Dutch-Bangla Bank valuation report: fair value BDT 11.1–105 (Oct 2026)

Fair value BDT 11–105, base 53: the price of 46.6 assumes a 14% sustainable ROE against our 15%

We put Dutch-Bangla Bank's (DBBL) fair value at BDT 11–105, base 53 (probability-weighted 55.6), against a price of 46.6 on 29-Oct-2026. Net profit doubled to BDT 9.65bn in FY2025 (EPS 9.98) after a provision-heavy 2024 [s1 p.149, p.171], and the first half of 2026 earned BDT 4.42bn against 1.05bn a year earlier (EPS 4.35 against 1.04) [2 p.4]. The DSE release for the nine months to September shows EPS of 6.15 against 2.52 [5].

DBBL's edge is funding. Current, savings and bills-payable balances were 80% of its BDT 622bn deposits at December 2025 and 79.5% at June 2026 [2 p.2], gathered through 243 branches, 334 sub-branches, 4,151 ATMs and 4,887 cash-recycling machines [1 p.26]. Interest on deposits cost 2.6% of average deposits in 2025 [1 p.222] while the banking system paid 6.21% [9]. The bank lends only 63% of its deposits at June 2026 and parks the rest in government securities yielding 8.6–9.3% [s2 p.2; s8], so investment income (BDT 22.0bn) passed net interest income (19.9bn) in 2025 [1 p.149].

Three things hold value back. Gross NPL jumped from 6.43% to 9.18% of loans in six months and special-mention loans reached BDT 45.1bn [2 p.26]. Funding cost is creeping up (3.41% of deposits and borrowings in 2025 against 2.39% in 2024), and on a BDT 740bn funding base each 1pp costs about BDT 7bn of pre-tax profit: our model puts fair value at 14 with 1pp higher funding cost. And the new Bangladesh Bank rule bars any cash dividend while paid-up capital is below Tk2,000 crore [6 p.1]; DBBL's is Tk1,015 crore, so it needs 97% more shares before cash can return.

Our base case assumes a sustainable ROE of 15.4% against a cost of equity of 17.1%, close to DBBL's own five-year median ROE of 16.1% (FY2021–25: 16.1, 14.4, 17.9, 9.5 and 17.3%) [1 p.26]. That justifies 0.83x adjusted book; the base value of 53.2 is 0.87x the June-2026 book of BDT 61.15 a share. The price implies a sustainable ROE of 14.3%. The range is wide on purpose: on a BDT 900bn balance sheet and BDT 62bn of equity, a 0.7pp move in funding cost and in bond yields together swings ROE by about 8pp, and the bear and bull cases apply those moves symmetrically.

DBBL turns 80% low-cost deposits into a BDT 319bn bond book; loans are only 63% of deposits

Dutch-Bangla Bank was set up in 1995 as a Bangladesh–Netherlands joint venture and opened in 1996 [1 p.79]. It is a mass-market retail bank: 12.8mn deposit accounts but only 137,000 loan accounts at December 2025 [1 p.26]. Its deposits come from branches and sub-branches in towns and upazilas, a network of 4,151 ATMs and 4,887 cash-recycling machines, agent banking (5,631 outlets, 7.8mn account holders) and Rocket mobile banking (39.6mn customers) [s1 p.26, p.107]. Those channels collect salary, remittance and small-saver money that pays little or no interest.

The lending side is mostly corporate and SME. In 2025 loan interest of BDT 37.1bn came 29% from industrial term loans (10.6bn) and 42% from SME and consumer finance (15.4bn) [1 p.221]. The bank's own yield on loans was 9.3% in 2025 [1 p.26], well below the system's 12.13% average lending rate [9]: DBBL lends to large borrowers and does not chase yield. Loans grew only 5.1% in 2025 and fell 2.9% in H1 2026 [s1 p.26; s2 p.2], while system loans grew 4.3% in the year to June 2026 [9].

So the surplus goes to government securities. Investments rose from BDT 128bn at end-2024 to 254bn at end-2025 and 319bn at June 2026 [s1 p.26; s2 p.2]. Investment income more than doubled to BDT 22.0bn in 2025, helped by BDT 3.8bn of capital gains on bonds sold as yields fell [1 p.222]. We forecast investment income at 78% of core income (net interest plus investment income) in FY2027F, above the 60% red-flag level in our bank framework: DBBL has become a deposit franchise with a bond portfolio attached.

