Executive finding
The payment rails were built and the risk markets were not, and FY26 to FY36 is decided by whether the second construction happens
Chapter 07 of 60 in the Bangladesh 2036 research base. Contents of the series.
The payment rails were built and the risk markets were not, and FY26 to FY36 is decided by whether the second construction happens
Bangladesh enters FY26 with a financial system that moves money at emerging market scale and prices risk almost nowhere. The mobilisation record is real: account ownership reached 43.3 percent of adults in the 2024 Findex wave, against 31.0 percent in 2014 [WB Findex 2024, account.t.d], the IMF Financial Access Survey records 238.68 million registered mobile money accounts that moved 1,737,411 crore BDT in 2024, equal to 34.7 percent of GDP [IMF FAS 2024], and bank deposit accounts outnumber adults, at 1,143.4 depositors per 1,000 adults in 2024 [IMF FAS 2024]. The risk pricing record is the opposite: Dhaka Stock Exchange market capitalisation ended FY24 at 662,137.26 crore BDT, 13.12 percent of GDP, down 14.24 percent over the year [BSEC 2024], small enterprise lending has fallen from 9.09 percent of GDP in 2017 to 5.98 percent in 2024 [IMF FAS 2024], and the share of firms financing investment through banks fell from 24.7 percent in 2007 to 18.8 percent in 2022 [WB ES 2022]. The thesis of this chapter is that the decade to FY36, the horizon the chapter 15 scenarios assume, is decided by whether institutions that price risk, an equity market with real liquidity, a corporate bond market, funded pensions, workable insurance and collateral light small firm credit, are constructed on top of the payment rails that already exist, or whether the system remains what it is today, a payments machine, a government debt machine and a bank cleanup in the middle, with the bank half of that story carried in chapter 06.
Access means payments, not credit
The access record is an asymmetry between saving and borrowing. Depositor accounts per 1,000 adults rose from 262.9 in 2006 to 1,143.4 in 2024, more than one account per adult, while borrowers from commercial banks per 1,000 adults went from 85.5 in 2004 to 96.4 in 2024, essentially flat for two decades [IMF FAS 2024]. Domestic credit to the private sector by banks stood at 35.7 percent of GDP in 2024 [WB WDI 2026, code]. The 2022 census says the same thing in one line. The lake's headline of 18.61 percent of the population with a financial institution account could not be confirmed, since the census district tables in the data lake average 24.34 percent across the 64 districts, but the mobile financial services account rate, 38.53 percent as the unweighted mean of the 64 district rates, sits above the bank account rate on either reading [BBS Census 2022]. The census measures persons, not households, and tabulates both rates by sex.
The Findex arc runs 31.0 percent of adults in 2014, 50.0 in 2017, 52.8 in 2021, then 43.3 in 2024 [WB Findex 2024, account.t.d]. A 9.5 point fall in measured ownership while registered mobile money accounts grew to 238.68 million is the first puzzle of this chapter, and the dormancy gap resolves it: only 88.89 million of the registered accounts were active in 2024 [IMF FAS 2024]. Accounts were opened for one flow, wage payment, remittance collection or a government allowance, and stayed open and idle, which a survey of active use reads as exclusion.
The equity market prices banks more willingly than the real economy
The FY24 record is a market shrinking in every dimension at once. DSE market capitalisation fell 14.24 percent to 662,137.26 crore BDT at end FY24, from 772,078.04 crore a year earlier, and stood at 13.12 percent of GDP against 17.39 percent the year before [BSEC 2024]. The DSEX fell 1,015.69 points, 16.01 percent, to 5,328.40, turnover fell 21.51 percent to 149,992.62 crore BDT, average daily turnover fell from 792.89 crore to 622.38 crore, BO accounts fell 4.60 percent to 1,775,146, and the market price earnings ratio compressed from 14.34 to 10.22 [BSEC 2024]. The fall continued into FY25: the DSEX closed at 4,838.39 at end June 2025, down a further 9.2 percent, with market capitalisation of 662,271 crore BDT and average daily turnover of 355 crore BDT in the April to June quarter [BSEC 2025]. The May 2026 exchange snapshot in the data lake lists 396 quoted instruments, records session turnover of 4,866 million BDT, about 487 crore BDT, across 133,735 trades [DSE 2026].
