Will APLD report Q4 FY2026 HPC base rent revenue above $50M?

Q3 base rent was $44.1M from 100MW on a full-quarter basis. The question asks whether Q4 exceeds $50M — a 13.4% increase on identical operational capacity. PF1 Building 2 has a July 1 RFS target, which falls after Q4 ends (May 31), so no Building 2 contribution is expected. No contractual escalators have been publicly disclosed. The most plausible path to $50M involves undisclosed escalators or very early Building 2 commissioning revenue in May, both of which are speculative. The $441K/MW/quarter unit economics established in Q3 should hold steady, yielding ~$44M again. There is a narrow possibility of CPI-linked or anniversary escalators not disclosed, but 13.4% growth on flat capacity is a high bar.

Taking the pessimistic view: the math is straightforward. 100MW at $441K/MW/quarter = ~$44.1M. To reach $50M from the same capacity requires either a ~$60K/MW/quarter escalator (a 13.4% step-up) or incremental revenue from a source that does not yet exist operationally. Building 2 is not scheduled for RFS until July 1, well after Q4 ends May 31. Even if construction runs a month ahead of schedule, commissioning revenue from a partially operational building in May alone would be modest. The CoreWeave lease terms are structured as fixed base rent — there is no disclosed mechanism for a 13%+ quarterly increase. Power pass-through ($8.1M in Q3) is reported separately. This is a relatively clear NO case with only tail-probability paths to YES.

The base case is $44M repeating on flat capacity, but there are legitimate tail scenarios. First, 15-year datacenter leases commonly include annual CPI or fixed escalators (2-4% typical in the industry), which APLD may not have disclosed granularly. A 3% annual escalator would add ~$1.3M/quarter — meaningful but insufficient alone. Second, APLD could reclassify some revenue between categories. Third, if Building 2 construction runs ahead by 4-6 weeks and partial commissioning begins in late April/May, even 20-30MW of early capacity could add $3-4M. The combination of modest escalators plus partial early commissioning is the most viable path. I assign slightly higher probability than the pure math suggests because datacenter leases commonly have escalators that management may not highlight until they appear in results.

The analytical picture is relatively clear: $44.1M base rent on 100MW with no capacity additions expected before Q4 ends. The $50M threshold requires a 13.4% jump from identical infrastructure. PF1 Building 2 targets July 1 RFS, after Q4's May 31 close. While undisclosed contractual escalators are possible — datacenter leases often include annual bumps — a 13.4% quarterly escalator would be unusually aggressive. Most CPI-linked escalators in the sector run 2-4% annually, which would add only ~$0.9-1.8M per year (negligible on a quarterly basis). The CoreWeave A3 credit rating and strong counterparty relationship are positive for payment certainty but do not create a mechanism for revenue growth on fixed capacity.

Taking the bearish view: the math makes this quite unlikely. Q3 established $44.1M as the run-rate for 100MW. Q4 has the same 100MW. Building 2 does not hit RFS until July 1. The $5.9M gap to $50M has no identified revenue source. Management has not disclosed escalators. Even optimistic scenarios (early Building 2, CPI bumps) would struggle to close a $5.9M gap. If Building 2 were somehow 6 weeks early and contributed May revenue at partial capacity, that might add $2-3M at most. A CPI escalator might add $0.3-0.5M quarterly. Combined best case is ~$47-48M — still short. The question specifically isolates base rent, excluding fit-out and power pass-through, narrowing the paths to YES even further.

The base case strongly favors NO given fixed capacity and no disclosed escalators. However, I give slightly more weight to unknown contract terms. APLD's lease with CoreWeave has not been fully disclosed publicly, and HPC datacenter leases in 2025-2026 have been structured with increasingly creative economics due to the AI infrastructure boom. Possible mechanisms: power density escalators if GPU utilization exceeds thresholds, performance bonuses, or stepped rent schedules that increase after the initial stabilization quarter. Q3 was the first full quarter — it is plausible the lease has a ramp structure. Still, 13.4% is a large jump and the base case is ~$44M repeating. Probability is low but not negligible due to contract opacity.

Q3 base rent was $44.1M on 100MW. Q4 has the same 100MW. Building 2 RFS is July 1 (after Q4). No disclosed escalators. The $50M target requires +13.4% growth on flat capacity. Possible upside from undisclosed escalators or early Building 2, but the math strongly favors staying near $44M. Probability is low.

The math is clear: $44.1M base rent on 100MW with no capacity changes expected. $50M requires $5.9M of incremental revenue from unknown sources. Industry standard CPI escalators would add at most $0.5M quarterly. Early Building 2 revenue would be minimal even if construction runs ahead. Base rent definition excludes fit-out and power, closing category spillover paths. This is likely a NO.

While the base case is clearly below $50M given fixed capacity, datacenter lease structures can include stepped rents, utilization bonuses, or annual escalators that may not be publicly detailed. The AI infrastructure boom has produced creative lease structures. However, even generous assumptions about hidden contract terms struggle to bridge a 13.4% gap in a single quarter. The most likely outcome is base rent of $43-46M, roughly tracking Q3.