Tornado Networks and the rupee price of a megabit built in dollars

Two hundred and fifty-five rupees to the dollar

On 26 January 2023 the Pakistani rupee fell 9.61 per cent against the dollar in a single trading session, closing at 255.43 in the interbank market, as the central bank abandoned an unofficial peg to satisfy the International Monetary Fund. CNN reported it as the currency's worst day in over two decades. It was not close to the bottom. By 5 September 2023 the interbank rate touched 307.1, and the rupee closed that year at 281.86 after starting it at 226.43. Stretch the window back to May 2021, when a dollar cost about 152 rupees, and the arithmetic becomes brutal: in twenty-eight months, the rupee price of anything invoiced in dollars roughly doubled.

In Pakistan, a megabit is invoiced in dollars. The telecom regulator said so in plain terms in a November 2023 consultation on wholesale bandwidth charging: wholesale internet bandwidth "is denominated / charged in US Dollars (USD) and payments are made in Pak Rupees (PKR) equivalent on the date of payment," so that "whenever Pak Rupee depreciates versus USD, the cost of Internet bandwidth for wholesale customers automatically increases with the same proportion." The same paper put the depreciation from January 2020 to August 2023 at 83 per cent — roughly 17 per cent a year — against average inflation of about 16 per cent. Every internet provider in the country sits inside that clamp: dollar costs on one side, rupee revenues on the other, and a customer base whose incomes fell in real terms throughout.

Most analysis of that squeeze stays at the top of the market, with the mobile carriers and the two submarine cable operators. This piece goes to the bottom of it. Tornado Networks (Pvt.) Limited operates a small fixed-wireless network in Bahawalnagar, a cotton-and-wheat district of 3.55 million people on Punjab's border with India. It holds a fifteen-year federal licence, an autonomous system with exactly 512 public IP addresses, and a published price list that runs from 999 rupees a month for one megabit to 5,999 rupees for eight. Between the archived copy of that price list from March 2023 and the copy from December 2025, not one figure changed — through the worst currency crisis in Pakistan's history, two IMF programmes and a near-doubling of the national electricity tariff. Either the company repriced everything except its website, or it absorbed the squeeze in margin because its market could not bear the pass-through. Both possibilities are worth pricing, and the evidence lets us get surprisingly close.

A licence, a street address and an autonomous system

The identity record is thin but coherent, and it is worth assembling before the economics, because the directory entry for this company rests on a live domain and little else. The legal name is Tornado Networks (Pvt.) Limited. The Pakistan Telecommunication Authority's register of Data Class Value Added Services licensees records the company as licensee number DIR(L)/CVAS-723/PTA/2015, granted on 17 November 2015 for internet services in Punjab, with its chief executive named as Mamoon ur Rasheed and its address as Street 12, Mohallah Nizam Pura Gharbi, Bahawalnagar. The January 2024 edition of the same register shows the licence running to 16 November 2030. The official website, tornado.com.pk, gives the same street, the same phone numbers with Bahawalnagar's 063 area code, and describes the firm as a wireless internet service provider working in point-to-point and point-to-multipoint radio.

The routing record matches. APNIC's registry shows autonomous system 139718, registered on 10 October 2019 to Tornado Networks (Pvt.) Limited, with contact addresses at tornado.com.pk in the name of Mamoon — the licence-holder's first name — and two blocks of IPv4 space, 103.144.10.0/24 and 103.144.11.0/24, registered the following day and announced continuously since. Traffic on the network shows the day-night rhythm of a residential eyeball network, per ipinfo's measurement, and the route origination is signed and valid. A PeeringDB record appeared on 1 April 2025: open peering policy, no exchange points, no listed facilities. The web history goes back further than the ASN. A September 2016 capture of the domain shows nothing but a support-ticket portal and an administrator login — a working ISP back office within a year of the licence grant, years before the company had its own address space.

