% of operating income, against 26.7% in FY2024 and 7.3% in FY2023. In H1 2026 investment income was BDT 18.9bn, of which 16.5bn coupons and 1.7bn gains [2 p.13].

The bond book rolls off with maturity. In June 2026 the government portfolio was BDT 280bn: 10-year bonds 98bn, 15-year 32bn, 20-year 23bn, 5-year 61bn, 2-3-year 16bn, and bills and short Islamic bonds about 27bn [2 p.12]. HFT holdings (138bn) are marked to market and HTM holdings (84bn) are carried at cost [2 p.12]. The long bonds keep their coupons; bills and bonds of up to seven years (BDT 126bn, 45% of the book) reprice to today's 8.5-9.3% within our forecast. So the investment yield (on average year-end balances) falls from 14.9% in FY2026F to 11.1% in FY2027F and 10.3% in FY2028F, and we assume no further trading gains. Net treasury contribution falls to BDT 15.9bn by FY2028F.

Held to maturity and held for trading. At end-2024 City held BDT 57.3bn of government securities as HTM and 52.7bn as HFT; at end-2025 47.1bn HTM and 115.5bn HFT plus 27.0bn of sukuk and bonds; at June 2026 84.1bn HTM, 138.1bn HFT and 8.5bn encumbered [s1 p.336; s2 p.12]. HFT was 48% of HTM plus HFT at end-2024, 71% at end-2025 and 62% at June 2026 (derived). Gains on government securities were BDT 1.6bn in FY2024, 6.1bn in FY2025 and 1.7bn in H1 2026 (2.7bn in H1 2025) [s1 p.361; s2 p.13]; the reduced tax rate on capital gains lowered FY2025 tax by BDT 1.2bn [1 p.365]. The filings do not split gains into sales and revaluation, nor give purchase yields or coupons by bond. BDT 110.0bn of government securities was counted as cash equivalents at end-2025 (61.9bn a year earlier) [1 p.312]. The revaluation reserve on securities rose from BDT 5.5bn to 6.6bn in H1 2026 and 10.3bn at September [s2 p.7; s20 p.7]. At 30-Sep-2026 HTM was 84.1bn and HFT 183.3bn [20 p.12].

By tenor (standalone), 10-year bonds were BDT 16.0bn at end-2023, 42.5bn at end-2024, 63.0bn at end-2025 and 98.1bn at June 2026; 20-year bonds 0.5bn, 10.7bn, 15.0bn and 22.6bn; 15-year bonds 11.8bn, 27.0bn, 23.1bn and 32.4bn; 5-year bonds 19.6bn, 28.0bn (end-2024 as restated in AR2025; AR2024 printed 33.8bn), 49.4bn and 60.5bn [s13 p.413; s1 p.337; s2 p.12]. Purchase yields and coupons by bond are not disclosed.

Cards, retail and fee income. Commission, exchange and brokerage was BDT 9.1bn in FY2025 (+11.3%) [1 p.361]. Credit card income was BDT 2.49bn, up 14.6%: card issue fees 970mn (+24%), merchant commission 562mn, late fees 341mn, interchange 250mn (+28%) [1 p.361]. The American Express model earns on both sides: as issuer City collects annual fees, interchange and revolving interest from 580,000+ cardholders; as the only acquirer of American Express in the country it earns the merchant discount on every Amex purchase at its 38,000 merchant points [1 p.130]. City's acquiring handled BDT 88bn of volume (+15%), a 23% share of acquiring and 24% of POS terminals, plus 33,000 Bangla QR points; in 2025 it launched Bangladesh's first metal card and Google Pay integration [1 p.130]. Card loans of BDT 17.7bn [1 p.178] add interest income on top of the fees.

We forecast card income +14% and +13% in FY2027-28F (FY2025 +14.6%), other commission (trade, LC, guarantees, service charges; BDT 5.0bn) +9% with trade volume (imports BDT 482bn, exports 420bn in FY2025 [1 p.103]), and exchange gains +5% after H1 2026 (BDT 1.04bn, +83% y/y [2 p.13]). Fees rise to BDT 12.6bn in FY2028F, 22.6% of operating income.

