Megaport 827M AI Raise|29% Premium, RSI 80 Who Holds the Risk?

Chapter 1: The Raise Anatomy — What 99% Institutional Take-Up Actually Signals

Megaport just completed the largest capital raise in its history. Eight hundred and twenty-seven million dollars, split across two tranches. Five hundred and eighteen million from institutions. Three hundred and nine million from retail. The institutional tranche closed with a 99% take-up rate. That number is being cited everywhere as proof of conviction. But there is a detail buried in that statistic that changes its meaning entirely.

Every institution that participated in the raise did so at fourteen dollars and thirty cents per share. At the time of writing, Megaport trades at eighteen dollars and forty-two cents. That is a twenty-nine percent premium to the price at which the smartest, most resource-rich investors on the ASX just bought in.

Here is the uncomfortable question that framing raises. When an institution commits to a raise at a twenty-nine percent discount to market, is that a signal of conviction in the business? Or is it a signal of conviction in the discount itself?

These are not the same thing. A fund that buys Megaport at fourteen dollars and thirty cents and sees it trade at eighteen-plus has already made twenty-nine percent before the business moves at all. The structural incentive to participate at 99% is not just about believing in AI infrastructure. It is about locking in a spread that retail cannot access on the same terms.

The retail entitlement offer is now open, also at fourteen dollars and thirty cents. Existing shareholders can participate and receive new shares at the same offer price. But here is the arithmetic that matters. The retail window closes. If an existing holder does not participate, they face dilution without the compensating benefit of the cheap entry price. And if a non-holder wants exposure today, the only path is on-market — at eighteen dollars, not at fourteen-thirty.

This is not the first time a major ASX raise has created a two-tier entry point. But the scale of the gap here — nearly four dollars per share — is unusually wide. It means the conversation about Megaport's conviction story has to be separated into two distinct questions. The first question is whether the business is genuinely inflecting. The second question is whether the current market price already prices that inflection — and then some.

Phase one of this analysis addresses the first question through the AI contract stack. Phase two addresses the second through the valuation. But the raise structure itself is the lens through which both must be read.

Chapter 2: The AI Contract Stack — Four Deals and a Business Model That Just Changed

Alongside the capital raise announcement, Megaport disclosed four new AI infrastructure contracts. Combined contract value: four hundred and fifty-eight point nine million dollars. That is not a revenue recognition number. It is a total contract value — spread over the life of the agreements. But it is the largest single-week contract announcement in the company's history.

Understanding why this matters requires a brief step back into what Megaport actually does. Megaport built its business on software-defined networking — the ability to connect enterprise customers to cloud providers without physical infrastructure commitments. The model works on consumption. Customers pay for what they use. It is capital-light on the customer side, and it scales with cloud adoption.

The problem through most of 2025 was that growth had plateaued. The stock hit an intraday low of six dollars and forty cents in April this year. From that low to today's price, the stock has tripled. The recovery narrative is almost entirely built on one pivot: AI infrastructure.

What the four new contracts represent is the first concrete monetisation of that pivot. Megaport is no longer just moving data between clouds. It is positioning itself as the connectivity and compute layer underneath AI workloads. The acquisition of Latitude.sh and the VAST Data AI Operating System partnership are both pointing in the same direction. This is a company in the middle of a business model transition.

That transition is the most important analytical fact in this story — and it is also the most dangerous one for valuation. When a company changes what it is, the metrics used to value the old business no longer apply cleanly. Megaport was valued as a networking infrastructure company. The market is beginning to re-rate it as an AI infrastructure company. Those two categories carry different multiple frameworks, different growth expectations, and different risk profiles.

The FY26 revenue guidance has been tightened to the three hundred and seven million dollar range. That is the anchor for the current year. But the four hundred and fifty-nine million in new contract value is forward-dated — it accrues into future periods. The market is pricing some portion of that future accrual today, which is part of what explains the premium to offer price.

Here is the reversal card. Most coverage of the AI contracts treats them as the validation that the stock deserved to triple. The less comfortable read is that four contracts worth $459 million in total contract value — spread across multiple years — is still a relatively modest number for a company that just raised $827 million. The raise is larger than the near-term contract haul. That means Megaport needs additional AI contract wins to justify deploying that capital productively. The capital raise is not the finish line. It is the starting gate for a strategy that still has to prove out at scale.

Chapter 3: Valuation Stress Test — RSI at 80 and the Risk Structure Nobody Is Pricing

Earlier this week, Megaport's relative strength index touched ninety-one. That briefly made it the most technically overbought stock on the entire ASX. The RSI has since cooled to approximately eighty following a five-point-two percent single-session pullback. An RSI of eighty is not a sell signal on its own. Momentum stocks can stay overbought for extended periods. But it is a signal that the price has run far ahead of any news cycle's ability to justify it in real time.

The fifty-two week high was twenty-one dollars and sixteen cents, hit last Friday. The stock has since pulled back to the eighteen-forty range. That pullback of roughly thirteen percent from peak happened in less than a week. What that tells us is that the market is not uniform in its conviction. There are sellers at twenty-one dollars who were not sellers at eighteen dollars. And there are likely more sellers between here and twenty-one if the stock attempts that level again.

Now layer in the offer price. At fourteen dollars and thirty cents, the raise was priced at a level the stock has already traded through by nearly thirty percent. If the stock were to retrace to offer price — not an outlandish scenario given the RSI and the post-high pullback — every on-market buyer above fourteen-thirty is underwater. That is not a prediction. It is a description of the risk structure.

The Macquarie note flagging Megaport as a potential forty-percent-plus returner from current levels gets cited in bullish commentary. But a forty percent return from eighteen dollars implies a target above twenty-five dollars. That would be a new all-time high by a substantial margin for a company still delivering its first meaningful AI contract revenues. The analytical work required to justify that target is considerable — it requires the AI contract pipeline to expand materially, revenue to compound at rates well above FY26 guidance, and the margin profile of the new compute business to be comparable to or better than the legacy networking margins.

None of that is impossible. But none of it is yet demonstrated in reported numbers. What is demonstrated is this: Megaport raised capital at fourteen-thirty because that was the price at which it could clear eight hundred and twenty-seven million dollars in a single week. The fact that the stock trades twenty-nine percent above that price does not mean fourteen-thirty was wrong. It may mean eighteen-forty is too high.

The honest framing for any investor sitting with this stock today is straightforward. The business model transition is real. The AI contract wins are real. The institutional take-up was near-total. Those are genuine positive signals. But the risk structure has shifted. This is no longer a recovery story at six-forty. It is an AI-infrastructure pivot at a twenty-nine percent premium to what institutions paid four days ago. The question of whether that premium is justified is the question this stock will spend the next two quarters answering. The leaning is that the near-term risk is asymmetric to the downside until at least one of the four AI contracts begins recognising material revenue — and until FY26 results demonstrate that the compute pivot is margin-accretive rather than margin-dilutive. The business may be worth the premium. But it has not yet earned it in reported numbers. The Chekhov signal to watch: FY26 full-year revenue versus the $307 million guidance midpoint, and whether any of the $459 million in contracted value begins appearing in reported quarterly revenue before the end of calendar 2026. If those numbers hold or exceed, the premium compresses into a justified re-rate. If they miss, the offer price becomes the floor that tests how much of the 99% institutional conviction was structural discount versus genuine belief.