Two clarifications matter. First, "Tornado" is also the name of a wireless broadband brand that COMSATS Internet Services launched in 2005; that is a separate, older, state-linked operator, and nothing connects it to the Bahawalnagar company beyond the name. Second, this is not an urban operator. Directory assumptions about Pakistani ISPs default to Karachi and Lahore; the evidence here points to a district town of under two hundred thousand people in a division where urban residents are 27 per cent of the population and roughly half the workforce farms. That location is not a footnote. It is the whole economic problem.

What a thousand rupees buys on Police Line Road

The product set, as the company's own site presents it, is small-town connectivity in full: home broadband over fixed wireless, television bundles of 150-plus and 200-plus channels, CCTV camera installation, and biometric attendance systems for schools and businesses. Local directory listings describe the firm as a wireless specialist working with Ubiquiti, MikroTik, D-Link, TP-Link and Alfa equipment — the standard toolkit of the subcontinental WISP — and the site names Wateen, Jazz, Ufone and PTCL as technology partners, which in this tier of the market usually means backhaul circuits and reseller arrangements rather than alliances. The published price card is five rows long: 1 Mbps for Rs 999 a month, 2 Mbps for Rs 1,999, 4 Mbps for Rs 2,999, 6 Mbps for Rs 3,999, 8 Mbps for Rs 5,999, all sold as unlimited data with round-the-clock support.

Read those numbers against the national market and the customer becomes visible. At interbank rates near 285 rupees to the dollar in mid-2026, the entry package costs about $3.50 a month and the top of the card about $21. The buyers are households that want steady video calling to relatives in the Gulf, shopkeepers running point-of-sale and CCTV, schools, ginning mills, and the district's small professional class. Pakistan has only about 3.6 million fixed broadband connections against 207 million total telecom subscriptions; in a district like Bahawalnagar the fixed-line share is thinner still, which means every customer is won from the real competitor — the mobile 4G bundle — and lost back to it the moment fixed service stops being visibly better per rupee.

What Tornado sells, at bottom, is contention engineering: it buys a trunk of expensive dollar bandwidth, splits it across radio sectors, and prices the split in rupees. One megabit at Rs 999 sounds archaic beside urban fibre, and it is — PTCL's Flash Fiber sells ten megabits for Rs 1,999 where its optical network reaches. But a wireless CPE on a rooftop reaches streets that fibre build-out economics will not touch for years, and an operator on Police Line Road answers its phone in Punjabi and sends a technician on a motorbike the same afternoon. The premium is real, and it is paid for proximity, not speed.

The dollar side of the ledger

Set against those rupee revenues, nearly every material cost line has a dollar inside it. Start with bandwidth, the largest recurring input. Until August 2023 Pakistan's entire international capacity was wholesaled by two firms, PTCL and Transworld Associates, both declared dominant in the wholesale IP bandwidth market by the regulator in July 2021; a third entrant, Cybernet's PEACE cable landing, began selling onward capacity only in late 2023. Those wholesalers pay their upstream counterparties in dollars — operations and maintenance, port charges, circuit charges — and invoice downstream operators in dollars payable in rupees at the day's rate. A district WISP sits at the end of that chain and eats every basis point of depreciation with no hedge available. When the rupee moved from 152 to 307, the rupee cost of an unchanged bandwidth commitment doubled, mechanically, before any negotiation.

Equipment is the second dollar line. The radios, sector antennas and customer-premises units that constitute a wireless network are imported, and their rupee prices track the exchange rate at import-parity: a Ubiquiti LiteBeam M5, the workhorse CPE of Pakistani WISPs, retails at about Rs 16,700 — almost exactly its dollar list price converted at the interbank rate, plus duties. Worse, for eighteen months the problem was not price but permission. From mid-2022 the State Bank rationed foreign exchange by slowing and refusing import letters of credit; banks were turning away importers by June 2022, and sector analyses of the period record telecom operators and tower companies facing outright import bans that slowed network builds nationwide. A company Tornado's size does not open its own letters of credit; it buys from Karachi and Lahore distributors whose shelves emptied and whose prices ratcheted with each clearing backlog. Growth, for a WISP in 2022 and 2023, was rationed at the port.

