Bangladesh Telecommunications Sector Analysis

The signal is everywhere; the money is not. Bangladesh's mobile network reaches almost the entire population, but average billing per user is only about $2.50 per month (GSMA Intelligence), one of the lowest figures in South Asia. That single number governs the sector. It explains why a market of 186.06 million subscribers generates just $4.2 billion in revenue, why operators hesitate on 5G, and why affordability rather than coverage is the live policy question. The lever that moves the sector is BTRC spectrum pricing, not additional towers.

Coverage is solved; affordability is not

4G covers 98.3% of the population (GSMA), so physical signal is not the constraint on internet use. Demand is. At $2.50 ARPU, the unit economics reflect what users can pay, not what the network can deliver. The subscriber headline also overstates reach: at 1.68 SIMs per user, the unique base is materially smaller than 186.06 million, with multi-SIM behavior driven by tariff arbitrage rather than additive connectivity. Closing the usage gap means lowering the effective price of data and devices, not extending coverage that already exists.

Thin ARPU and front-loaded spectrum costs starve 5G of capital

The 5G business case fails at $2.50 ARPU. Consumer enhanced mobile broadband offers little revenue uplift when per-user revenue is this thin, so the only viable near-term case is enterprise: industrial IoT for garment exporters, port logistics, and precision agriculture. Building that case requires capital, and the sector's capital is being absorbed before it reaches the network. High upfront auction prices divert investment from deployment into license fees. In a low-ARPU market with long payback horizons, that front-loading is effectively a tax on network quality. With 1,247 MHz allocated and no 5G auction date announced, the binding decision sits with the regulator, not the operators.

A real contraction from policy, not demand

Subscriber counts swung for regulatory reasons, and the contraction was real, not cosmetic. The June 2025 SIM cap (reduced from 15 to 10 SIMs per NID) and a re-verification drive removed inactive and duplicate SIMs. Internet subscriptions fell from a July 2025 peak of 135.99 million to 128.27 million in February 2026, the lowest level since early 2023, then recovered to 129.62 million in March 2026 (BTRC; Financial Express, 2026). The SIM-cap policy is sound on fraud and KYC grounds, but it means headline subscriber growth is a poor read on real connectivity right now. ARPU and affordability are the cleaner signals.

Recommendations

1. BTRC: shift to phased, usage-based spectrum fees benchmarked to South Asian peers. Replace high upfront auction prices with fees spread over the license term and tied to deployed capacity. Expected effect: capital that today pays license fees instead funds network densification, improving the 5G payback math at $2.50 ARPU. Success signal: a published fee schedule with a per-MHz benchmark against at least three South Asian peers within the next license cycle.

2. BTRC: set and publish a 5G spectrum auction date and finalize network-slicing rules under a reviewed National Telecommunications Policy. The enterprise 5G case (industrial IoT, ports, agriculture) cannot be built without assigned spectrum and slicing frameworks. Expected effect: operators can plan enterprise deployments instead of deferring capex indefinitely. Success signal: a gazetted auction date and slicing rules issued for consultation.

3. ICT Division and operators: target the affordability gap directly. Pair device financing and zero-rating of essential public services with the existing universal service fund. Expected effect: raise data usage among low-income users without waiting on network expansion that is already complete. Success signal: a measurable rise in average data consumption per user reported in BTRC quarterly data.

4. BTRC: raise mandated tower-sharing ratios toward South Asian peer levels. Sharing ratios remain below peers, leaving operators with duplicative capex on a base of roughly 35,500 towers (BTRC reference). Expected effect: lower per-operator deployment cost frees headroom for 5G and rural coverage without raising total industry capital expenditure. Success signal: a published co-location mandate with a target sharing ratio and a compliance reporting cadence.

What would change this view

If a 5G auction is announced with phased fees and operators commit enterprise capex, the investment-starvation thesis weakens and the constraint shifts from regulation back to demand. If verified ARPU rises materially above $2.50, for example through data-volume growth offsetting price compression, the case for fee reform softens. And if the post-re-verification recovery holds and internet subscriptions climb back above the July 2025 level of 135.99 million, the policy-contraction read would need revisiting.

Sources: BTRC monthly statistical reports (March 2026); GSMA Intelligence Bangladesh; ITU World Telecommunication/ICT Indicators; The Financial Express (2026).