By capitalisation, manufacturing took 43.0 percent, services and miscellaneous 32.3 percent, the financial sector 23.6 percent and bonds 1.2 percent in FY24 [BB AR 2024]. By trading, banks and insurers took eight of the twenty highest value instruments in the May 2026 snapshot [DSE 2026], a flow that is a symptom of the distress trading chapter 06 documents as much as a channel for pricing the non financial economy.
The mobile money rail is the genuine construction project of FY10 to FY25, and it holds 0.26 percent of GDP by design
Registered mobile money accounts grew from 0.13 million in 2011 to 238.68 million in 2024, active accounts from essentially zero to 88.89 million, and agent outlets from 2,551 to 822,726, a footprint that dwarfs the 19,829 automated teller machines and 15.87 machines per 100,000 adults recorded in 2024 [IMF FAS 2024]. Transaction value reached 1,737,411 crore BDT in 2024, 34.7 percent of GDP [IMF FAS 2024]. The balance tells the structural story: mobile money balances held 0.26 percent of GDP in 2024, against commercial bank deposits at 37.89 percent [IMF FAS 2024]. The rail is a payment system, not a savings system, and the institutional design keeps it that way, since the providers, bKash and Nagad among them, operate under payment rules that hold customer funds in bank accounts rather than intermediating them. The formalisation dividend chapter 03 documents, the 26.83 percent remittance jump to 30,328.80 million USD in FY25 after the float [BB Econ 2025, table XVIII], ran over this rail, which is why the rail matters beyond finance: it is the pipe between chapter 03's external cushion and chapter 04's transfer programmes.
The two retail channels moved in opposite directions: microfinance grew and bank agents closed
Agent banking outlets went from 448 in 2015 to 21,601 in 2023, then fell for the first time to 21,248 in 2024 [IMF FAS 2024]. The card stock keeps expanding, 39.57 million debit cards and 2.67 million credit cards in 2024 [IMF FAS 2024], but the rural agent channel's reversal in the first year of the cleanup matches the commission structure: agents are paid out of transaction revenue from remittance and government flows, and the weak banks that sponsor most agent networks cut the channel when their balance sheets tighten.
The microfinance tier is the intermediation that worked. Its loan book stood at 3.19 percent of GDP in 2024, from 2.69 percent in 2021 and after a 3.35 percent peak in 2023, delivered through 26,070 branches with 28.96 million female borrowers [IMF FAS 2024]. Grameen Bank alone held 10.27 million members, 96.85 percent of them women, with cumulative disbursement of 275,628 crore BDT and an annual recovery rate of 92.84 percent in 2022 [Grameen Bank / BBS 2022].
The non bank financial institutions are the retail intermediation that failed, on the same related party networks chapter 06 traces into the banking book. The 35 finance companies held assets of 1,007.17 billion BDT at end June 2024, up from 996.85 billion a year earlier, while their deposits fell to 475.14 billion BDT from 477.22 billion [BB AR 2024]. Classified loans reached 33.83 percent of the sector's loans at end 2024, up from 15.03 percent in 2020, and the sector's capital adequacy ratio ended 2024 at minus 6.46 percent, from plus 14.17 percent in 2020 [BB FSR 2024]. A tier with negative aggregate capital cannot intermediate, and its run off is the funding vacuum between banks and microfinance.
The four markets that would price risk are thin, absent or insolvent
The corporate bond market barely exists as a listed venue. Sixteen corporate bonds were listed at end FY24, up from eleven a year earlier, alongside eight debentures, and bonds carried 1.2 percent of DSE market capitalisation in FY24 [BSEC 2024] [BB AR 2024]. The outstanding stock relative to bank credit has no series in the data lake and could not be confirmed; the Bangladesh Bank Financial Stability Report would settle it [BB FSR 2024].