Costs. Operating expenses rose 25.2% in FY2025 and 23.7% y/y in H1 2026 [s1 p.109, s2 p.9]. Staff cost, 58% of the total, rose 35% in FY2025 with 898 more staff and the year-end salary adjustment [s1 p.103, p.227, p.360]. In its latest pay review the bank 'restructured salaries of all grades' and introduced a critical-illness benefit [1 p.185]; the filings do not split the rise into one-off and recurring parts. We model staff cost +10% a year and other costs +10% after H2 2026; in our model cost-to-income rises from 43.7% to 55.0% [8].

NPL fell to 2.50% with BDT 3.4bn excess provisions; a 50% haircut on net NPL cuts book by BDT 5.6bn

Gross NPL fell from BDT 16.6bn (3.72%) in 2024 to 12.7bn (2.50%) in 2025: sub-standard 0.35%, doubtful 0.48%, bad/loss 1.67% [s1 p.166, p.400]. The fall came largely from write-offs: BDT 43.5bn of loans were classified during the year and 47.4bn left NPL through recovery, rescheduling and BDT 10.1bn of write-offs [1 p.178]. Special-mention loans, the next in line, were BDT 8.3bn, two-thirds of the NPL stock [1 p.178].

Two write-off figures. The loans note shows BDT 13.37bn of loans written off in 2025 (4.43bn in 2024) [1 p.342]; the Pillar 3 movement table shows BDT 10.13bn of fully provided loans written off (3.50bn) [s1 p.178; s13 p.247]. The difference, 3.24bn, is within BDT 0.03bn of the 3.22bn of interest suspense written off in 2025 [1 p.353]; the filings do not reconcile the two tables. Reductions in NPL were BDT 9.8bn in 2023, 16.6bn in 2024 and 47.4bn in 2025, and additions 10.6bn, 18.7bn and 43.5bn [s14 p.284; s13 p.247; s1 p.178]. Cumulative written-off loans were BDT 36.6bn at end-2025, and loans under law suit 59.9bn [1 p.342].

Special rescheduling. The auditor's key audit matter says Bangladesh Bank gave 'policy support to borrowers who become classified/non-performing through BRPD Circular No. 07 dated 16 September 2025', and that 'Significant proportion of the Bank's classified borrowers availed these special rescheduling and policy support facilities and were consequently reported as unclassified as at the reporting date. However, despite being unclassified these borrowers reflect high risk of default and hence may require additional provision.' [s1 p.297–298]. The rescheduled balances are not disclosed (searched the loans notes and Pillar 3 tables [s1 p.164–178, p.339–343]). Interest suspense on standard loans rose from BDT 5.5bn to 7.2bn, while suspense on classified loans fell from 3.4bn to 1.2bn [1 p.353]. The bank had identified 15 wilful defaulters by end-2025, 7 of them in 2025 [1 p.376].

What the FY2025 provisions were for. Of BDT 8.13bn, 3.73bn was for unclassified loans (2.11bn in FY2024), 3.91bn for classified loans (2.51bn), -0.22bn for the value of investments (+0.52bn) and 0.72bn for banks, financial institutions and other assets (0.12bn) [1 p.364]. General provision held rose from BDT 8.8bn to 10.0bn, BDT 3.4bn above the requirement [1 p.342].

By sector, NPL was concentrated in industrial credit (BDT 4.1bn), CMSME (3.5bn), consumer finance (1.6bn) and services (1.5bn) [1 p.178]. CMSME NPL of 3.5bn on a 90bn book (3.9%) is the highest ratio among the large segments [1 p.178]. Provisions held were BDT 14.4bn against 11.0bn required, an excess of 3.4bn [1 p.343], and the bank reports provision coverage of 127% [1 p.74]. Specific provision covered 35% of gross NPL; the requirement is set after eligible security [1 p.343].

