Centuria Industrial REIT FY26 slides: 30% leasing spreads, data centres

Centuria Industrial REIT (ASX:CIP) reported FY26 FFO of $114.1m (18.2c/unit), up 4% YoY, with distributions rising to 16.8c/unit and a 92% payout ratio. Leasing momentum drove 30% re-leasing spreads and like-for-like NPI up 6.1% to $204.0m, while the portfolio remains ~17% under-rented with ~55% of leases expiring in the next three years to drive rental reversion. Despite higher finance costs ($65.9m vs $59.0m in FY25), the trust refinanced $450m at 10–20 bps tighter margins and extended weighted average debt expiry to 3.6 years. FY27 guidance calls for FFO of 18.8–19.2c/unit (+up to 5.5% vs FY26) and distributions of 17.3c/unit, with units still trading at ~23% discount to net tangible assets ($3.08 vs $4.01).

Analysis

CIP’s core industrial book looks like a slow-burn compounding story rather than a rerating catalyst, but the market may still be underappreciating how much embedded rent upside is already locked in versus what is being paid today. The better second-order winner is any listed owner of scarce urban land with power proximity and permitting optionality; in Australia that favors the higher-quality logistics compounders and data-center-adjacent landlords over generic yield names. The flip side is that the apparent discount to NTA can persist if investors keep treating the trust like a bond proxy while debt costs remain elevated.

The near-term catalyst path is mostly about execution, not narrative. Over the next 1-3 months, the stock should trade on funding clarity for the buyback, continued premium asset sales, and whether management can show a credible route from landbank to powered capacity; without that, the data-center angle is just out-of-the-money optionality. Over 6-18 months, a funded JV or long-dated powered land lease could justify a materially higher multiple, but failure to secure power/planning would force the market back to core industrial REIT metrics where higher leverage and lower hedge coverage matter more.

Consensus may be too fixated on the headline discount to NTA and not enough on conversion friction: utility queues, capex intensity, and the risk that Australia captures the real estate rent while the AI economics accrue offshore. That means the market could be overpaying for "AI adjacency" before there is visible monetization. Relative value looks better than outright beta: CIP is interesting only if the discount persists and asset-sale validation continues; otherwise the cleaner expression is a quality industrial/data-center leader with proven execution, not a turnaround story with long-dated permitting risk.

Key Decisions for Investors

• Accumulate a starter long in CNECF/CIP only on pullbacks while the stock still trades at a material NTA discount and the buyback remains active; target a 10-15% re-rating over 3-6 months, but cut if the discount widens beyond 30% or premium asset sales stop.
• Pair trade: long CNECF/CIP vs short a broad Australian REIT proxy such as VAP for 1-3 months to isolate idiosyncratic re-rating from rate-sensitive beta; thesis fails if bond yields sell off and the whole REIT complex reprices higher.
• Prefer GMG over CIP for structural exposure to industrial logistics and data-center growth; CIP has cheaper entry valuation, but GMG offers cleaner execution and lower permitting/monetization risk over 12-18 months.
• Treat the data-center pipeline as an alert, not a standalone thesis: do not add aggressively until there is concrete proof of power allocation, planning approval, or a funded JV; without that, rallies should be faded rather than chased.