The Bangladesh Telecommunication Regulatory Commission keeps its memory in lists, and one of them, a summary of every divisional internet service provider licence in force, carries at entry 316 a company called ZX Online Limited, registered office 102, Aziz Co-operative Super Market, Shahbagh, Dhaka-1000, licence BTRC/LL/ISP-Central Zone (49) ZX/2008-44, dated 22 July 2008. The address matters more than it seems. Aziz Co-operative Super Market is a low-rise arcade of small shops near Dhaka University, better known for books and panjabis than for infrastructure. In July 2008, somebody running an internet operation out of unit 102 did the thing that separates Dhaka's licensed operators from the uncounted informal cable networks strung between the city's rooftops: they put a limited company on paper, paid the regulator's fees, and took a licence.

To see why the act mattered, recall what Dhaka connectivity looked like in 2008. The city's internet travelled through a barely regulated ecology of cyber cafés, building-by-building LAN operators and cable resellers, most of them unregistered, paying no fees, filing nothing, and buying their upstream capacity second- or third-hand from whoever held paper. The regulator's answer was to sell paper in tiers — national, zonal, later divisional, district and upazila classes — so that formality itself became a graded product: the more territory and legitimacy an operator wanted, the more it paid. Thousands eventually bought in at the bottom of the ladder; the association that organises the licensed industry now counts 2,196 member firms, most of them holding the smallest class of licence, while an unknown number of informal networks still operate below the paperwork line entirely. Zx Online bought in early, one tier from the top, in the capital. What follows is the ledger of that decision.

That act — the formalisation itself — is the asset this piece prices. Not the fibre, which in Mohammadpur is a commodity; not the brand, which has been rebuilt at least twice; but the wrapper of incorporation and licence that lets a small operator buy wholesale bandwidth lawfully, hold its own numbering resources, bill a bank, and survive a regulator's inspection. Bangladesh is an unusually clean laboratory for the question, because the state has fixed nearly every other variable. Retail broadband prices are capped by decree under the one-country-one-rate directive; wholesale bandwidth and transmission prices are capped by a published tariff schedule; even the contention ratio — how many customers may share a megabit — is written into approvals. When the product, the price and the quality floor are all fixed by the state, what remains to differentiate one operator from a thousand others is the paperwork tier it occupies and what that tier permits it to sell.

Zx Online's own price card states the answer with unusual clarity. On its current site the company lists shared home fibre at 1,050 taka a month for 100 megabits — 10.5 taka per megabit — and, a few lines down, dedicated small-business internet at 5,000 taka for the same 100 megabits: 50 taka per megabit, a premium of 4.8 times, both list prices, same seller, same card, same month. The gap between those two numbers is not a bandwidth cost; upstream, the megabits are identical. It is the price of everything the licence and the company's paper make it possible to promise: a service-level guarantee, a public IP address, a signed contract a bank's auditor will accept. The rest of this article works out what it costs to be allowed to charge that premium, and whether the arithmetic still closes as the regulator rewrites the rules.

Three addresses, two dead brands, one continuous licence

Every identity trail on this company leads back to the licence, because almost nothing else has stayed still. The regulator's divisional summary has it in Shahbagh. The Asia-Pacific numbering registry has it at 68/4 Green Road, Dhaka-1205 — the address on the APNIC records behind AS58889, the autonomous system registered to Zx Online Ltd, whose abuse-contact object was last touched in May 2026 and therefore still administered. The company's current website puts its office in a third place: House 9, Main Road, Mohammadi Housing, Mohammadpur, Dhaka-1207, with coverage claimed in Mohammadpur, Dhanmondi and Shyamoli. The archived record explains the sequence: the 2016 and 2020 site captures both carry the Green Road address, so the company moved from the Shahbagh arcade to Green Road sometime after licensing and westward to Mohammadpur in the 2020s. Same phone thread throughout: the hotline on today's site, 01707 255 477, is the number on the company's trade-association record.

The licence number itself repays reading. BTRC/LL/ISP-Central Zone (49) ZX/2008-44 reads as the forty-ninth licence of the regulator's Central Zone series and the forty-fourth issue of 2008 — an early joiner, either way, of a class that no longer exists. The December 2020 guideline retired the old zonal geography and redrew the tiers along administrative divisions, and its conversion clause let incumbents roll into the new classes without paying fresh evaluation or acquisition fees. Zx Online's 2008 paper thus became a Dhaka Division licence at no new capital cost — which is why a document headed Central Zone still anchors a company that today calls itself divisional. The licence is the one identifier that has never changed hands, addresses or spelling in eighteen years, and every other record hangs off it.