The third line is power, which in Pakistan is a second currency exposure wearing a domestic disguise. Grid tariffs are driven by imported fuel and by capacity payments to independent producers that are largely dollar-indexed, so IMF-mandated cost recovery turned the exchange rate into an electricity bill: the national base tariff rose Rs 5.72 in one step to Rs 35.50 per kilowatt-hour from July 2024, with upper residential slabs near Rs 49. A wireless network is a distributed electrical load — every tower site, every relay, every switch cabinet draws continuously — and in rural Punjab it draws through scheduled and unscheduled load-shedding, which means batteries, and increasingly solar panels, each replacement cell and panel imported and dollar-priced. Bandwidth, hardware, power: three cost lines, one exchange rate.

A rate card frozen in a falling currency

Against that cost stack, hold up the revenue record. The Wayback Machine's captures of tornado.com.pk show the same five-row price card in March 2023, weeks after the January devaluation, and in December 2025, and on the live site today. Rs 999 for a megabit was $4.40 when 2023 opened at 226 to the dollar, $3.25 at the September 2023 trough of 307, and $3.50 at mid-2026 rates. Meanwhile domestic inflation compounded at rates that touched 38 per cent year-on-year in May 2023. In real rupees, Tornado's published prices fell by more than a third across the window. In dollars — the currency of its costs — they fell by a fifth. No line of this analysis matters more than that one, so it deserves precision about what it can and cannot prove.

It cannot prove that street prices never moved. Small Pakistani ISPs transact in cash, negotiate door-to-door, and treat the website as a brochure; the same captures show the company's about page still carrying the web template's Latin filler text four years after launch, and the site's footer still crediting a developer in Kotli with a 2022 copyright. A rate card that stale could simply be unmaintained while actual invoices climbed. But the surrounding market says otherwise. PTCL froze and only modestly revised its own broadband tariffs through the crisis for the same reason every operator did: the customer's wallet is denominated in rupees too, and in a district where half of household labour is agricultural and literacy is 57 per cent, a broadband subscription is among the first line items cut when flour and electricity double. Pakistani retail bandwidth is one of the cheapest in the world per megabit not because costs are low but because willingness to pay is the binding constraint. The squeeze is real even if the card is nominal: whatever the street price did, it did not track the dollar, because it could not.

There is also a quieter tell in what the card sells. One to eight megabits, in 2026, is a pre-fibre product ladder. The site's boilerplate claims speeds "up to 100+ Mbps," but the priced offer stops at 8 Mbps — a shape that suggests a network still built on 5 GHz unlicensed radio with limited backhaul, where every additional sold megabit has a visible marginal cost. Compare Login.Me, the Faisalabad-region operator that (as shown below) sells Tornado its transit: its own retail card starts at 15 Mbps for Rs 1,649. Tornado's price per delivered megabit is five to ten times its upstream's retail rate. That differential is the arithmetic signature of a capacity-constrained wireless operator in a thin market — and it is exactly the position most exposed when the input currency halves.

The arithmetic of a small wireless network

Assemble the unit economics from what the record gives. The evidence: a five-tier rupee price card; 512 public IPv4 addresses; a CPE that costs Rs 16,700 retail; wholesale bandwidth denominated in dollars; an electricity base tariff of Rs 35.50 per unit; a fifteen-year licence with modest fees; and a claimed customer base of home users plus corporate clients. The inference, flagged as such: subscriber count, contention ratio and staffing are estimated, not reported, because no filing discloses them.