ATMs, cards and Rocket earn 10% of operating income; their real value is the low-cost deposits they gather

DBBL pioneered shared ATMs, online branches and mobile banking in Bangladesh. Card, ATM and related service income was BDT 3.97bn in 2025 (+8.0%), Rocket mobile banking BDT 1.09bn (+18.6%) and agent banking BDT 0.19bn (+30.6%) [1 p.223]. Together that is 10.1% of total operating income, and none passes our 10% test on its own (Channels sheet). The ATM fleet has shrunk from 4,934 in 2022 to 4,151 while cash-recycling machines grew from 743 to 4,887 [1 p.26]; POS merchant turnover fell 12.6% to BDT 42.0bn as card payments shift to QR codes such as DBBL's NexusPay (9.0mn users, 182,240 merchants) [s1 p.40, p.107].

Rocket. DBBL launched mobile banking in March 2011 and rebranded it Rocket in 2016; it is a division of the bank, not a subsidiary, and has no separate capital or accounts [1 p.79]. The bank reports 39.6mn Rocket customers, 17.0mn of them women [1 p.107]. The market is led by bKash (a BRAC Bank subsidiary) and Nagad: reported by The Daily Star, 14 providers are licensed, bKash claims 4.20 crore active users and Nagad 2.82 crore, and Nagad has been run by a Bangladesh Bank-appointed administrator since 2024 [10]. Rocket's BDT 1.09bn of income is 2.1% of DBBL's operating income and its costs are not disclosed, so we do not value it separately. If it lost half its income for good, the after-tax loss of about BDT 0.33bn a year (against about BDT 10bn of profit), capitalised at Ke − g, would take roughly BDT 3 a share off fair value (our estimate).

Regulatory position. Rocket operates under Bangladesh Bank's permission of April 2010 inside the bank [1 p.79]. We have not verified whether current Bangladesh Bank MFS rules would require a bank-run service to move into a separately capitalised subsidiary; that is a labelled gap. Such a move would ring-fence some capital. Rocket also brings in low-cost wallet balances and transactions, which DBBL does not disclose separately, so the BDT 3 a share above is a floor on the downside: if losing Rocket also cost, say, 2% of deposits (our assumption), the risk table's deposit and funding lines show how much more value is at stake. The line Rocket moves directly is fee income (commission, exchange and other), which we forecast to fall 8% in FY2026F after H1 2026 came in 7.8% lower, then grow 8–9% a year.

DBBL paid 2.6% for deposits in 2025 against a 6.21% system rate; funding cost rose to 3.4% and we take it to 3.6% by FY2028F

Deposit cost is the centre of the DBBL story. Interest paid on deposits was BDT 14.8bn in 2025 on average deposits of about BDT 572bn, a cost of 2.6%, up from 1.8% in 2024 [1 p.222]. Savings deposits cost about 1.5% (BDT 3.9bn on about 255bn) and fixed deposits carried most of the bill (BDT 7.9bn) [1 p.222]. The system's weighted deposit rate was 6.21% at June 2026 [9], so DBBL's funds cost less than half the market's.

The mix is drifting. Term deposits rose from 19.7% of deposits at December 2025 to 20.5% at June 2026 [2 p.2], and subordinated debt of BDT 19bn costs about 12.6% (BDT 2.4bn of interest in 2025) [1 p.222]. Total funding cost (interest expense over deposits and borrowings) rose from 1.69% in FY2022 to 3.4% in FY2025 and was 3.30% annualised in H1 2026 [2 p.4]. We assume 3.4% for H2 2026, 3.5% in FY2027F and 3.6% in FY2028F. With T-bill yields at 8.6–8.8% and the 10-year bond at 9.27% [8], the gap between funding cost and bond yield is about 5.5–6pp.

Deposit growth drives the balance sheet. Deposits rose 19.2% in 2025 and 11.9% in H1 2026 [s1 p.26; s2 p.2]; we forecast 13% in FY2027F and 12% in FY2028F, close to system growth of 10.7% [9] plus DBBL's network gains. That takes deposits to BDT 934bn and investments to BDT 547bn by FY2028F.