The sovereign sukuk programme is the one tested issuance machine. The government issued its maiden Bangladesh Government Investment Sukuk in December 2020 as an Ijara structure with Bangladesh Bank as special purpose vehicle, and total outstanding BGIS stood at 190.00 billion BDT, 19,000 crore BDT, at end June 2024, when a further 10.00 billion BDT Social Impact Sukuk was accepted out of 44.14 billion BDT of bids [BB AR 2024]. It is government paper; it has not created a corporate market.
Insurance is the missing risk transfer market, and its regulator's own arithmetic makes the point. Total gross premium was 18,226.77 crore BDT in calendar 2023, up 5.38 percent on 2022, split 12,273.49 crore life and 5,953.28 crore non life, which puts penetration at 0.41 percent of GDP, down from 0.44 percent in 2022, with about 16.49 million people insured out of a population of 173 million [IDRA 2024]. A sector with premium under 1 percent of GDP and coverage under one person in ten neither mobilises savings nor prices catastrophe at scale.
The institutional bid is equally thin. Total assets of mutual funds stood at 10,729 crore BDT at end FY24, closed end funds 5,285 crore and open end funds 5,444 crore, down from 11,623 crore a year earlier [BSEC 2024]. Foreign portfolio investors have been net sellers for the market's life: since April 1992, gross inflow through non resident taka accounts reached 537.0 billion BDT against gross outflow of 545.9 billion, and in FY24 alone the outflow of 40.6 billion BDT exceeded the inflow of 19.1 billion [BB AR 2024].
Firms finance themselves from retained earnings because the bank and market channels both failed
In the 2022 enterprise survey wave, 38.7 percent of firms used banks to finance working capital, up from 29.9 percent in 2013 but still below the 43.1 percent of 2007, and 18.8 percent used banks to finance investment, down from 19.8 percent in 2013 and 24.7 percent in 2007 [WB ES 2022]. Self finance and retained earnings do the rest, which is the micro foundation of the investment rate stall chapter 01 measures. The price at which any corporate paper would clear is set off a sovereign risk benchmark that moved from 605 basis points on the five year CDS estimate in 2023 to 480 basis points in 2024 [IMF Art IV 2024].
Family firms will not list, funded buyers do not exist and the sovereign crowds the curve: why every risk market starves
The equity market stayed small because both sides of the trade had better options. On the supply side, the real economy's surplus sits in family controlled firms that fund expansion from retained earnings, garment cash flow and land collateral, and that face a governance discount in public listings, disclosure, minority shareholder rights, related party discipline, they do not face privately; the mechanism that would force listings, bank capital charges on concentrated credit, was neutralised by the lax classification regime chapter 06 documents. On the demand side, the natural buyers, pension funds and insurers, either do not exist as funded pools or are too small to carry a market: civil service pensions are pay as you go, the Universal Pension Management Act 2023 [MinLaw 2023] created the authority and the scheme but the fund's asset stock could not be confirmed from the sources this chapter draws on, and the mutual fund sector's 10,729 crore BDT [BSEC 2024] is a residual. Between the two sits a trust deficit with dates: the 1996 and 2010 to 2011 crashes each wiped out a generation of retail investors, and the floor price imposed in the 2022 stress, its July 2022 imposition date not confirmed from the sources this chapter draws on, which the commission ordered lifted during FY24 [BSEC 2024], froze price discovery deep into the down market. The margin loan channel connected the two: bank merchant bank arms lent against shares, the related party networks of chapter 06 used the same channel, and the classification reset withdrew margin finance at the same time.
The rails grew because the state and the migration economy gave them flows to carry. The agent of agent model put cash in and cash out within walking distance of every village at a cost per transaction that branches cannot match, and the flows arrived: remittances formalised after the 2024 float that chapter 03 documents, wage payments moved as the garment sector digitised payroll, and government allowances moved onto the same pipes. The rail's economics are transactional, which is why it holds no money: balances at 0.26 percent of GDP [IMF FAS 2024] are the regulatory design working as intended, customer funds swept into bank accounts, and also the design's ceiling, since a system that cannot hold money cannot intermediate it. Dormancy is the rail's unfinished half: the second account opening, into savings, credit and insurance products, is the access frontier now, and the Findex fall to 43.3 percent [WB Findex 2024, account.t.d] says the frontier is not crossing itself.
