Megaport Shares Lead ASX 200 Today as Halt Ends: What Happened?

Megaport shares (ASX:MP1) jumped out of a trading halt today, leading the ASX 200 after unveiling a fully underwritten $827 million equity raise to fund a strategic transformation into a distributed AI infrastructure provider, marking one of the most ambitious pivots in Australian tech this year.

The Megaport share price closed up 10.11% at A$18.29, significantly outperforming the broader market after resuming trade following a three day halt. The rally came despite the dilutive nature of the capital raise, with the offer price of A$14.30 per share representing a substantial discount to current trading levels. It did look to have been even better for MP1 holders at the open, with shares hitting an intraday, and new five year high of A$21.16 within the first half hour of trading.

The positive market response suggests investors are looking through near-term dilution to focus on the company’s long-term positioning in the rapidly expanding AI infrastructure market. Trading volumes surged as the stock emerged as the ASX 200’s best performer, with the institutional component of the raise demonstrating exceptional demand at a 99% take-up rate among eligible shareholders.

The Capital Raise

The fully underwritten accelerated non-renounceable entitlement offer will issue approximately 57.9 million new shares at A$14.30 each, raising A$827.3 million in total. The institutional component, which closed earlier this week, raised approximately A$518 million through the issue of around 36.2 million new shares. The near-complete institutional take-up, with excess demand covering the small shortfall, signals strong institutional confidence in the strategic direction.

The retail entitlement offer, expected to raise approximately A$309 million, opens on 11 June 2026 for eligible shareholders in Australia and New Zealand at the same A$14.30 price. Eligible retail shareholders will have the opportunity to apply for additional shares beyond their entitlement if a shortfall exists, subject to company discretion on scale-back. New institutional shares commence normal trading on 15 June 2026, while retail shares will begin trading on 7 July 2026.

From Connectivity to AI Infrastructure

The capital raise marks a fundamental transformation for Megaport from a pure-play Network-as-a-Service provider into a distributed AI inference cloud operator.

The company plans to deploy the proceeds to fund four new AI infrastructure contracts with combined total contract value of approximately A$458.9 million and to establish an on-demand GPU compute pool requiring roughly A$350 million in capital expenditure, primarily for NVIDIA GPUs. This infrastructure will be deployed across Megaport’s existing footprint of more than 1,100 connected data centres spanning 31 countries, creating a globally distributed AI platform that differentiates the company from centralised neocloud competitors.

The strategic shift builds on Megaport’s recent acquisition of Latitude.sh, a bare-metal cloud platform specialising in GPU and CPU compute for AI workloads. Since joining the Megaport group in November 2025, Latitude.sh has delivered tens of thousands of servers globally and driven on-demand annual recurring revenue growth of approximately 31% to US$58.7 million by late April 2026.

The four new AI infrastructure contracts, which include GPU, CPU, storage, and network resources for US-based technology customers running AI inference at scale, are expected to begin ramping meaningfully in the first half of 2027.

Contract Economics

The economics underpinning the AI infrastructure contracts appear compelling on paper. Analyst commentary points to an approximate 27-month payback period on the associated GPU and infrastructure capital expenditure, with internal rates of return described as well north of 20% on the contracted component.

These returns are based on contracted revenue streams versus committed capex, providing a degree of visibility unusual for a company historically reliant on more variable connectivity revenues. The separate A$350 million on-demand GPU pool sitting alongside the committed contracts could drive incremental consumption revenue at potentially higher marginal returns if utilisation proves strong, though this remains an execution-dependent upside scenario.

The shift from asset-light connectivity infrastructure to capital-intensive GPU ownership represents a fundamental change in Megaport’s business model and risk profile. Where the company previously acted as connectivity plumbing between data centres and hyperscalers, it now owns and operates the compute infrastructure itself, moving closer to the AI spend rather than simply taking a connectivity toll. The distributed nature of the deployment across 1,100 data centres is positioned as a key differentiator for latency-sensitive inference workloads, data residency requirements, and enterprises seeking bundled private connectivity and compute solutions.

Analyst Response

Analysts have responded enthusiastically to the announcement, with multiple firms lifting price targets materially and maintaining overweight or outperform ratings. JPMorgan retained its Overweight rating while raising its target price from A$16.00 to A$28.00, citing the AI infrastructure pivot as materially de-risking the balance sheet through equity funding and contracted revenues, as UBS retained its Buy rating and lifted its target from A$16.70 to A$24.20, viewing the Latitude acquisition thesis as fully vindicated by contract momentum.

RBC Capital maintained its Outperform rating while raising its target from A$20.00 to A$22.00, as Macquarie retained its Outperform rating and raised its target from A$26.30 to A$27.80, emphasising that contracted compute structurally reduces earnings volatility compared to legacy connectivity revenues. The firm argued that procurement risk is lower than market perception given the contracted nature of the deals and strong institutional support for the equity raise.

Macquarie also flagged potential ASX 100 inclusion as an additional catalyst if market capitalisation and liquidity continue rising, which could drive passive index inflows and broaden the shareholder base.

What to Watch

The bull case rests on several key pillars: successful execution of the four major AI contracts with revenue ramping as planned from early 2027; strong utilisation of the on-demand GPU pool generating incremental high-margin revenue; validation of the distributed AI inference model as customers prioritise latency and data residency; and potential upside from additional contract wins as the platform scales. If these elements align, the current valuation could prove conservative relative to the earnings power of a scaled AI infrastructure platform with contracted multi-year revenues.

The bear case centres on execution risk in a complex new business model; slower-than-expected customer ramps or utilisation of GPU infrastructure; competitive pressure from hyperscalers and specialist clouds eroding differentiation; and the possibility that the company has over-invested in capacity relative to sustainable demand. Capital intensity and returns on invested capital will come under scrutiny if the AI infrastructure proves more costly to operate or refresh than currently modelled.

Near-term, markets will watch for any updates on the retail entitlement take-up rate when it closes in late June, contract implementation progress as the company moves toward the first-half 2027 ramp period, and any additional contract wins that would validate the land-and-expand thesis. Quarterly updates on GPU utilisation rates, both contracted and on-demand, will become critical metrics for assessing whether the transformation is tracking to plan.