Net interest income falls to BDT 8.5bn by FY2028F while investment income rises to 45.0bn

Net interest income fell 32% to BDT 19.9bn in 2025 as funding cost rose and loan yields slipped from 10.0% to 9.4% [1 p.149]; it fell again to BDT 8.7bn in H1 2026 from 11.2bn [2 p.4]. We forecast 15.5bn in FY2026F and 8.5bn in FY2028F: all funding cost sits in interest expense while the new deposits go into bonds, whose income is booked as investment income. The loan yield eases from 9.0% to 8.7% as T-bill rates near 8.7% pull lending rates down.

Investment income was BDT 14.2bn in H1 2026 (9.9% annualised on average investments) [2 p.4]. We use 9.0% for H2 2026, the yield on new T-bills and bonds rather than the H1 rate that included trading gains, then 8.9% and 8.8%, with no repeat of the 2025 trading gains: new money earns the 8.6–9.3% on offer [8]. Core NIM (net interest plus investment income over loans and investments) falls from 6.6% in FY2025 to 5.4% in FY2028F, but the larger balance sheet lifts core income.

Commission, exchange and other income was BDT 10.1bn in 2025, including a BDT 1.5bn treasury dealing gain against a loss in 2024 [1 p.223]; it was 7.8% lower in H1 2026 [2 p.4]. Staff cost rose 21.8% in H1 2026 [2 p.4]; we take +18% for FY2026F and +10% a year after. Cost-to-income stays near 53%.

Cross-check with the nine-month release. The DSE release of 29-Oct-2026 gives nine-month EPS of 6.15, about BDT 6.24bn of profit, and Q3 EPS of 1.80, about 1.83bn [5]. Our second half of BDT 4.52bn (built from the June balance sheet: investment yield 9.0%, funding cost 3.4%, credit cost 2.4%, tax 46%) gives FY2026F of 8.94bn and implies about BDT 2.7bn in Q4, 0.9bn more than Q3. We assume no write-back. The named driver is cost seasonality: operating expenses are lowest in the fourth quarter (BDT 5.4bn in Q4 2025 against about 6.8bn in each of Q1–Q3 [1 p.150]); a similar 1.4bn dip is worth about BDT 0.76bn after 46% tax. That leaves about BDT 0.11bn of the step unexplained by a named driver. If Q4 2026 only matched Q3, FY2026F EPS would be about 0.9 lower.

Gross NPL rose from 6.43% to 9.18% in six months; we provide 2.4% of loans in H2 2026, then 1.6% and 1.4%

DBBL cleaned up hard in 2024: provisions of BDT 17.0bn took profit down to 4.7bn as NPL rose to 7.73% [1 p.26]. In 2025 NPL fell to 6.43% (BDT 29.0bn) and a BDT 6.1bn write-back in the fourth quarter left full-year provisions at 10.2bn [s1 p.150, p.26]. Specific provisions of BDT 26.6bn covered classified loans with a surplus of 0.9bn [1 p.233].

The first half of 2026 reversed the trend. Classified loans rose to BDT 40.2bn (9.18%), of which 26.1bn bad/loss, 11.1bn sub-standard and 2.9bn doubtful, and special-mention loans were BDT 45.1bn (10.3%) [2 p.26]. Specific provisions rose to BDT 29.0bn and general provisions to 9.4bn, a total of BDT 38.4bn against a requirement of 32.4bn, so the bank holds a BDT 6.0bn surplus [s2 p.27, p.33]. Interest suspense grew from BDT 16.3bn to 20.3bn [2 p.26]. For context, the system's overdue loans were 34% of loans at June 2026 [9]: DBBL is far cleaner than the sector, but it is not immune.

We set credit cost at 2.4% of average loans for H2 2026 (H1: 1.9% annualised) as part of the special-mention book migrates, then 1.6% in FY2027F and 1.4% in FY2028F, above the FY2021–23 average of 1.0% [1 p.26]. The NPL ratio peaks at 9.5% in FY2026F and eases to 8.0%. In the valuation we deduct half of the net NPL held at December 2025 (BDT 1.2bn) and use a long-run credit cost of 1.3%. A credit cost of 2.5% in FY2027–28 would take fair value to 42; NPL of 9–10% with 3% credit cost to 36.