Among those hanging records, the one held by the Internet Service Providers Association of Bangladesh does the heavy lifting a company registry search normally would. ISPAB lists ZX Online Limited as general member G-074: BTRC licence type divisional, licence number matching the regulator's list, establishment 15 July 2008, a Dhaka South City Corporation trade licence issued in 2019, a business identification number for VAT, a taxpayer identification number, and a contact whose email address is simply chairman@ the company domain. It also carries a quiet confession: the profile shows no registered addresses and asks the member to complete its directors' details for a verified badge. The one query this reconciliation could not complete is the corporate registry itself. The Registrar of Joint Stock Companies' search portals did not respond to repeated attempts on 3 July 2026 — the main site unreachable, the application host silent — so the company's paid-up capital and shareholder list stay dark. Nothing in the visible record contradicts the incorporation the name asserts; the licence, trade licence, VAT and tax identifiers all presuppose it. But the registry's silence means the ownership behind the chairman email cannot be independently confirmed, and this analysis says so rather than assuming.

The brand archaeology is where the record turns eloquent. The corporate domain, zxonlineltd.com, was registered on 12 February 2009 and has been renewed continuously since — paid through 2027. Its earliest capture, from March 2009, is a period piece: a template site, lorem ipsum still in two columns, declaring Zx Online "Largest ISP in Bangladesh" and listing its points of presence — which turn out to be eight named third parties, cyber cafés and neighbourhood networks from Mirpur to Chawk Bazar that resold its bandwidth. A second domain, zxonlinebd.com, served the era's real product: by 2012 it redirected to a movie-and-software download server called FreeDownloadBD, the FTP honeypot that anchored every Dhaka ISP's value proposition before cheap international bandwidth. Both of those side brands are now dead in instructive ways. The download domain lapsed and was re-registered in January 2026 through a drop-catching service. The alias domain lapsed and was re-registered in May 2025 through a registrar favoured by domain speculators; it now serves a Chinese-language template for a betting brand dressed as a power-installation firm. A public directory that records this company as two domain references and one live website is, without knowing it, recording a company that let half its historical web presence rot while renewing the half that carries the licence's name. The wrapper survived; the ornaments did not.

What formality costs, line by line

The costs of the wrapper are unusually legible, because the regulator publishes them. Under the ISP licensing guideline of December 2020, a divisional licence — Zx Online's class, converted from the old Central Zone category the guideline retired — costs 10,000 taka to have an application evaluated, 200,000 taka to acquire, 100,000 taka a year to keep, 200,000 taka to renew each five-year term, and a further 200,000 taka locked in a bank guarantee that the regulator may encash if fees fall late; a licensee that prepays its annual fees for the full term escapes the guarantee. Miss a payment and a 15 percent per-annum late levy runs for sixty days before cancellation proceedings begin. In steady state, then, the regulatory cost of being Zx Online is 500,000 taka in annual fees plus a 200,000 taka renewal every five years: 700,000 taka per cycle, or 140,000 a year — about US$1,150 at mid-2026 exchange rates. The same guideline prices the alternative tiers around it: a nationwide licence at 250,000 a year, a district licence at 50,000, and an upazila licence at 10,000 — the last being the 25,000-taka-acquisition tier through which the regulator formalised hundreds of small-town operators.

The regulator is only the first counter. Incorporation drags a tail of parallel obligations, each visible in the company's own identifiers. The 2019 trade licence from Dhaka South City Corporation renews annually. The business identification number means value-added-tax registration and monthly returns — and the company's price card shows the tax passing through: every home tier is a round number multiplied by exactly 1.05, so 600 becomes the listed 630 and 2,500 becomes 2,625, the arithmetic signature of the 5 percent VAT on internet service. (That rate itself wobbled in January 2025, when the revenue board slapped a 10 percent supplementary duty on ISP services and then withdrew it completely within two weeks under public pressure — a reminder that the fiscal terms of formality can change twice in a month.) The taxpayer identification number means corporate income-tax filings; a limited company must also file audited accounts with the registrar each year, an audit that costs real money whether or not anyone ever reads it — and, as noted, the registrar's portal would not let this piece read it.

