The long game behind Malaysia’s 5G network

AT first glance, the government’s exit from Digital Nasional Bhd (DNB) looks like a quiet admission that Malaysia’s 5G experiment has gone wrong.

The single wholesale network (SWN) – later reshaped into a dual-network model – now leaves the country’s largest telcos facing a loss-making national platform.

For investors, the optics are uncomfortable: a policy-driven project shifting financial strain from the state to the balance sheets of CelcomDigi Bhd, Maxis Bhd and YTL Power International Bhd.

Each has two months from Dec 1, 2025 to pay RM327.87mil to acquire the Ministry of Finance Inc’s (MoF) stake in DNB following the exercise of a long-embedded put option.

The immediate concern is clear, earnings pressure, dividend resilience and valuation risk.

Inside DNB, executives argue this interpretation misses the bigger picture.

Far from signalling failure, they say the government’s exit marks the end of the riskiest phase of a deliberately front-loaded national infrastructure build and the beginning of the phase where scale, utilisation and economics start to align.

DNB chief executive officer Datuk Azman Ismail, chief corporate officer Datuk Ahmad Zaki Zahid and chief technology officer Ken Tan tell StarBiz 7 the network was never designed to deliver short-term profitability, nor was the government’s exit an admission of defeat.

Instead, Azman argues Malaysia is only now entering the most economically meaningful phase of 5G deployment, one that would have been far costlier, riskier and slower had telcos been left to build competing networks from the outset.

He points to the surge in adoption: from a 1% penetration rate in January 2023 to 84.2% by October 2025, translating into almost 29 million 5G subscriptions.

The core misunderstanding, Azman says, lies in how DNB’s mandate has been interpreted. The network was conceived as a state-led, supply-driven intervention to ensure Malaysia did not miss the 5G window, not as a commercial venture optimised for early profits.

Malaysia has limited mid-band spectrum, and international standards require a minimum 100MHz block for true 5G performance.

Fragmenting that spectrum across multiple mobile network operators, Azman argues, would weaken efficiency, raise device costs and risk networks reverting to enhanced 4G.

This is why the SWN was originally configured around 200MHz from day one, even though only half of that spectrum could initially be utilised. The network was built for scale before scale arrived.

When regulators later required DNB to operate with only 100MHz, the impact was immediate. Demand surged, but supply was capped.

Azman explains that congestion increased, revenues lagged potential and losses deepened, not because 5G failed to attract users, but because the network was prevented from monetising its full capacity.

That policy decision now sits at the heart of DNB’s defence against claims that it is structurally unviable.

With the reassignment of another contiguous 100MHz restoring the network to its original design, Azman argues many of the financial pressures were policy-induced rather than operational.

The pent-up demand is already visible. National 5G penetration has surged past 80%, among the fastest adoption curves globally, even though enterprise use cases have yet to be fully monetised.

“40% aggressive demand growth for 5G, that’s what we see, that’s very key, and as well as the assignment of the additional contiguous 100Mhz spectrum. That’s really where we’re now able to have the capacity to meet the demand,” he says.

Cost discipline

Cost discipline forms the second pillar of DNB’s counter-narrative.

Since inception, the company says it has reduced total cost of ownership by about RM4bil, bringing long-term network costs down from RM17.2bil to RM13.2bil without compromising coverage or performance.

Tan attributes this not just to scale, but to the advantage of building a greenfield, AI-driven network from day one – a structural edge legacy operators rarely enjoy.

The result, he says, is one of the lowest wholesale costs per gigabyte globally, at roughly 13 sen, supporting DNB’s argument that 5G, once fully utilised, is inherently more profitable than 4G.

That said, additional capital will be required to support the new spectrum. DNB estimates about RM1bil is needed for equipment installation for the additional 100Mhz rollout.

“It’s going to be close to about RM1bil. We are now undergoing what we call a tender process. So that price could well be much lower, but this is just a budget that we put in,” Azman says.

He does not see this as an undue burden on shareholders. Financing options are being evaluated, with cost efficiency central to the decision.

“The board will decide which is the best ratio in this perspective. How much would that be from equity and how much would that be from debt? I think we’ll find whatever ratio it is, what is the least cost of debt that everyone has to bear,” he adds.

In internal modelling shared with shareholders, Ahmad Zaki says rising traffic volumes are expected to compress unit costs steadily, pushing the network towards breakeven.

“Given the demand curve as it stands, it’s just a matter of a few more years before that break-even can come through. Eventually, 5G will overtake 4G as well.

“So if the cost element is lower to produce 5G and demand keeps going up, then DNB will benefit from that and reduce our losses further and get to break-even,” he explains.

Defending utilisation

Questions also remain over how DNB intends to defend utilisation levels and retain wholesale and enterprise clients.

“With the 200 MHz spectrum, that’s going to be a big uplift. So we are knocking on the doors – via our access seekers – of the seaports, the airports, and in-building as well. Those are the growth revenues, the growth pillars that we see,” Azman says.

He strongly rejects the notion that these losses would not have existed had DNB never been created.

Had each operator built its own network, Ahmad Zaki argues, capital outlays would have been higher, duplication unavoidable and consumer pricing pressure more intense.

Malaysia might have ended up with slower rollouts, patchier coverage and higher tariffs, outcomes that would have undermined long-term sector valuations just as surely as today’s accounting losses.

The company does not deny that policy reversals, timing missteps and communication failures have scarred public perception. What it rejects is the claim that the experiment has failed.

As Azman puts it, demand for 5G is real. With the cost base under control, what comes next is execution, not reinvention.

For the market, the question is whether investors are willing to look beyond the near-term drag and recognise that Malaysia’s 5G bill may already have been paid.