In H1 2026 City charged BDT 4.6bn of provisions, of which 3.1bn general provision [2 p.13], 1.76% of loans annualised. The interim does not disclose NPL; we assume gross NPL of 3.0% at end-2026 and 3.2% after, under the 2025-26 rules (the auditor cites BRPD circulars 04/2024 and 08/2025 for write-offs and 15/2024 for classification and provisioning [s1 p.297–298]). Credit cost: 1.6% in H2 2026, 1.4% in FY2027F and 1.3% in FY2028F, against 1.71% in FY2025 and a FY2022-25 average of 1.16%. If NPL rises above our path, the model adds provisions at the 35% specific-coverage rate on the increase. For adjusted book we deduct 50% of forecast net NPL (gross NPL less specific provision), BDT 5.6bn; there is no provision shortfall to deduct.

CRAR of 16.3% is ample, but cash dividends need Tk2,000 crore of paid-up capital

Standalone CRAR was 16.26% in June 2026 and Tier-1 11.10%, against a 12.5% minimum including the conservation buffer [2 p.13]. Tier-1 includes a BDT 4.0bn perpetual bond; Tier-2 rose to 28.6bn after a new 13.7bn subordinated bond [2 p.12]. Capital above 12.5% was BDT 15.7bn at end-2025 [1 p.400]; on our forecast it is BDT 19.7bn in FY2027F and BDT 15.8bn in FY2028F as Tier-2 bonds amortise. RWA density fell from 73% in 2021 to 56% in June 2026 because government bonds carry no credit-risk weight [s1 p.102, s2 p.13].

Tier-2 bonds: the 4th subordinated bond (BDT 7.0bn) is held by Agrani Bank 2.6bn, Janata Bank 2.0bn, Sonali Bank 1.0bn, Mercantile Bank 0.65bn, National Life Insurance 0.5bn and Simanto Bank 0.25bn, and BDT 80mn of the 3rd bond remains with Dhaka Bank [1 p.346]; note 13.1 and the Pillar 3 capital pages print no coupon, tenor or redemption schedule for the 3rd or 4th bonds [s1 p.346, p.173]. The perpetual bond pays the 20-year T-bond standard tenor yield plus 2%, within 6-10%, out of retained earnings, and converts to equity only if consolidated CET1 stays below 4.5% for three quarters [1 p.346]. BSEC consented to a Tk1,200 crore unsecured Tier-II bond (12,000 bonds of Tk10 lakh) on 12-Apr-2026 [19]; bonds on the balance sheet rose from BDT 11.1bn to 20.5bn in H1 2026 [2 p.7]. Coupon, tenor and investors of the new bond are not in the interim.

The dividend rule. Bangladesh Bank issued a circular on 23-May-2026 under which banks with paid-up capital below Tk2,000 crore may not declare cash dividends, starting with dividends for the year ending 31-Dec-2026; banks above the line may pay at most half of a declared dividend in cash. Both the rule and its timing are reported by TBS (23-May-2026) and the Daily Star (24-May-2026) [9][10]; we have not read the circular itself. City's paid-up capital is Tk1,749 crore [2 p.7]. Its board proposed raising authorised capital from Tk2,000 crore to Tk3,000 crore at an EGM on 4-Oct-2026 [12][6][7], but on 29-Sep-2026 cancelled the EGM after new Bangladesh Bank directives on company shareholdings in banks and board representatives; the bank says the proposal is not abandoned [11], and as of 30-Oct-2026 DSE news showed no new EGM called.