Set against these outflows is the side of formality the fee schedules never show: what the documents are worth at a bank counter. An informal cable network is unbankable almost by definition — no audited turnover to underwrite, no licence to survive a compliance review, no registered assets to charge. A limited company with seven years of filed accounts, a regulator's licence and a corporate client book can, at least in principle, borrow working capital against receivables and equipment, and in a business where growth is bought in drums of fibre and crates of subscriber terminals, access to credit is the difference between compounding and treading water. The record shows no borrowing by Zx Online either way — the registrar's silence hides any charges filed against its assets — so the credit dividend stays a mechanism rather than a measurement here. But it belongs in the ledger, because it is the one benefit of incorporation that scales with ambition rather than with paperwork.

Then comes the cost few informal operators ever see: the numbering registry. Zx Online holds its address space and autonomous system directly from APNIC, and APNIC charges by formula. For holdings of this company's size — three registered IPv4 /24 blocks and a sliver of IPv6 — the published fee schedule works out to roughly AUD 2,000 a year, call it 160,000 taka. That number, computed here from the registry's formula rather than from an invoice, deserves a pause: the international registry costs this company more each year than its national regulator does. Sum the visible lines — regulator 140,000, registry about 160,000, plus trade licence, audit, VAT administration and the accountant who keeps it all reconciled — and the standing cost of formality lands plausibly around 400,000 to 500,000 taka a year, of which the licence proper is barely a third. The question the rest of the ledger must answer is what that half-million buys.

What formality sells: the fifty-taka megabit

Start with what it buys upstream, because the purchasing rights are the licence's hard core. Bangladesh's bandwidth market is a chain of licensed monopsonies: international bandwidth flows through licensed international gateways, domestic transmission through licensed nationwide transmission networks, and the regulator's August 2021 tariff order — effective 1 September 2021 for five years — prices what a licensed ISP pays each of them. Only a licensee sits at that counter. The routing table shows Zx Online sitting there twice: its autonomous system announces two /24 prefixes reached exclusively through two Bangladeshi gateway operators, Windstream Communication and Exabyte — a modest but genuinely multihomed arrangement that matches the redundancy language on its site. The same paper trail is what lets the company hold those addresses at all: APNIC contracts with legal entities, not with a person who owns switches in a stairwell.

Downstream, the licence's yield is the right to sell promises, and the price card separates the promises from the plumbing with rare precision. The home tiers — 630 taka for 30 megabits up to 2,625 for 250 — sell shared capacity at 10.5 to 21 taka per megabit, with a 1,000 taka installation charge and, on every tier, 100 megabits of domestic and cache traffic for video and local exchange content. The small-business tiers sell the same network with the promises attached: 3,000 taka for 60 dedicated megabits up to 7,500 for 150, exactly 50 taka per megabit at every step, installation 2,000, with the site promising a 99.9 percent service-level guarantee, symmetric uncontended bandwidth, public IP addresses and a named support engineer. One 100-megabit business account books the revenue of nearly five 100-megabit households. The megabits are the same megabits. What the customer pays 4.8 times more for is the contract — and only an entity with a licence number, a VAT registration and audited books can sign it. The company says those customers are banks, hospitals and factories served since 2008; that claim is the company's own, unverified, and the honest reading is narrower: the visible corporate product exists, priced and specified, and it is the only line on the card that can be paying for the paperwork.

The buyer's side of that corporate counter enforces the paperwork gate as strictly as the seller's. A scheduled bank's vendor enlistment, a hospital group's procurement office or a government tender all demand the same folder — licence copy, trade licence, VAT registration, tax certificate, audited statements — before a connectivity contract can even be evaluated, and the folder is precisely what the informal competitor cannot produce at any price. Zx Online's site adds one more small monetisation of its paper: residential customers may rent a real public IP address for what it calls a nominal monthly fee, a product that exists only because the company holds registry-delegated address space in its own corporate name. Formality, in other words, is not one gate but a nested series — and each layer of the series shows up somewhere on the price card as a line the unlicensed cannot copy.

The households, meanwhile, explain the ceiling. A residential subscriber in Mohammadpur faces an installation charge of 1,000 taka and no contractual lock-in worth the name; switching to any of the dozens of operators wiring the same lanes costs one technician visit. Retail churn discipline in this market comes not from contracts but from the cap: since the July 2025 revision the state prices entry broadband at 400 taka for 5 megabits, 700 for 10 and 1,100 for 20, and after the trade association's April 2025 pledge the 500-taka tier doubled to 10 megabits nationwide. A capped, homogeneous, switchable product is nobody's pricing power. The corporate tier, uncapped and paper-gated, is where the licence collects.