A network with two /24 blocks and carrier-grade address translation comfortably serves one to four thousand subscribers; the residential traffic rhythm on the ASN and the size of the town make a working estimate of 1,500-2,500 connections defensible, and anything above 5,000 would likely have left a bigger routing and hiring footprint. Take 2,000 subscribers at a blended Rs 1,600 a month — weighted toward the Rs 999 and Rs 1,999 tiers — and gross revenue is about Rs 3.2 million a month, or roughly $11,000. Sold capacity at an average 2 Mbps per subscriber is 4,000 Mbps; at a typical 1:10 contention the trunk requirement is about 400 Mbps. Small-volume transit delivered upcountry in Pakistan plausibly costs $2-4 per Mbps per month once Karachi wholesale, national transport and last-mile circuits are stacked — inference again, anchored by the regulator's observation that international prices are falling while the rupee cost rises. That puts bandwidth at $800-1,600 a month — Rs 230,000 to Rs 460,000, or seven to fifteen per cent of revenue — before a watt is billed or a salary paid, and with every rupee of it repricing on the day the exchange rate moves. Power across a headend and a dozen relay sites might run Rs 200,000-400,000 a month at current tariffs with battery replacement amortised; a lean staff of eight to twelve — riggers, support, recovery men — perhaps Rs 500,000-800,000. Net those lines out, subtract rent, vehicles, fuel, television content fees and the ordinary leakage of a cash-collection business, and what remains is plausibly Rs 1 million to Rs 2 million a month before capital spending: real money in Bahawalnagar, invisible at national scale, and thin enough that a 20 per cent currency move eats a large bite of it through the bandwidth line alone.

The capital line is where the currency bites hardest. Connecting one new subscriber requires a CPE, mounting hardware and a technician-day: call it Rs 20,000 all-in, of which perhaps 80 per cent is imported kit. Against the Rs 999 entry tariff that is a twenty-month payback on hardware alone; against the blended Rs 1,600 it is a year. Every rupee of depreciation lengthens that payback because the CPE repriced overnight while the tariff could not. This is why the LC crisis was existential arithmetic rather than an inconvenience: a WISP that cannot buy radios at pre-crisis prices simply stops adding subscribers, and a WISP that stops adding subscribers in a market with churn to 4G is shrinking. The rational responses are visible in the record — sweat the installed base, stretch equipment life, hold nominal prices to protect volume, and defer every dollar of capex the network can survive without. A frozen rate card is what that strategy looks like from outside.

Five upstreams in six years

If margin defence is invisible in prices, it is visible in routing. The history of AS139718's transit relationships reads like a procurement diary. From October 2019 to May 2020 the network bought transit from LINKdotNET Telecom, a legacy national ISP. From mid-2020 to December 2021 it moved to Redtone Telecommunications Pakistan. From December 2021 it bought from PTCL itself, the incumbent. From November 2023 it added Login.Me (Pvt) Ltd, a fast-growing operator out of Jaranwala in Faisalabad division — a regional wholesaler two tiers below the submarine cable operators — which took over as primary transit while the PTCL relationship wound down through 2024. And from March 2025 a second leg appears: CMPak, the China Mobile subsidiary that operates Zong, via its enterprise network — alongside, for the first time, a downstream of Tornado's own, Shahram Telecom Private Limited, a newly licensed local-loop operator with a footprint in Haroonabad, forty minutes down the road.

Each hop tells an economic story. Turning to a Faisalabad wholesaler in late 2023 — weeks after the rupee bottomed at 307 — is what an operator does when the dollar-linked invoice from the incumbent stops being payable and a regional aggregator offers a sharper rate by bundling many small WISPs into one wholesale purchase. Adding a Zong enterprise circuit in 2025 buys redundancy against the single-homed fragility that a one-upstream network lives with, and hints at fibre backhaul reaching the district on infrastructure the state subsidised: the Universal Service Fund's build-out lists ten optical fibre node locations across Bahawalnagar district's tehsils, from Chishtian to Fort Abbas to Dunga Bunga. And taking on Shahram Telecom as a paying downstream inverts the squeeze one notch: Tornado, the price-taker in every dollar negotiation, becomes a price-maker in rupees to an operator even smaller than itself. The wholesale chain in Pakistani connectivity is a ladder of exactly such relationships — Karachi sells to Faisalabad, Faisalabad sells to Bahawalnagar, Bahawalnagar sells to Haroonabad — and at every rung the dollar content of the invoice falls while the service content rises. Climbing one rung is the single most effective hedge available to a company like this, because reselling to a neighbour converts dollar bandwidth into rupee wholesale revenue at a spread Tornado controls.