The board's 13-Aug-2026 proposals also included raising the number of non-independent directors from 8 to 10; it withdrew that item on 24-Aug-2026 'following further consideration of the legal, regulatory, governance, shareholder, contractual and other consequential aspects' [19]. The cancellation notice names 'fresh directives concerning companies' shareholdings in banks and the appointment of company representatives on the Board' without saying which holders they affect [11]. Three directors are company nominees: IFC (4.95%), A-One Polymer (2.00%) and Partex Corporate (2.00%) [1 p.369]. Directors and sponsors held 30.55% at end-2025 [1 p.93], and each non-independent director held between 2.00% and 4.95% [1 p.369]. The filings do not state compliance with the 30% collective / 2% individual holding rule. On 15-Oct-2026 sponsor Mohammed Shoeb declared an intention to sell 3,516,029 of his 45,775,541 shares by 29-Oct-2026, and on 26-Oct-2026 the company reported 4,219,235 bonus shares of the late sponsor Deen Mohammad (holding 73,937,078 shares) being transmitted to him [19]. Share pledges are not disclosed.

Two routes to Tk2,000 crore. City needs 250.6mn new shares. Authorised capital of Tk2,000 crore is exactly 2,000mn shares [12], so either route fills it to the last share and leaves no room for later stock dividends until authorised capital is raised; that is what the higher Tk3,000 crore ceiling would enable, not the threshold itself. A stock dividend of 14.3% raises no cash: each holder owns 1.143 shares for every one, value per holding and ROE are unchanged, and value per post-bonus share is BDT 16.0. A rights issue at an assumed Tk10 par (no offer has been announced; par is the lowest price allowed and our assumption) would raise BDT 2.5bn. Invested at the 10-year T-bond after tax, it lowers FY2027F ROE from 13.1% to 12.9%, and with the new shares priced below our 18.3 base value per share, value per share falls 5.7% to BDT 17.2 in our model [8]; a holder bears that loss only by neither subscribing nor selling the rights. Paid-up capital is above the threshold either way; the stock dividend costs existing holders nothing in cash, the rights issue adds about half a point of CRAR. The dividend rule, not capital adequacy, is the reason to act.

The timing of a capital increase is not known [11]; we treat the FY2026 dividend as a scenario. Base: no FY2026 cash dividend (a stock dividend lifts capital above the line), then cash payout of 20% in FY2027F and 22% in FY2028F (DPS 1.12 and 1.25 on current shares). Bear: no cash until FY2028F. Bull: capital is raised before the FY2026 declaration and 17% is paid in cash. Retained profit stays in book: in our model the rule moves dividend yield, not residual-income value [8]. Cash payouts from FY2027F also need a new EGM to raise authorised capital: the rule caps cash at 50% of a declared dividend [9][10], and each cash payout then needs an equal stock dividend, and the 250.6mn shares that reach Tk2,000 crore already fill the Tk2,000 crore authorised capital. This affects timing of dividends, not fair value.

Cash interest receipts covered 100% of FY2025 loan interest; no regulatory action on the board

Interest received in cash was BDT 55.0bn in FY2025 against interest income of 54.7bn (100%) [1 p.307], so booked loan interest is being collected. City is under no Bangladesh Bank board reconstitution or provision deferral. Financial statements follow Bangladesh Bank rules where they differ from IFRS: perpetual bond coupons are charged to retained earnings, and since an October 2025 BB directive listed shares are carried at cost [s2 p.13–14].

Reported operating cash flow. Net cash from operating activities was BDT 13.7bn in FY2022 and 3.4bn in FY2023, when customer deposits brought in 49.8bn and 59.8bn, loans used 68.4bn and 41.3bn, bank borrowings added 27.5bn and then took out 22.1bn, and net securities purchases in investing were 11.5bn and 10.0bn [14 p.402]. It was BDT 89.0bn in FY2024 and 70.5bn in FY2025: customer deposits brought in 112.0bn and 131.4bn, loans used 48.9bn and 61.2bn, and borrowings from banks added 8.8bn and then took out 6.3bn [1 p.312]. Net purchases of securities in investing were BDT 13.4bn in FY2025 [1 p.312], while government securities counted as cash equivalents rose from 7.5bn (end-2022) and 7.8bn (end-2023) to 61.9bn and 110.0bn [s14 p.402; s1 p.312]: part of the bond buying sits inside cash and cash equivalents, not investing. In H1 2026 operating cash flow was BDT 78.0bn, with customer deposits 97.9bn and bank borrowings 18.5bn [s2 p.11, p.14]; for 9M 2026 operating cash flow was BDT 127.9bn, with customer deposits 142.0bn, bank borrowings 8.0bn and loans -38.9bn, while net securities purchases in investing were 60.2bn and government securities in cash equivalents rose to 191.5bn [20 p.11]; the company attributes the rise 'mainly' to 'higher deposit acquisition from customers' [20 p.14]. The model's balance-sheet cash flow treats all investments as investing and differs from the reported figures [8].