Small firm credit shrank for the same reason big firm credit was mispriced, in mirror image. SME loans peaked at 9.09 percent of GDP in 2017 and fell to 5.98 percent by 2024 [IMF FAS 2024], a decade of decline that ran through three mechanisms. First, the state's answer to the SME gap was directed refinance schemes, which subsidise the price of credit without building the underwriting capacity to price it, so the capacity was never built and the schemes' share of the book, which could not be confirmed from the sources this chapter draws on and which Bangladesh Bank scheme reports would settle, stayed the load bearing wall. Second, collateral first lending concentrates any bank's portfolio in land and plant owners, which in a system with weak foreclosure is collateral that does not collect, so the rational bank lends against relationships and scale, not against small cash flows. Third, the classification reset of chapter 06 repriced risk across the banking book, and repricing lands first on the small tickets whose information cost per taka is highest. The result is the firm survey record: bank financed investment down to 18.8 percent of firms [WB ES 2022] and the SME share down to 5.98 percent of GDP [IMF FAS 2024], with the credit bureau and digital identity rails that could change the information cost still being wired, which chapter 13 carries.
The state's own footprint completes the mechanism. The treasury's borrowing, chapter 04's financing triangle, is met from the same deposit pool the access agenda mobilised, the savings certificates pay administered rates that compete with every private paper the market might issue, and the government's own yield curve is the only benchmark curve the system has, which chapter 05 documents repricing after the lending cap went. A corporate bond market needs that curve plus a banking system with underwriting balance sheets, which is the cleanup's spillover into this chapter: banks that stop warehousing legacy loans become the first underwriters, chapter 06's market deepening upside. Until then the sovereign risk price, 480 basis points on the five year CDS estimate in 2024 [IMF Art IV 2024], is the benchmark off which any private risk paper would clear, and at that base rate the domestic corporate market cannot price.
Four decisions with named authors separate the FY27 to FY36 scenarios: pensions, underwriting, conduct and rail licensing
The pension decision, authored by the Finance Division, is whether the scheme under the Universal Pension Management Act 2023 [MinLaw 2023] accumulates funded assets and whether private employer pension vehicles are licensed; funded pools are the domestic institutional bid the equity and bond markets have never had, and the reform scenario in chapter 15 assumes they exist at scale by the early 2030s. The bond market decision, authored jointly by Bangladesh Bank, the securities commission and the Finance Division, is whether the post cleanup banking system underwrites corporate paper off a market determined government curve; the IMF programme reviews carry the financial sector benchmarks [IMF Art IV 2025]; no Financial System Stability Assessment exists for this cycle, the last being February 2010 [IMF FSSA 2010], and the 2025 Article IV package names no assessment programme [IMF Art IV 2026], so the sequencing question has no current Fund document behind it. The market conduct decision, authored by the securities commission after the 2024 leadership reset, is whether enforcement replaces floor prices as the response to price stress [BSEC 2025]. The rail regulation decision, authored by the central bank as payment system regulator, is whether mobile money providers are licensed into savings and credit products or held as payment channels; the rail's decade is decided in this rule book more than in any market.
The Eighth Five Year Plan carried capital market deepening targets, market capitalisation and bond stock shares of GDP, that the data lake does not hold, so the specific numbers could not be confirmed and the plan document would settle it [GED Plans 2020]. What the verified record supports is the order of operations the reform scenario assumes: resolve the banks, so the margin and related party overhang leaves the market; build the institutional bid, so primary issuance has a buyer; regulate the rail into intermediation, so the 88.89 million active accounts [IMF FAS 2024] become addressable customers for products rather than channels for flows. The baseline scenario assumes none of the four decisions lands cleanly, the rail keeps processing payments, the equity market keeps clearing bank and insurance paper, and the SME share keeps drifting down; the stall scenario assumes a price shock brings back price floors and the market freezes again, the mechanism the 2010 to 2011 record demonstrates.