Paid-up capital of Tk1,015 crore bars cash dividends from FY2026; DBBL needs 97% more shares to reach Tk2,000 crore

Bangladesh Bank's SPCD circular 06 of 23 May 2026 says a bank whose paid-up capital is below Tk2,000 crore may not pay any cash dividend, and a bank above it may pay at most 50% of its total declared dividend in cash; the rule applies to dividends for the year ending 31 December 2026 and later [6 p.1]. DBBL's paid-up capital is Tk1,015 crore (BDT 10.15bn) after the 5% bonus for 2025 [s2 p.2, p.6]. The gap is Tk985 crore, 97% of today's capital: the largest gap, relative to size, among the banks we cover.

Capital actions in the last 24 months [5]: 10% cash + 10% stock for FY2024 (DSE 30-Apr-2025; AGM 19-Jun-2025); a BDT 12bn floating-rate 5th subordinated bond (BSEC consent Jan-2025, raised in H1 2025, taking sub debt to BDT 19bn; DSE 22-Jan-2025) [2 p.5]; 25% cash + 5% stock for FY2025 (DSE 29-Apr-2026; AGM 16-Jun-2026; paid, DSE 13-Jul-2026); and on 25-Aug-2026 (DSE 27-Aug-2026) the board proposed raising authorised capital from Tk1,500 crore to Tk3,500 crore, put to an EGM on 15-Oct-2026 subject to Bangladesh Bank's no-objection. We have not seen the EGM result. The board also decided in Aug-2025 to buy a 21.5-storey Motijheel building for a head office at about BDT 10.16bn, subject to Bangladesh Bank approval; that is 17% of December-2025 equity and its status is not disclosed. No rights issue has been announced.

The authorised-capital increase is the signal: at Tk1,500 crore DBBL could not have crossed Tk2,000 crore at all. Our base case assumes a stock dividend of about 97% for FY2026, paid in mid-2027 from retained earnings of BDT 39.1bn [2 p.2], and no cash for FY2026. From FY2027 we assume a total payout of 40% of profit, half in cash, so cash is 20% of profit (DPS about BDT 1.17 on the enlarged share count). A stock dividend moves no value: base value per post-bonus share would be 27.0. The alternative, a Tk10 rights issue for the whole gap, means about one new share for every share held; at today's price the theoretical ex-rights price would be 28.6.

The dividend rule matters less to value than to yield. Under our method, profit held back earns the bank's return, so cash paid out or retained changes little: keeping the old 25% cash payout would give 55.8 and no cash at all in FY2026–28 51.6, against a base of 53.2. Capital is ample: CRAR was 17.1% at December 2025 [1 p.26] and rises to 18.7% by FY2028F as profit is retained.

Sponsors and directors own 83.6%; the free float is 16.4%, worth about BDT 7.8bn

Sponsors and directors held 807.8mn of 966.7mn shares (83.6%) at December 2025, down from 87.0% a year earlier; institutions held 6.9% and individuals 9.5% [1 p.23], general public 16.44% in total [1 p.215]. Three holders own more than 10% each: Ecotrim Hong Kong Ltd 24.74%, Horizon Associates Ltd 22.84% and Mr Mohammed Sahabuddin Ahmed 20.21% [1 p.68]. The chairman, Ms Sadia Rayen Ahmed, holds 2.0% [1 p.215]. Only 8,331 shareholders are on the register and 40 of them hold 94.6% of the shares [1 p.215].

At 46.6 the free float is worth about BDT 7.8bn. That keeps trading thin and index weights small, and it means a few holders decide every special resolution, including the authorised-capital change and any rights issue. We do not add a governance discount to the valuation: the bank is rated AAA by CRAB (DSE 24-Jun-2026) [5], its CRAR is well above the minimum and its NPL has stayed far below the sector's. But the low float can keep the share price away from fair value for long periods, and our cost of equity (beta 1.05) does not capture liquidity. A new MD and CEO, Md. Ahteshamul Haque Khan, took office on 7-Feb-2026 (DSE 8-Feb-2026) [5]; the auditor added an emphasis-of-matter paragraph to the FY2025 report, which we have not read (labelled gap).