Other assets rose from BDT 31.9bn to 43.5bn in FY2025: advance tax 15.6bn (7.1bn), interest receivable 10.2bn, of which 5.7bn on government securities, card-operation receivables 6.4bn (4.9bn), remittance advances 2.9bn (1.8bn) and investments in subsidiaries 6.3bn [s1 p.344–345]. Non-banking assets from 34 court-awarded properties were BDT 0.45bn [1 p.345]. Subsidiaries and projects: City put BDT 195mn into City Credit Bureau PLC in 2025 [1 p.370]. The 2021 board decision to set up a Singapore payments subsidiary (SGD 2mn) and the 2023 decision to invest up to BDT 277.8mn in a digital-bank consortium [17] do not appear in the FY2025 subsidiary or investment notes [s1 p.344, p.336]; their status is not disclosed. Related parties: IFC, whose nominee is a director, lent BDT 6.1bn [1 p.370].

CRAB rated City AA1 in June 2024, upgraded it to AAA in May 2025 and kept AAA / ST-1 with a stable outlook in June 2026 [18][19]; Moody's rates it B2 with a negative outlook, 'aligned with Bangladesh's sovereign rating' [1 p.115]. CRAB's rationale reports are not in our document set. The auditor's key audit matters are loan provisioning (including the special rescheduling above) and interest-income recognition [s1 p.297–298]. Without modifying its opinion, the auditor draws attention to note 2.1 (basis of preparation, including Bangladesh Bank rules on loan classification and impairment provision) and note 2.14 (presentation of comparative information) [1 p.297]. Not checked: share pledges and DSE category. The Islamic window earned BDT 3.1bn of pre-tax profit in FY2025 and the offshore unit 0.7bn [1 p.368].

We forecast EPS of BDT 5.85 in FY2026F, in line with the H1 2026 run rate of 5.93

FY2026F is H1 2026 actual (net profit BDT 5.18bn, EPS 2.96 [2 p.9]) plus our H2 forecast from the June balance sheet: net profit BDT 10.2bn, EPS 5.85, down 22% from FY2025. FY2025 benefited from 6.1bn of bond gains, a low 30.6% tax rate and H2 provisions of only 1.7bn [s1 p.109, p.361]; H1 2026 tax was 45.8% [2 p.9] and we assume 40% in H2. We then forecast BDT 9.8bn in FY2027F (EPS 5.58) and BDT 10.0bn in FY2028F (EPS 5.69); ROE eases from 15.6% to 12.0% as bond income rolls off faster than fees and loans grow. Our FY2026F EPS compares with an H1-annualised run rate of 5.93 and a five-year average (FY2021-25, restated) of 4.49.