Three risks print in monthly series and the upside is a single institutional act: turning payroll into assets
Risks. First, rails without intermediation: the payment system formalises flows while savings sit in cash, land and gold, and the account ownership measure keeps eroding from its 43.3 percent 2024 level [WB Findex 2024, account.t.d]; the revealing indicator is mobile money balances to GDP, 0.26 percent in 2024 [IMF FAS 2024], which stays flat if the rail remains a pipe. Second, a market freeze: a price shock met with floor prices again collapses liquidity below the 487 crore BDT session level of the May 2026 snapshot [DSE 2026], kills the IPO pipeline, which raised only 8.4 billion BDT from nine listings in FY24 [BB AR 2024],; the revealing indicator is daily turnover and the floor price notice itself. Third, the SME famine: if small enterprise loans fall below 5 percent of GDP [IMF FAS 2024] while bank financed firm investment stays under 20 percent [WB ES 2022], the employment intensive segments of chapters 08 and 09 lose their financing, and the investment stall of chapter 01 becomes structural.
Upside. First, funded pools: a pension reform that converts payroll into assets creates the institutional demand for equity and corporate bonds that fifteen years of market development plans could not manufacture; the revealing indicator is the first published pension asset stock in the Finance Division accounts. Second, underwriting capacity: banks freed of the legacy book become bond underwriters, and the corporate bond market's decade starts where chapter 06's resolution ends; the revealing indicator is the outstanding corporate bond stock in the Financial Stability Report [BB FSR 2024]. Third, data led credit: transaction ledgers from the mobile money rail, attached to digital identity, make cash flow underwriting of small firms cheaper than collateral machinery, and the 28.96 million female microfinance borrowers [IMF FAS 2024] are the identifiable graduation pipeline from micro to enterprise credit, the mechanism chapter 06 names as the inclusion dividend; the revealing indicator is the SME loan share recovering from 5.98 percent of GDP toward its 2017 level of 9.09 percent [IMF FAS 2024].
What to watch: five thresholds that separate rail from system
- Account ownership. Current value 43.3 percent of adults in the 2024 Findex wave [WB Findex 2024, account.t.d]. Threshold: a recovery above 50 percent signals the access regime rebuilding on products rather than accounts; a fall below 40 percent confirms the exclusion trend the 2024 wave began.
- Mobile money transaction value to GDP. Current value 34.7 percent in 2024 [IMF FAS 2024]. Threshold: sustained growth above 40 percent with rising balances to GDP signals the rail becoming a financial system; a stall below 30 percent signals the rail saturated as a payment pipe.
- SME loans to GDP. Current value 5.98 percent in 2024 [IMF FAS 2024]. Threshold: a recovery above 8 percent is the small firm credit turn the reform scenario targets by around FY31; a fall below 5 percent is the famine regime.
- Equity market turnover. Current value 622.38 crore BDT average daily turnover in FY24 [BSEC 2024], 355 crore BDT in the April to June 2025 quarter [BSEC 2025], a single session of 487 crore BDT in the May 2026 snapshot [DSE 2026]. Threshold: sustained sessions above 1,000 crore BDT signal the liquidity regime returning; price floors or sessions below 300 crore BDT signal the freeze.
- Agent banking outlets. Current value 21,248 at end 2024, the first fall in the series [IMF FAS 2024]. Threshold: renewed growth above 25,000 signals the rural channel expanding under stronger sponsors; a second consecutive fall signals the branch and agent network contracting faster than digital substitution replaces it.