The squeeze between two price caps

Now assemble the unit economics, stating plainly which numbers are documents and which are assumptions. The documents: Zx Online's published card (630 taka entry, 1,050 for 100 megabits shared, 5,000 for 100 dedicated, all inclusive of the 5 percent VAT pattern on home tiers). The state's retail floor-and-ceiling structure just described. On the input side, the 2021 wholesale schedule set gateway bandwidth for a Dhaka ISP at 365 taka per megabit per month at the smallest volumes, falling to about 330 at ten gigabits, and transmission at 200 to 300 taka per megabit for small metro volumes down to the low teens at hundred-gigabit scale; in mid-2025 the government cut wholesale another 15 to 25 percent at the gateway level. And the regulator writes the sharing rule into tariff approvals: when it approved a fresh card for Sam Online — entry number 2 on the same divisional list as Zx Online — in April 2026, it specified 30 megabits at a 500-taka maximum, 100 at 1,000, 250 at 3,000, all at a contention ratio of 1:8.

Put those documents against each other and the shared-tier arithmetic looks impossible at list wholesale prices. A 100-megabit home plan at 1:8 contention implies 12.5 megabits of provisioned capacity per subscriber; even at the deepest 2021 gateway slab that is more than 4,000 taka of international bandwidth against 1,000 taka of ex-VAT revenue. The gap is closed by two things the caps do not capture. First, market wholesale prices sit under the ceilings — the ceilings are what the schedule caps, not what volume buyers pay, and the 2025 cut ratified a fall that had already happened; any figure for what Zx Online actually pays its two gateways would be an estimate, and none is offered here. Second, and decisively, most bytes never cross the gateway. The 100-megabit domestic-and-cache lane advertised on every Zx Online tier is the economics speaking: video platforms' cache servers and the domestic exchange fabric serve the overwhelming share of consumer traffic at near-zero marginal cost, so the scarce input is not the megabit but the international sliver of it. On plausible assumptions — one to two megabits of true gateway demand per busy-hour household, wholesale somewhat under the capped slabs — the entry-tier household yields a gross margin in the low hundreds of taka before staff, fibre repair and electricity, which is thin, while the 100-megabit household at 1,050 yields more absolute margin for barely more gateway demand: the card is built to migrate customers upward at a flat 10.5 taka per megabit. Those consumption figures are inferences and flagged as such; every price in this passage is a document.

Two smaller lines complete the retail arithmetic. The 1,000-taka installation charge cannot cover a fibre drop, an optical terminal and a technician's afternoon — hardware alone plausibly exceeds it, though no primary price for the kit is on record and the point is stated as inference — so each new household begins life as a small negative balance recovered over the first months of billing, which is why churn in a capped market quietly destroys value even when the subscriber count holds steady. And beneath the bandwidth sits the transmission line item: where an operator leases metro capacity rather than owning glass, the 2021 schedule prices small-volume metro transmission at 200 to 300 taka per megabit falling steeply with scale, a cost curve that punishes exactly the small books and rewards exactly the consolidators. The fixed block — the network operations centre the company says runs around the clock, the support staff, the office rent across three successive addresses — spreads over however many subscribers there are, which is the next section's problem.

The dedicated tier transforms the same inputs. A 100-megabit uncontended account books 5,000 taka against a worst-case input of 100 megabits of genuinely reserved capacity; even at capped wholesale the account clears its bandwidth bill, and at realistic volume prices it is the profit centre the home book cannot be. This is the formality dividend in a single comparison: the licence's fixed costs — the 140,000-taka regulatory year, the registry, the audit — spread invisibly across a few thousand capped households, but they are recovered many times over by each corporate contract that only a licensed, incorporated counterparty can win. At 3,000 subscribers (an assumption used for scale, not a claim), the entire regulatory stack costs about four taka per subscriber per month. The licence is cheap. What is expensive is everything the licence forces the company to be — audited, addressed, taxable — and what is valuable is the small book of customers who will only buy from exactly such a company.

Sizing a business that publishes no accounts

No revenue figure for Zx Online exists in the public record, so any estimate must be built and labelled. Two independent routes are available, and they roughly agree. The first divides the regulator's market by its own licence count. The commission's subscriber table records 14.46 million fixed internet subscriptions as of July 2025, up from 10.05 million in August 2021; the association that organises the industry lists 2,196 member firms. The crude mean is about 6,600 subscriptions per member, but the distribution is violently skewed by a handful of nationwide operators with six-figure books, so the plausible median for a three-neighbourhood divisional operator sits far lower — a band of one to five thousand connections is consistent with the shape of the market, and no better than that.