Q3 2026, from the statements. Standalone profit after tax was BDT 3.38bn in July-September 2026 against 4.04bn a year earlier (EPS 1.93 against 2.31) [20 p.9]. Operating income rose by 1.30bn to 14.57bn: net interest income was 0.97bn against -0.22bn, investment income 10.82bn against 11.11bn, and commission, exchange and brokerage 2.68bn against 2.31bn [20 p.9]. Gains on government securities were about 2.11bn in Q3 2026 against 2.82bn in Q3 2025 (nine-month less half-year figures, derived [s20 p.13; s2 p.13]). Against that, operating expenses rose by 1.01bn to 6.37bn (salaries 3.46bn against 3.08bn), provisions by 0.42bn to 2.83bn, with 0.04bn charged for the value of investments against a 0.65bn release a year earlier, and tax by 0.53bn to 2.00bn, including 0.09bn of deferred tax against a 0.11bn credit [20 p.9]. For nine months the company attributes higher profit 'mainly' to 'incremental commission and fess income as well as income from interest income / profit on investments and investment income' [20 p.14]; it gives no reason for the lower Q3. Book value: standalone equity rose from BDT 64.92bn at June to 71.71bn at September, +6.79bn: Q3 profit 3.38bn, the revaluation reserve on government securities +3.61bn (6.65bn to 10.26bn), the perpetual-bond coupon -0.20bn and other items +0.002bn [s20 p.7, p.9; s2 p.7] (derived). NAV was 40.99 a share standalone and 41.51 consolidated [20 p.14]. Our FY2026F EPS of 5.85 less nine-month standalone EPS of 4.89 leaves 0.96 for Q4 2026, against 3.42 in Q4 2025 (FY2025 standalone 7.46 on current shares less 9M 2025 4.04, derived; sheet 'Q3 2026' [8]). We have not re-run the forecast.

Tax. The FY2025 effective rate of 30.6% (39.0% in FY2024) reflects a 37.5% statutory rate, admissible expenses of -25.1pp, inadmissible items of +21.7pp, a 6.6pp saving from the reduced rate on capital gains, deferred tax of +5.8pp and a 2.4pp prior-year release [1 p.365]. In H1 2026 deferred tax was BDT 0.87bn against 0.03bn a year earlier, and current tax 3.51bn [2 p.13]; the interim gives no reconciliation, and attributes higher current tax to 'higher operating profit' [2 p.14].

Fair value BDT 7–39: base 18 = 0.49x book for a 13.3% sustainable ROE

Method. Residual income on adjusted book is 70% of the blend, the peer median P/B applied to City's adjusted book 15%, and P/E on normalised EPS 15%. Sum of the parts does not apply: subsidiaries are 1.4% of group profit. Cost of equity is 17.7%: the 10-year T-bond cut-off of 9.27% [5] plus a Blume-adjusted beta of 1.21 (raw 1.31, two years of weekly returns against the DSEX [3]) times our standard 7% equity risk premium. Terminal growth is 7%. Range convention: bear and bull are full recalculations of the model with each scenario's drivers, not a percentage band around the base.

Why 70/15/15 rather than our default 60/25/15. The default gives 25% to a peer P/B-ROE regression. Across City's eight listed peers that fit has an R² of 0.09: P/B barely moves with ROE in today's DSE bank market, and the fitted multiple mostly restates the peer average [3]. We therefore replace it with the plainer peer median P/B, applied to adjusted book, and cut its weight to 15%; the 10 points go to residual income, the only method that reads the forecast and the normalised ROE directly. At the default 60/25/15, with the peer median in the 25% slot, the base would be BDT 18.8.

Adjusted book. Equity at end-2025 was BDT 61.1bn. We deduct the 5.5bn revaluation reserve on bonds (the above-market coupons it reflects are already in our forecast) and 50% of forecast net NPL (BDT 5.6bn), giving adjusted book of BDT 50.0bn. At June 2026 the same adjustments give BDT 30.1 a share against reported NAV of 37.11 [2 p.14]; the stock trades at 0.99x adjusted book.

Sustainable ROE. Our bank method anchors on the five-year record. City's reported ROE for FY2021-25 was 15.8%, 14.1%, 17.6%, 26.0% and 24.4% [1 p.103], median 17.6%. FY2024-25 were lifted by bond gains of BDT 1.6bn and 6.1bn [1 p.361]; stripping them after tax gives 23.6% and 16.5%, and the median becomes 16.5%. Credit cost over those years averaged about 1.2%, close to the 1.3% through-the-cycle rate we use; we make no further adjustment [8]. The forward view is lower: FY2028F with the bond yield reset to 8.77% (10-year less 0.5pp), half that fall passed to funding cost, and the FY2022-28F average credit cost gives 10.1%. We weight the two equally: 13.3%. The record includes 2021-23, when City earned 14-18% without the bond windfall but also with lower deposit costs (2.6-3.5% [1 p.103]); the forward view carries today's 6%+ funding cost.