The dependency picture that remains is narrow but no longer trivial. Upstream, the company depends on one regional wholesaler for most of its route table and one mobile carrier's enterprise arm for the rest; either relationship repricing with the exchange rate would land directly on margin. Downstream, one small operator now depends on Tornado the same way. The 512 public addresses, held and announced without interruption since 2019 and signed valid under RPKI, are a quietly appreciating dollar-denominated asset in their own right: IPv4 blocks trade globally at prices that would make Tornado's two /24s worth several years of its estimated profit, a fact that functions as an unusual form of balance-sheet insurance for a company whose other assets are depreciating radios on rooftops.

Competing with the incumbent that sells you transit

Competition in Bahawalnagar is a three-front affair, and on two of the fronts the competitor is also a supplier. The first front is the incumbent. PTCL's copper reaches the district as it reaches everywhere, and its fibre-to-the-home brand lists Bahawalnagar among its coverage cities; where the optical build actually passes a street, Rs 1,999 buys ten megabits against Tornado's two. Fibre coverage in tertiary cities is patchy — a few feeder routes and colonies rather than blanket build — but every street it reaches converts the wireless value proposition from "the fast option" to "the flexible option," and the price umbrella under which a WISP charges Rs 2,999 for four megabits folds accordingly. That PTCL was simultaneously Tornado's transit provider for three years is the ordinary, uncomfortable geometry of this market: the incumbent wholesales to the same small operators it retails against.

The second front is mobile substitution, which in Pakistan is less a front than a flood. The country has 158 million mobile broadband users against 3.6 million fixed connections; a Jazz or Zong 4G package delivers usable video for a few hundred rupees, and for a large share of Bahawalnagar households the question is not which fixed provider to choose but whether to have fixed service at all. Mobile data is the reservation price for everything Tornado sells. It caps the rate card far more effectively than any regulator, and it explains the card's shape: unlimited data at low headline speeds is precisely the product that a capped 4G bundle cannot match, sold to the household that has exhausted its monthly gigabytes by the third week.

The third front is horizontal: other WISPs and cable operators in the same streets, and the regional wholesaler one district over. Login.Me's own retail card — 15 Mbps at Rs 1,649 — shows what happens when a Faisalabad-tier operator pushes fibre into a town: the wireless price ladder collapses. Nothing in the public record shows Login.Me retailing in Bahawalnagar today, and its transit relationship with Tornado argues against near-term invasion; wholesalers rarely eat their customers while the customer is still growing. But the PTA's licence register lists hundreds of Data CVAS licensees across Punjab, entry costs a few hundred thousand rupees, and the moat around a district WISP is made of rooftop access, installed CPEs and personal trust rather than anything a competitor cannot replicate with eighteen months and a spectrum-free radio plan. Switching costs for customers are one visit from a rival's technician. What keeps churn low in practice is the same thing that keeps prices sticky: in a cash economy, the provider who knows your house, extends a week's grace on the bill, and shows up during a canal-season storm is not lightly abandoned for ninety-nine rupees of savings.

Power, taxes and the second squeeze

Two further mechanisms transmit the macro crisis into this business, and both arrive through the bill rather than the exchange ticker. The first is electricity. IMF-driven tariff resets took the national base rate from Rs 27.78 to Rs 35.50 per unit in July 2024, and commercial connections of the kind an ISP's sites use sit above the base. For the company this is a direct cost line across every powered site. For its customers it is a budget shock that competes with the broadband subscription: the same July 2024 reset raised the bill of a modest urban household by thousands of rupees a month. When the electricity bill doubles, the Rs 999 subscription survives only by being the cheapest entertainment and communication bundle in the house — which, against a cinema ticket, a cable bundle and four SIM cards of mobile data, it arguably is. The perverse effect is that grid failure sells fixed broadband's complements too: a household that installs solar and batteries to ride through load-shedding suddenly has uninterrupted power for a router, and a WISP whose towers stay lit through an outage is selling reliability the grid cannot.

