Applied Digital is a pure-play AI data center build-to-suit developer whose contracted franchise reached five campuses on 2026-06-08 with the Delta Forge 2 lease — a 15-year, 210 MW take-or-pay deal with the SAME U.S. high investment-grade hyperscaler that already anchors Delta Forge 1 and Polaris Forge 3. Total contracted base-term revenue is now ~$36B across five campuses (1.4 GW critical IT load, ~2.15 GW utility power); ~$86B if all renewals are exercised. The single most important new datapoint is that APLD now discloses a contracted annual NOI pipeline of ~$2.1B — already at the CEO's prior internal $2B aspiration and double the public $1B/5-year target — which converts the story from 'can the platform reach $1B NOI?' to 'how fast does $2.1B of contracted NOI convert to operating cash flow, and at what multiple does the market capitalize it?' The book composition has WORSENED on concentration even as it grew: the DF1/PF3/DF2 IG hyperscaler is now $20.2B / ~56% of the $36B backlog (the 2026-06-08 open question — same vs new counterparty — resolved as SAME, their third lease), CoreWeave is ~$11B / ~31% (A3-rated SPV with lockbox), and the PF2 IG hyperscaler ($5B) is the only meaningful diversifier at ~14%. Q3 FY2026 delivered the first adjusted profit ($0.09 EPS, $44.1M EBITDA) on HPC base rent of $44.1M (3.5x QoQ), and ChronoScale spinoff closed 2026-05-05 (APLD retains ~97% of CHRN), removing the cloud drag. The stock round-tripped from $48 (post-PF3, May 21) to ~$41 after the 2026-06-10 -7% reaction to a $1.59B 7.0% senior secured notes pricing (ELN-04, PF1 fourth building, 150 MW, repays the Goldman bridge, closes ~June 16) — the market is now weighing accelerating leverage against the franchise re-rating. Analyst targets have run further ahead: Needham $83 (from $66) and Lake Street $90 (from $70) post-DF2, both reiterating Buy. The near-term execution calendar that gates the thesis: PF1 Building 2 RFS July 1 (data halls energize July/Aug/Sep, first material revenue Q1 FY2027); PF2 $2.15B escrow release requires ESA finalization by June 30 (substation is the 'longest pole'); and the $1.59B notes fund PF1's fourth building. The bear case requires execution slippage at the multi-site build OR a credit/escape event at the now-$20.2B single IG hyperscaler — a concentration that is the dominant idiosyncratic risk and grew, not shrank, with DF2. Competitively, the crypto-miner-pivot cohort keeps matching APLD at scale (Hut 8's 352 MW / $9.8B Beacon Point lease with an AA- tenant funded at 6.129% vs APLD's 7.0% — a real cost-of-capital gap), so the IG-hyperscaler 'scarcity' premium continues to erode even as APLD's contracted NOI compounds. By early July the stock had fallen a further ~20% to ~$33 (2026-07-02) on neocloud/leverage sentiment with no adverse fundamental change — into the ~$32-36 add band and at ~14x FY2027E EV/Revenue (mature-DC-REIT parity for a 50%+ grower); the debate is whether ~$2.1B contracted NOI plus a clean July execution print re-rates the equity toward the bull case, or whether single-name concentration and a 7% incremental cost of debt cap the multiple.