Core Scientific to acquire Polaris DS, expand Oklahoma campus

Core Scientific plans to expand its Muskogee, Oklahoma campus to about 1.5 gigawatts of gross power and has agreed to acquire Polaris DS LLC, which brings 440 megawatts under contract with Oklahoma Gas & Electric. The company has already started construction on a second 82.5 MW building, secured about 250 acres, and expects the acquisition to close in Q3 2026 with initial delivery from the new building in Q4 2027. The update reinforces Core Scientific’s AI infrastructure buildout, while analysts still see 97% revenue growth for fiscal 2026 and maintain an outperform view with targets of $20 to $40.

Analysis

This is less a pure growth story than a capital formation story around scarce power. The market is starting to price CORZ as a utility-like option on interconnection rights and contracted megawatts, which helps explain the multiple rerating, but it also means execution risk is being pushed several years out while valuation is being pulled forward now. The nearest-term upside is not from the new campuses themselves, but from continued repricing if the company keeps converting legacy mining assets into contracted AI infrastructure faster than peers can secure power.

The second-order winner is NVDA-adjacent demand: every incremental AI campus completion tightens the bottleneck on power rather than chips, which indirectly supports premium pricing for high-density deployments and makes liquid cooling, switchgear, transformers, and grid services the real supply chain choke points. That should also help peers with similar power pipelines and hurt smaller colo operators that lack balance-sheet capacity or utility relationships. The financing package is important because it signals the asset is becoming financeable at infrastructure-like terms, but that same leverage cuts both ways if commissioning slips or customer take-up slows.

The key risk is timeline slippage. This is a 12-24 month catalyst stack with limited visible cash conversion in the near term, so the stock can de-rate hard on any regulatory, permitting, or commissioning delay even if the long-term thesis remains intact. The weakest point in the bull case is gross margin: if power and build costs stay elevated while customer ramps arrive later, the market may decide the stock is trading more on narrative than near-term earnings power.

Consensus is probably underestimating how much optionality is embedded in the land/power portfolio versus the operating business today. But it may also be overestimating how quickly that optionality turns into per-share value, especially after a strong run. The cleanest edge here is to own the secular winner while fading the timing risk through options or relative-value structures rather than chasing outright equity at elevated expectations.

Key Decisions for Investors

• Maintain a tactical long CORZ, but size it as a 6-12 month catalyst trade rather than a core position; upside remains if power conversion milestones land, but any delay before 2H26 likely drives a 15-25% air pocket.
• Buy CORZ Jan-2027 calls / sell Jan-2027 puts on weakness to express upside from multi-year campus monetization while limiting exposure to near-term execution misses; best entry on post-rally consolidation.
• Pair trade: long CORZ vs short a basket of lower-quality colocation names without comparable power access; the spread should widen if investors continue rewarding scarce interconnection rights over generic rack growth.
• Add NVDA on pullbacks as a second-order beneficiary of AI infrastructure buildout; the best risk/reward is through diversified demand exposure rather than single-project execution risk.
• Avoid chasing CORZ after sharp up days; wait for confirmation from financing closes or commissioning progress, since the stock can easily retrace 10-20% on any delay before revenue realization.