The second mechanism is taxation, where the state's fiscal desperation lands directly on the retail price. Every telephone and internet bill in Pakistan carries a 15 per cent advance income tax withheld at source, and the Finance Bill 2024 raised the rate to 75 per cent for non-filers named on the tax authority's enforcement lists. In a district with 57 per cent literacy and a workforce that is half agricultural, registered tax filers are a small minority; the practical effect is that the state adds between a seventh and three-quarters on top of every rupee Tornado invoices, on rates that current tax tables confirm. A frozen rate card is therefore not even a frozen price: the delivered price of connectivity in Bahawalnagar rose by government action while the operator's share of it stood still. Between the exchange rate taking the cost side and the withholding regime taking the demand side, the band of viable prices for a small ISP has narrowed from both ends through the entire crisis.

When the country's internet breaks upstream

Beyond cost, a district ISP carries risks it can neither price nor hedge, because they originate two rungs up the ladder or in Islamabad. Pakistan's international connectivity runs through seven submarine cable systems, and when two of them failed in the summer of 2024 — a fault on SMW-4 near Karachi and a rerouting outage on AAE-1 — the regulator itself announced that the whole country would live with degraded service into October. A Bahawalnagar customer who bought eight megabits from Tornado got whatever fraction of that survived the national congestion, and the refund conversation happened at a shop counter on Police Line Road, not at a cable landing station. The same summer, businesses and providers blamed a state-installed filtering apparatus for further slowdowns; monitoring groups priced Pakistan's 2024 outages and throttling at $1.62 billion in economic damage, among the highest in the world that year. None of this appears on a small operator's ledger, and all of it lands on its churn.

The political geography adds its own tail risks. Bahawalnagar is a border district; the state shuts mobile networks around security events with some regularity, and while fixed networks are rarely switched off deliberately, every national-level blackout teaches customers that connectivity in Pakistan is contingent — a lesson that suppresses willingness to prepay annual plans, the cheapest working capital a small ISP can raise. Licensing is the milder exposure: the Data CVAS regime that covers Tornado is light-touch and cheap, its licence runs to 16 November 2030 on the current register, and the regulator's enforcement attention concentrates on the carriers. The heavier hand is fiscal — the withholding rates and duty schedules that change every June with the federal budget — and monetary, where the company's entire cost base is one IMF review away from repricing. A firm this size does not manage such risks; it absorbs them, and its capacity to absorb is exactly the margin cushion estimated above, a few hundred thousand rupees a month standing between a currency event and a shrinking network.

What the quiet record says

The unofficial record around Tornado Networks is defined by its silences, and the silences are informative. The company's Facebook presence is a bare listing; local directory mirrors such as BizSouthAsia carry the same brief "wireless expert" copy with no accumulated customer reviews; no job advertisements surface under the company's name on any Pakistani jobs board; and the corporate website has carried its template's Latin placeholder text and a 2022 copyright for four years. The licence register reaches the company at a Gmail address while the routing registry reaches it at the founder's own mailbox. A crowd speed-test host page exists for the network, confirming that real users test it, though the archive behind it is not openly readable. One flag in the measurement record — ipinfo notes BitTorrent and VPN activity on some of the address space — is the ordinary signature of a residential eyeball network rather than evidence of anything amiss.

Read together, these signals suggest an owner-operated business of modest headcount that acquires customers through walk-ins, word of mouth and WhatsApp rather than marketing; that has never needed to hire publicly because staff come through personal networks; and that invests engineering attention in the network rather than the shopfront. The counter-reading — that the quiet marks a business in decline — is weakened by the routing record, which shows the opposite of decay: a new transit relationship in 2023, a second redundant upstream in 2025, a PeeringDB record created the same spring, and a downstream operator signed on. Companies winding down do not add suppliers, registries and customers of their own. What the quiet record cannot settle is scale: nothing public separates a 1,200-subscriber operation from a 4,000-subscriber one. A district-level count of active connections — the kind of figure the regulator collects but does not publish per licensee — would settle it, as would something as mundane as a hiring notice for a second support shift or a fibre-splicing crew.