Sources used
[WB Findex 2024] World Bank Global Findex database, 2024 wave via bdpolicy.db parquet findex_bd_full, series: account.t.d. [IMF FAS 2024] IMF Financial Access Survey via bdpolicy.db, series: imf_fas_mobile_money_registered_accounts, imf_fas_mobile_money_active_accounts, imf_fas_mobile_money_tx_value_bdt, imf_fas_mobile_money_tx_pct_gdp, imf_fas_mobile_money_balances_pct_gdp, imf_fas_agent_banking_outlets, imf_fas_commercial_bank_depositors_per_1k, imf_fas_commercial_bank_borrowers_per_1k, imf_fas_sme_loans_pct_gdp, imf_fas_mfi_loans_pct_gdp, imf_fas_mfi_borrowers_female, and the card, ATM, MFI branch and deposit series cited. [DSE 2026] Dhaka Stock Exchange quote board snapshot captured 2026-05-03 via bdpolicy.db parquet dse_latest_snapshot. [BSEC 2024] Bangladesh Securities and Exchange Commission Annual Report 2023-24, market depth, listed securities, mutual fund assets and floor price lifting orders, via ocr_text/bsec. [BSEC 2025] Bangladesh Securities and Exchange Commission Quarterly Review, April to June 2025, DSE operational statistics, via ocr_text/bsec_deep. [BBS Census 2022] Bangladesh Bureau of Statistics Population and Housing Census 2022 via bdpolicy.db indicators table (bbs_census_financial_account_pct, provenance not recorded) and the district workbook bbs_census/bangladesh_bbs_population-and-housing-census-dataset_2022_admin-02.xlsx (census2022_mfs_account_national_avg_pct, unweighted mean of 64 district rates). [WB ES 2022] World Bank Enterprise Surveys, Bangladesh 2022 wave via bdpolicy.db parquet wb_enterprise_surveys_full, series: IC.FRM.BKWC.ZS, IC.FRM.BNKS.ZS. [BB Econ 2025] Bangladesh Bank Monthly Economic Trends October 2025, table XVIII, remittances FY24 23,912.22 and FY25 30,328.80 million USD. [BB AR 2024] Bangladesh Bank Annual Report 2023-24 via ocr_text/bb/annual_report/ar2023-2024.txt, chapter 6 BGIS sukuk, chapter 8 finance companies, chapter 9 capital market, IPO and NITA flows. [BB FSR 2024] Bangladesh Bank Financial Stability Report 2024 via bdpolicy.db, series: bb_fsr_nbfi_classified_loans_ratio, bb_fsr_nbfi_car; resolving source for outstanding corporate bond stock. [Grameen Bank / BBS 2022] Grameen Bank membership and disbursement series via bdpolicy.db, series: grameen_bank_members_total, grameen_bank_female_member_pct, grameen_bank_cumulative_disbursement_crore, grameen_bank_annual_recovery_rate_pct. [IDRA 2024] Insurance Development and Regulatory Authority Annual Report 2023-24 via ocr_text/idra_deep, gross premium, penetration table and coverage. [IMF Art IV 2024] IMF Article IV via bdpolicy.db parquet bd_sovereign_spreads, series: bd_sovereign_spreads_cds_5y. [IMF Art IV 2025] IMF Article IV and programme reviews, financial sector benchmarks. [IMF FSSA 2010] IMF and World Bank, Bangladesh: Financial System Stability Assessment, Country Report No. 10/38, February 2010, on information to 30 July 2009; the prior assessment was the unpublished 2003 FSAP. cr1038.pdf, read 2026-09-06. [IMF Art IV 2026] IMF, Bangladesh 2025 Article IV Consultation, Country Report No. 26/24, Board 26 January 2026; the package carries no FSAP or FSSA reference. 1bgdea2026001, read 2026-09-06. [MinLaw 2023] Universal Pension Management Act, 2023, Act text on bdlaws.minlaw.gov.bd; title and year confirmed on the National Pension Authority laws page, upension.gov.bd/Public/Laws, accessed 2026-09-06. [WB WDI 2026] World Bank indicator snapshot updated 2026-09-04 via bdpolicy.db; domestic credit to private sector by banks via parquet wb_financial_sector_bd, code FD.AST.PRVT.GD.ZS. [GED Plans 2020] General Economics Division, Eighth Five Year Plan, capital market deepening targets, resolving source.
Verified line by line against primary sources: 86 claims checked, 5 corrected.
Previous: 06 Banking sector
Cite / Reproduce
BDPolicyLab Research. (2026). 07 Capital markets and access to finance. Bangladesh Policy Laboratory. https://bdpolicylab.com/publications/2026-09-06-bangladesh-2036-ch07-capital-markets
Method and source
Source: Primary sources cited at point of use in the publicationAs of 6 Sep 2026