The first route also carries a warning about direction. The fixed market grew 44 percent between August 2021 and July 2025 on the regulator's own table; a company whose coverage claim spans three neighbourhoods it has served for years is, absent expansion, a shrinking share of a growing market, and the 2024-dated corporate relaunch on its site reads like a response to exactly that arithmetic.

The second route reads the company's own registry footprint. Zx Online announces 512 public IPv4 addresses and holds route objects for 768; in a market where household connections live behind address translation and public addresses are sold as add-ons and burned on infrastructure and corporate accounts, holdings of that size are the signature of a small operator — bigger books hold multiple times more. The 2009 capture showing eight reseller points-of-presence, and today's coverage claim of three contiguous neighbourhoods, point the same way. Put the two routes together and a defensible band emerges: low thousands of subscriptions, blended monthly revenue per account somewhere between the 630-taka entry tier and the corporate book's pull, hence annual revenue on the order of 15 to 50 million taka — roughly US$120,000 to US$400,000. Both routes are inference stacked on observation; the band is stated only because the two stacks, built from different materials, land in the same place. Against that band, the proposed 6.5 percent of gross that the next licensing regime would claim (of which more below) is one to three million taka a year — an order of magnitude more than the licence costs today.

The unofficial record, read closely

The signals outside the filings sketch the company's temperament. The 2009 boast — "Largest ISP in Bangladesh," floating above lorem ipsum — was false the day it was typed, but the list beneath it was true and more interesting: eight named cyber cafés and stairwell networks buying bandwidth from a one-year-old company, which is to say that Zx Online began as a wholesaler to the informal sector it had just formally left. Its licence let it buy at the licensed counter and resell to operators who could not. That intermediary posture — formality as arbitrage — is the oldest business model the licence supports, and the company's own 2020 homepage still advertised it, boasting of sign-ups with "leading local internet distributors/cyber cafés."

The association record adds a seniority signal the filings understate. Membership number G-074 makes Zx Online roughly the seventy-fourth general member of a body that now counts over two thousand — the early-adopter cohort of Bangladeshi internet formality, joined the same month the licence was issued. Longevity of that kind is itself information in a sector where licences churn: whatever else the company has neglected, it has paid the association and the regulator through eighteen years, two address moves and, on its own telling, a rebuild of its plant into the fibre backbone both the 2016 and current sites describe.

The same archive shows the cost of standing still. The homepage captured in January 2020 is byte-for-byte the story of 2016: the newest news item is dated February 2013, the copyright line reads 2009–2016, and the navigation still leads with FTP servers, antivirus resale and web design — a site administered but not authored for half a decade, spanning precisely the years when Bangladesh's fixed-broadband base grew from a few million toward ten. The relaunch visible today — a modern single-page site whose own timeline jumps from 2016 to 2024, when it says dedicated SME and corporate tiers launched — reads as a company reawakened by exactly the corporate product this article prices. Meanwhile the housekeeping stayed uneven: the association profile missing its addresses and directors; the alias domain surrendered to a betting-page squatter in May 2025; the FTP brand drop-caught in January 2026. None of this is distress evidence — the licence renews, the routing table answers, the IRT object was touched six weeks ago — but it is the profile of a firm that maintains what the licence requires and neglects what it does not. What would settle the picture: a completed association profile naming directors, a corporate reference from a named bank or hospital client, or the company's tariff appearing in the regulator's approval notices the way its list-neighbour Sam Online's did in April 2026.

A crowded market on both sides of the wrapper

Whatever moat the licence digs, it is not exclusivity. The divisional summary that carries Zx Online at entry 316 carries 141 licensees in Dhaka Division alone as of September 2023, inside a national stack of 3,573 telecom licences across 27 categories. Above the divisional class sit the nationwide operators — the Link3s and Dot Internets whose sales teams work the same Mohammadpur lanes; below it, upazila licensees and the surviving informal networks compete for the same 500-taka household; beside it, 121.5 million mobile internet subscriptions mean every marginal fixed customer already carries a substitute in a pocket. Price is regulated, the product is homogeneous, and switching costs one visit. Competition therefore migrates into the seams: cache and domestic throughput (hence the 100-megabit local lane on every tier), installation speed (the site promises 24 to 48 hours), support theatre, and — for the only customers with real switching costs — the corporate relationship, where a bank that has certified one vendor's paperwork does not casually re-run diligence questions on another's. The supplier side is concentrated in the other direction: two gateway upstreams visible in the routing table, a transmission market of a few nationwide networks, and a gateway sector fragile enough that the regulator reported 29 of its members owing it 2.05 billion taka. A small ISP in Dhaka is a price-taker at both ends of its business; the licence buys it the right to stand between two administered prices and keep the difference.