What would change the judgement

The judgement, as it stands: Tornado Networks is a real, licensed, decade-old wireless operator of village-utility scale, run tightly by its founder, that has survived Pakistan's currency collapse by freezing rupee prices, sweating dollar assets, repeatedly re-shopping its transit downmarket, and beginning to climb the wholesale ladder itself — a rational, margin-thin adaptation that leaves it dependent on the exchange rate staying inside the 275-290 band it has held through the first half of 2026. Several discoveries would move that assessment materially.

Evidence of actual street pricing would come first. A single verified customer invoice from 2023 and another from 2026 would show whether the frozen web card is a price commitment or a dead page; if invoices climbed 40 per cent while the card stood still, the margin-compression thesis softens into an ordinary stale-marketing story and the demand-side reading of this piece weakens accordingly. Second, a subscriber count from any credible source — a regulator disclosure, a court filing, an acquisition memo — would collapse the widest error bars in the arithmetic above; below a thousand connections this is a family livelihood rather than a firm, and above five thousand the profit estimates roughly double. Third, the fibre map matters: confirmation that PTCL's optical build or Login.Me's expansion has passed meaningful street-kilometres inside Bahawalnagar city would start the clock on the wireless ladder's obsolescence, while confirmation that fibre stalled at the feeder routes would extend Tornado's pricing umbrella for years. Fourth, the currency itself: a renewed slide toward and past 307 rupees — the 2023 record — would reopen every dollar wound this piece documents, and the company's response (a repriced card, a fire-sale of an IP block, a lapsed transit session) would be visible in exactly the public records used here. And fifth, succession: everything registry-visible about this company routes through one person; a changed contact name in the APNIC objects or the licence register would be a small line in a database and a large fact about the company. On the other side of the ledger, signs of the state loosening the squeeze — the regulator forcing rupee-denominated wholesale bandwidth, as its own consultation contemplated, or import duties on network equipment falling — would improve the economics of every operator in this tier at a stroke, Tornado included.

Sources and signals

The evidence for this piece is public, and the load-bearing items are few enough to list with their function. On identity and licensing: the PTA Data CVAS licensee register of June 2020 and its January 2024 successor establish the legal name, the 2015 licence, its 2030 expiry, the chief executive's name and the Bahawalnagar address; the official site and its price page carry the current offer; the COMSATS page separates the unrelated older brand of the same name.

On the network: APNIC's registry record for AS139718 and the address-space record date the routing identity to October 2019; ipinfo's ASN profile documents the two /24 blocks, RPKI validity and residential traffic pattern; RIPEstat's neighbour history records the five-upstream procurement trail and the Shahram Telecom downstream, with ipinfo's page on that operator supplying its identity; the PeeringDB record dates the company's interconnection housekeeping to April 2025; the USF node list maps subsidised fibre backhaul into the district.

On the history: Wayback Machine captures from 2016, 2022, March 2023 and December 2025 anchor the operating timeline and the frozen rate card. On the macro squeeze: the PTA wholesale bandwidth consultation is the central document tying bandwidth costs to the dollar; Business Recorder and CNN date the rupee's fall; Profit's reporting on letters of credit and the TowerXchange country profile document the import rationing; Profit's tariff coverage and EY's Finance Bill alert carry the power and tax mechanics, with current withholding tables confirming rates.

On market context: ProPakistani's fibre package tables and the Flash Fiber coverage tool price the incumbent's competing offer; Login.Me's site prices the upstream's own retail; subscriber statistics size the fixed-broadband niche; census tabulations and the district profile describe the market's demography; equipment retail listings and 2026 exchange-rate tables supply the hardware and currency inputs to the arithmetic. The soft signals — the bare Facebook listing, the reviewless directory mirrors, the speed-test host page — are read in the body for what they suggest about scale and style, not as facts about performance. Subscriber numbers, street prices, staffing and margins in this piece are labelled estimates throughout; every dated fact traces to one of the records above.