Residual income. FY2026-28F profit after the perpetual coupon falls short of the 17.7% charge on adjusted book: explicit residual income is BDT -1.5bn [8]. The 13.3% sustainable ROE justifies 0.59x adjusted book in the terminal year. Value at end-2025 is BDT 29.4bn: adjusted book BDT 50.0bn, plus explicit residual income, less BDT 19.1bn of terminal value destruction: in the residual-income method a bank that earns below its cost of equity is worth less than its book [8]. Rolled to 30-Oct-2026 and net of the FY2025 dividend paid, that is BDT 18.0 a share.

Cross-checks. The eight listed peers trade at a median 0.77x book [3]; applied to City's adjusted book of 30.1 that is BDT 23.0. A regression of peer P/B on ROE has an R² of 0.09 on 8 points [3]; we show it on the Peers sheet as context and do not value on it. P/E: City's year-end P/E was 2.8-6.1x over FY2021-25, median 4.3x [1 p.103]; on normalised EPS of 3.46 that gives BDT 14.9. The three legs agree within about BDT 8.

Departures from our default and the pass-through test. We set the long-run bond yield 0.5pp below the 10-year T-bond, not the default 1pp: half of City's government book is 10- to 20-year paper bought above today's yields [2 p.12]; on the default the base would be BDT 17.4. With funding pass-through of 0%, 30% or 70% instead of 50%, the base becomes BDT 15.5, 17.1 or 19.4; the residual-income leg stays positive in all three.

Scenarios change drivers, not outputs. Bear (6.9): deposits and loans grow 8-9%, bond yield 9.9% by FY2028F, credit cost 1.5-1.8%, NPL rising to 5.0% with a 100% haircut, cards and fees slower, Ke +1pp and the default 1pp treasury spread. Bull (39.3): bond yield 11.0%, cost of funds 5.9%, credit cost 1.0%, card income +17-18%, Ke -1pp and no treasury discount to the T-bond. Probability-weighted (25/50/25) 20.7. At 18.3 the base implies 0.49x June 2026 book and 3.1x FY2026F EPS. The price of 29.8 implies a sustainable ROE of 19.1%.

Six risks could cut fair value by up to 61%; sticky funding costs matter most

Every risk below changes one driver in the model, with everything else at base, ranked by effect on fair value. Funding costs come first: City funds BDT 860bn+ of assets with deposits and borrowings against equity of about 80bn; in our model 1pp on cost of funds cuts FY2027F EPS to 2.36 and fair value by 61% [8]. Deposit growth at half the system's pace leaves less money to invest (fair value -28%). Credit: gross NPL of 6-8% adds provisions and a bigger book haircut (FY2027F EPS 3.84, value -28%); a 2.5% credit cost on its own cuts value 17%.

Treasury yields are the most likely risk: a 2pp lower investment yield in FY2027-28 cuts FY2027F EPS to 3.24 and value 25%, smaller than its EPS effect: our sustainable ROE already assumes yields normalise [8]. Fees: flat card and commission income takes 11%. Regulation: the Tk2,000 crore paid-up rule delays cash dividends (yield, not value). Opportunities: card and payments growth, a faster fall in deposit rates than we assume (70% pass-through adds BDT 1.1), and continued share gains in deposits.

Five signposts to April 2027

• Investment income in the FY2026 accounts: above BDT 38bn without trading gains supports the bull case; below 35bn the bear.
• Cost of funds: H2 2026 below 6.5% annualised means pass-through is faster than our 50%.
• Gross NPL at end-2026: our base is 3.0%; above 4.5% moves us toward the bear case.
• Fee income: card income growth above 15% in FY2026.
• Paid-up capital above Tk2,000 crore before the FY2026 dividend declaration.