We forecast net profit of BDT 8.9bn in FY2026F and 13.9bn in FY2028F, with ROE near 17%

Each line comes from its own driver. Interest income = loan yield × average loans; interest expense = funding cost × average deposits and borrowings; investment income = investment yield × the investment book, which is whatever deposits and equity are left after loans, cash and fixed assets. Fees, staff cost and other costs grow from their 2025 base; provisions = credit cost × average loans; tax at 46% for H2 2026 (the H1 2026 rate) and 42% after. FY2026F is the H1 2026 actual plus a second half built from the June-2026 balance sheet.

The result: EPS of 8.81 in FY2026F, 11.54 in FY2027F and 13.72 in FY2028F on 1,015mn shares, and ROE of 14.1%, 16.2% and 16.6%. FY2027F profit jumps because credit cost falls from 2.1% to 1.6% and the bond book keeps growing. Scenarios move the same drivers as the risk table, symmetrically around the base: funding cost ±0.7pp in FY2027–28 (it rose about 1pp in 2025 alone), investment yield ∓0.7pp (half the 2025 move), loan yield ∓0.3pp, deposit growth ∓5pp a year, credit cost +0.6/0.8pp in the bear and −0.6/0.8pp in the bull, and NPL to 11% or 5.5%. Tax stays on the base path. Beyond FY2028 the cases fade back to DBBL's own history: the sustainable ROE is anchored at 9.5% in the bear (the FY2024 trough) and 18.5% in the bull (near the FY2023 best of 17.9%) [1 p.26]; the base uses the normalised ROE. Terminal growth stays at 7% (6% bear); at 9.5% ROE that needs 63% retention, which the bank can fund.

Fair value BDT 11–105, base 53: a 15.4% sustainable ROE justifies 0.83x adjusted book

Method. Residual income on adjusted book (80%) and a justified P/E on normalised EPS (20%). The cost of equity is 17.1%: the 10-year T-bond cut-off of 9.27% [8] plus a Blume-adjusted beta of 1.05 (raw 1.07 against the DSEX over two years of weekly returns, R² 0.36) [7] times a 7.5% equity risk premium. Adjusted book is December-2025 equity of BDT 60.1bn less half of the net NPL (1.2bn) [s1 p.26, p.233]. DBBL has no subsidiaries, and Rocket, cards and agent banking are divisions none of which reaches 10% of income, so a sum of the parts does not apply.

Sustainable ROE. We take FY2028F profit, mark the investment yield to the T-bond less 1pp with 30% of that fall passed through to funding cost, and use a long-run credit cost of 1.3%: normalised ROE is 15.4%. DBBL's reported ROE was 16.1%, 14.4%, 17.9%, 9.5% and 17.3% in FY2021–25 (median 16.1%) [1 p.26]; 2024 was the clean-up trough. Our figure sits slightly below the median: cheaper funding than peers offsets a lower loan yield and a higher NPL than in 2021–23.

Terminal. Growth is 7% (6% in the bear case). Reinvestment = g / ROE_s, so the economic payout is 1 − 7%/15.4% = 54.6%; the SPCD rule caps cash at 50%, and the remaining 4.6% is excess capital that earns only the cost of equity, so it neither adds nor destroys value. Justified P/ABV = (ROE_s − g)/(Ke − g) = 0.83x. Because ROE_s is slightly below Ke, the terminal term subtracts 9.4bn; the FY2026–28F residual income adds -0.6bn. Rolled forward to 30-Oct-2026 and less the FY2025 cash dividend paid, residual income gives 52.6 a share.

Justified P/E. Payout × (1+g)/(Ke − g) = 5.8x, applied to normalised EPS of 9.62 = 55.6, consistent with the residual-income leg because it uses the same ROE, g and payout. We do not use DBBL's own historical P/E (3.9–8.9x in FY2021–25 [1 p.26], partly DSE floor-price years). The peer P/B leg is dropped: peers' book values are available to us only from the screener, not from their filings on the same basis. Blend: 0.8 × 52.6 + 0.2 × 55.6 = 53.2, 0.87x June-2026 book and 6.0x FY2026F EPS.

