Plans to build data centre at Drax power station

Drax Power Limited has submitted a planning application to North Yorkshire Council to build two data centre buildings totalling about 4,960 sq metres on storage land north of the Drax power station near Selby; the proposal includes three-storey structures, a sprinkler plant room and tank, access roads, loading bays and parking. Planning documents describe the development as "major in scale" but state it would not cause significant or complex environmental effects; the application reflects rising data-centre demand driven in part by AI, with potential local implications for power consumption and site asset utilisation.

Analysis

Market-structure: The Drax proposal (site ~4,960 sq m) is small relative to hyperscale campuses but is a strategic signal that power producers will vertically integrate into data-hall hosting to capture high-margin, baseload demand. Winners: Drax (LSE:DRX) for land monetization and PPA leverage, hyperscale REITs (DLR, EQIX) for continued enterprise demand, and UK grid/renewables suppliers (NG, SSE) who capture network upgrades and PPAs. Losers: traditional office REITs and local distribution networks facing upgrade costs; expect modest upward pressure on regional power prices over 1–3 years as colocations scale.

Risk assessment: Immediate risk (days–months) is planning rejection or delay—assign ~30–50% chance of >6‑month delay given local objections and environmental reviews. Medium-term (6–24 months) risks include grid connection constraints and PPA pricing stress if wholesale power or carbon prices spike; tail risks include punitive local regulation or an ESG-driven moratorium that could strand assets. Hidden dependency: value hinges on firm PPAs and guarantee of grid capacity—without them economics evaporate even if planning is granted.

Trade implications: Tactical direct plays: small-cap UK exposure to DRX (1–2% position) to capture upside from development/lease revenues within 6–18 months; larger overweight in DLR (NYSE:DLR) or EQIX as secular AI demand plays over 12 months. Use pair trades: long DLR vs short office REITs (LSE:LAND or LSE:BLND) to express rotation from office to hyperscale real estate; implement 9–12 month call spreads (10–20% OTM) on DLR for convex exposure while limiting premium risk.

Contrarian angles: The market underestimates grid/utility winners—National Grid/SSE may rerate as capex visibility improves, while markets may overestimate immediate revenue impact to Drax (small footprint). Historical parallels (e.g., Northern Virginia data‑centre buildout) show multi-year utility rate cases and local pushback; that implies prefer optionality via liquid REITs/option structures rather than large direct equity bets until PPAs and grid agreements are announced within 6–12 months.

Key Decisions for Investors

• Establish a tactical 1–2% long position in Drax Group plc (LSE:DRX) within 30–90 days to capture land/hosting optionality; trim to zero if no signed PPA or tenancy announcement within 9 months or if planning is officially refused.
• Overweight Digital Realty (NYSE:DLR) by 3–5% as a secular AI/data‑centre play over 12 months, or buy a 9–12 month call spread 10–20% OTM to limit premium; take profits if DLR outperforms the broad REIT index by >8% within 6 months.
• Implement a pair trade: long DLR (or EQIX) vs short Landsec (LSE:LAND) 1:1 notional for 6–12 months to capture expected rotation to hyperscale real estate; close if the spread tightens to <2% or macro liquidity tightens (yield shock >75bp move higher in gilts).
• Add 2–3% exposure to UK grid/renewable beneficiaries (National Grid LSE:NG, SSE LSE:SSE) to capture likely network capex and PPA volumes; sell down if government/regulator issues a moratorium or materially reduces allowed returns within 12 months.