NextDC share price: is M3 delayed or dead?
A power delay at M3 has taken ~$2.6bn off the NextDC share price, and it shows how much risk data centre stocks carry when the grid can’t keep up.
NextDC (ASX: NXT) has dropped 37.9% from $9.18 to $5.70 today, taking ~$2.6bn off its market value. The problem is power, not demand. Jemena, the electricity distributor that supplies NXT’s M3 Melbourne data centre (colloquially known as Mordor), has suspended work on a 66kV line upgrade because of community opposition. It does not expect to restart before February 2027, so the next increase in grid supply to M3 now arrives in FY28. A data centre without power is an expensive shed. Is the selling overdone or justified?
NXT has taken the affected M3 contracts out of guidance. FY27 underlying EBITDA has been cut from $385m–$410m to $245m–$265m, flat on the $248.8m reported in FY26. The August guidance included ~$140m of EBITDA from M3. FactSet consensus EBITDA still sits at $401.5m, so forecasts need to come down ~36%. Net revenue guidance is cut from $615m–$640m to $475m–$500m. CapEx guidance down from $5.25bn–$5.75bn to $4.5bn–$5.0bn, and the board has put all new development projects and land acquisitions on hold.
Why NXT is not as cheap as it looks
The de-rating has been fast. NXT closed at $13.87 after the FY26 result on 28 August before falling over 30% to settle yesterday at $9.18. Today’s $5.70 is ~59% below the post-result price. On FactSet numbers, NXT trades on 15.8x FY27 EV/EBITDA, or enterprise value to EBITDA, down from more than 30x on 28 August. Global data centre peers trade on an average of ~19x trailing EV/EBITDA, so NXT looks cheap. It isn’t, yet. The 15.8x still uses the old $401.5m consensus. Applying the same enterprise value to the new $255m guidance midpoint gives ~25x, and counting September net debt, including the $1.1bn convertible, gives ~31x. Adding back the M3 EBITDA brings that to ~20x, in line with peers. On the multiples, NXT is only cheap if M3 survives. FactSet puts the shares on 0.7x FY27 book value. That book includes revaluation gains, and NXT has flagged an M3 writedown at the half year.
What the NXT share price is pricing in
The size of the fall shows what the market assumes. If M3 is a 12-month delay, NXT will lose a year of ~$140m EBITDA, plus any service level credits and a writedown on M3’s book value. That costs hundreds of millions of dollars, not billions. If the customer walks away, the ~$140m goes permanently. At the ~19x peer average, that costs ~$2.7bn, slightly more than today’s $2.6bn fall. The market is pricing the M3 contracts as lost, not delayed. Our read is that the sell-off is overdone if the company’s description of a delay holds.
Contract and funding risks at NXT
Nothing confirms that yet. The contracts allow service level credits for late delivery and, in some circumstances, termination. NXT is still checking whether force majeure relief applies. The Victorian Government will not decide on planning for M3 Stages 6 and 7 until 2027, after the 28 November state election.
Funding is the second unknown. NXT had $8.7bn of liquidity at 30 September and meets its covenants. It now says the M3 delay could limit its ability to draw the $6.7bn of undrawn senior debt. FactSet consensus, set before today’s crash, has net debt rising from $2.43bn at June to $13.24bn by FY29, with statutory losses in each of FY27, FY28 and FY29. If the debt becomes harder to draw, equity fills the gap. At $5.70, a raise would dilute shareholders far more than April’s $1.5bn at $12.70.
The next answers come from the M3 customer, and from the board’s CapEx review at the annual general meeting on 20 November. Until then, the share price will trade on the worst case. Are we buying? Day traders only until the dust settles.