Harworth H1 2026 slides: powered land gains offset residential drag

Harworth reported H1 2026 EPRA NDV per share of 214.8p, down 4.3% from year-end, and a negative 3.7% total accounting return, driven by a £15.8m loss on major residential developments and £14.9m total portfolio valuation losses. Net debt rose £44.1m to £190.0m, lifting LTV to 20.3%, although it improved to 17.1% by August; the interim dividend was increased 10% to 0.592p per share. Management is accelerating an exit from residential to focus on powered land and industrial logistics, supported by 0.8GW of accepted power offers and a data-centre land pipeline with £293m of potential future profit not reflected in NDV. The strategic pivot is underpinned by industrial/logistics ROCE of 24.0% annually in 2023-25, but execution, residential disposal timing and construction-cost risks remain.

Analysis

HWG’s strategic value is increasingly driven by scarce grid-connected land rather than conventional UK development economics. If management can monetize power capacity before energization, it converts a multi-year construction and leasing risk into earlier cash realization; this should warrant a higher capital-turn multiple than residential land banks, but only after a second transaction establishes pricing and repeatability beyond the initial hyperscale counterparties. The relevant read-through is modestly positive for MSFT’s UK capacity buildout, but more importantly pressures competing industrial-land owners without secured power, whose apparent land-bank value may prove less monetizable.

Near-term NAV marks may remain noisy as residential disposals and planning-stage industrial sites absorb construction-cost inflation. A pivot away from housing can improve earnings quality, but it also removes a source of recurring plot-sale liquidity during a period of elevated development spend; the key balance-sheet test is whether asset sales and powered-land deposits fund capex without LTV re-expanding. Over 1-3 months, catalysts are a binding hyperscale sale, planning approvals, and disclosed proceeds from non-core residential exits; over 6-18 months, the valuation rerating depends on proving realized returns and lower overhead, not on pipeline GDV.

The contrarian risk is that the market is already capitalizing a meaningful portion of the AI/data-centre optionality despite no independently verified valuation for the remaining powered-land pipeline. Grid connection dates, planning conditions, customer concentration and tenant financing can turn accepted power offers into long-duration options rather than near-term cash flows. Source-quality and quote-currency inconsistencies also make a spot valuation conclusion unreliable; verify the London share price, current discount/premium to NDV, and sale economics before committing capital.

Key Decisions for Investors

• Place HWG on a catalyst watch rather than initiate at market: buy only if verified trading price implies at least a 20% discount to reported NDV and the next powered-land transaction discloses binding consideration/deposit economics. Target a 15-25% rerating over 6-12 months; exit if LTV moves back above 25% or a major data-centre counterparty delays commitment.
• If the above entry conditions are met, express the thematic view as long HWG / short equal-value UK housebuilder basket (TW., BWY) for 6-12 months. The pair isolates capital rotation toward powered industrial land from UK housing demand; review or close if residential divestments occur below book value or housing demand materially re-accelerates.
• Do not use MSFT as a direct sympathy long on this development. Treat additional UK data-centre land commitments as a capex-intensity watch item: incremental commitments are strategically supportive but could raise investor scrutiny of free-cash-flow conversion if AI revenue evidence does not keep pace.
• Monitor UK planning decisions, grid-connection milestones, construction-cost indices and UK gilt yields monthly. A 50-75bp rise in long-end yields or further grid-connection slippage would likely compress industrial-land valuation assumptions before any operating benefit from the portfolio pivot is visible.