Funding cost is the largest risk: 1pp higher takes 74% off base fair value

Each risk below is a change to a model driver; the value shown is the base fair value after recalculating the workbook with that change alone. DBBL's balance sheet is BDT 900bn against BDT 62bn of equity, so small changes in funding cost or bond yields move value a lot.

• Regulation: a rule moving bank-run MFS into subsidiaries would ring-fence capital for Rocket (capital, not earnings); a change to the SPCD dividend rule would change only the cash/stock split.
• Head-office building (BDT 10.16bn, pending Bangladesh Bank approval) would add roughly BDT 0.3–0.4bn a year (our estimate) of depreciation and running cost to other operating expenses if bought; not in the base.
• Liquidity: a 16.4% free float can keep the price far from any fair value.
• Interest-rate cuts: lower T-bill yields hit investment income faster than they lower DBBL's already-low deposit cost.
• FY2026 delivery: our FY2026F needs about BDT 2.7bn in Q4 2026 after 1.83bn in Q3 [5].

FAQ

What is Dutch-Bangla Bank's (DUTCHBANGL) fair value?

We estimate BDT 11–105 a share, base 53 and probability-weighted 56, against 46.6 on 29-Oct-2026. It blends residual income on adjusted book (80%) and a justified P/E on normalised earnings (20%).

Can Dutch-Bangla Bank pay a cash dividend for 2026?

Not under the current rule: Bangladesh Bank's SPCD circular 06 of 23 May 2026 bars cash dividends for banks with paid-up capital below Tk2,000 crore from the FY2026 dividend onward. DBBL's paid-up capital is Tk1,015 crore, so it needs about 97% more shares. The board has proposed raising authorised capital from Tk1,500 crore to Tk3,500 crore (EGM 15-Oct-2026).

Why is DBBL's deposit cost so low?

About 80% of its deposits are current, savings and bills-payable balances collected through 243 branches, 334 sub-branches, over 9,000 ATMs and cash-recycling machines, agent banking and Rocket. Deposit interest cost 2.6% of average deposits in 2025 against a 6.21% system deposit rate.

How important is Rocket to Dutch-Bangla Bank?

Rocket earned BDT 1.09bn in 2025, 2.1% of DBBL's operating income, with 39.6mn customers. It is a division of the bank, not a subsidiary. bKash and Nagad lead the market; Rocket's main value to DBBL is the low-cost deposits and transactions it brings.

How bad is DBBL's NPL?

Gross NPL rose from 6.43% at December 2025 to 9.18% at June 2026, with special-mention loans of BDT 45.1bn. Provisions held were BDT 38.4bn, a BDT 6.0bn surplus over the requirement. The system's overdue loans were 34% of loans at June 2026.

Sources

1. Dutch-Bangla Bank PLC Annual Report 2025 (year to 31-Dec-2025); bank and offshore banking unit, no subsidiaries; page numbers are PDF pages · link · accessed 2026-10-31
2. Dutch-Bangla Bank PLC financial statements (un-audited), half year to 30-Jun-2026 · link · accessed 2026-10-31
3. Dutch-Bangla Bank PLC Annual Report 2024 · link · accessed 2026-10-31
4. Dutch-Bangla Bank PLC Annual Report 2022 · link · accessed 2026-10-31
5. Dhaka Stock Exchange news archive for DUTCHBANGL, Nov-2024 to Oct-2026 (dividends, subordinated bond, authorised capital and EGM, head-office building, Q1-Q3 2026 results, MD appointment) · link · accessed 2026-10-31
6. Bangladesh Bank SPCD Circular No. 06, 23-May-2026, dividend declaration policy for banks (Bangla) · link · accessed 2026-10-31
7. AnalystBro market data: DSE prices (DUTCHBANGL, DSEX) and bank screener, 29-Oct-2026 · link · accessed 2026-10-31
8. Bangladesh Bank call money and treasury auction pages, 29-Oct-2026 (10-year bond cut-off 9.2680% on 25-Oct-2026) · link · accessed 2026-10-30
9. Bangladesh Bank, Scheduled Banks Statistics Apr-Jun 2026 (Tables 101-102) · link · accessed 2026-10-30
10. The Daily Star, 'Tk 6,000cr moves daily, not every wallet winning', 11-Jul-2026 · link · accessed 2026-10-31
11. AnalystBro model (workbook on this page) · accessed 2026-10-31