When the flat fee becomes a revenue tax

The forward risk is that the state is repricing the wrapper itself. In April 2025 the commission unveiled a plan to collapse its 27 licence categories into three service classes, with existing gateway-era licences expiring mostly by 2027 and small ISPs pushed toward a lighter enlistment track. The draft fixed-telecom framework that followed proposes that broadband operators share 5.5 percent of annual revenue with the regulator and pay 1 percent more into a social obligation fund — terms mobile operators have long lived with and fixed ISPs never have. The industry association's arithmetic says retail prices would need to rise about 18.4 percent to carry the levy, in a market where the same state caps retail prices — an equation with no solution that does not compress somebody's margin. For Zx Online the stakes are precise: its current regulatory cost is a flat 140,000 taka a year regardless of revenue; a 6.5 percent gross levy on even the low end of the revenue band estimated above would multiply that several-fold, and it would fall hardest on exactly the corporate book that pays for formality today. A licence bought as property in 2008 — a one-time admission ticket with a small annual toll — would become a franchise taxed on turnover. The old regime made the suffix cheap to keep; the proposed one prices it like a partnership with the state.

The ladder logic beneath the proposal deserves a moment, because it explains why the divisional class was the shrewd rung to hold. A nationwide licence costs two and a half times the divisional annual fee for coverage rights a three-neighbourhood operator cannot use; a district licence saves half the fee but caps the map; the upazila class beneath is cheap enough to have formalised the informal sector wholesale, flooding the market the divisional tier once had to itself. The 2025 policy paper — published that September and archived before the regulator's portal reshuffled its files — would flatten this ladder into broad service classes with an enlistment track for the smallest operators, which cuts both ways for a firm like Zx Online: enlistment would strip the small competition of licence property while pushing licensed mid-tier firms toward obligations sized for companies many times larger. Where the current guideline treats a divisional licensee as a fee-payer, the draft treats a fixed-telecom licensee as a taxable franchise with rollout duties. Between those two readings sits most of the value of an eighteen-year-old piece of paper.

The regime risk is not hypothetical for operators who lived through recent years. In July 2024 the previous government severed the country's connectivity outright for days during the uprising — broadband as well as mobile — demonstrating that a licence is also an obedience instrument: the network exists at the state's sufferance, revenue stops when the state says so, and the compensation ladder in the 2025 tariff order (bills halved after five days of outage, zeroed after fifteen) applies to a provider's failures, not the government's. Add the January 2025 two-week supplementary-duty whiplash, the five-yearly renewal gate, and a licence that can be cancelled for late fees, and the shape of the asset clarifies: the divisional licence is a real and cheap asset under the current rules, a costlier and more conditional one under the rules now in consultation.

What would change this reading

A handful of documents would move this assessment, in either direction. The registrar's file, if the portal ever answers, is first: paid-up capital, shareholders and filed accounts would replace the revenue band's two triangulations with a number, and would confirm or complicate the single-family reading suggested by a chairman-addressed inbox. Second, a Zx Online entry in the commission's tariff-approval notices — as its list-neighbour received in April 2026 — would establish the regulated ceilings for its own card and show whether the 630-taka entry tier is compliant pricing or a premium the market tolerates. Third, the fate of the draft licensing framework: enactment of the 5.5-plus-1 percent levy as drafted would convert the thesis of this piece from "formality is cheap and yields a 4.8x corporate premium" to "formality is a tax whose payers must consolidate," and a small divisional book would be an acquisition target rather than an annuity. Fourth, the 2028 renewal window: a renewal on schedule would confirm the wrapper's maintenance; a lapse, after the pattern of the lapsed domains, would say the owners had concluded the arithmetic no longer closes. And fifth, any transaction — a sale of the company, its customer list or its licence position — would do what no analysis can: put an observed market price on the suffix itself. Until one of those documents surfaces, the record supports a modest, specific conclusion: an eighteen-year-old licence, continuously held, cheaply kept, and earning its keep on the one tariff line the informal market cannot copy.

Evidence register