CORDIANT DIGITAL INFRASTRUCTURE LIMITED (CORD) — Investment research note

Executive summary

CORD is a London-listed closed-ended investment company that owns six digital-infrastructure platforms in Europe and North America (Emitel in Poland, CRA in the Czech Republic, Speed Fibre in Ireland, Hudson in NYC, Belgian Tower Company, and a 37.4% stake in DCU in Belgium), covering broadcast/mobile towers, fibre networks and data centres. Across the five years of filings, the company has built EBITDA from ~£100m to a £168m run-rate (12 months to Dec 2025), grown the dividend from 1.5p to 4.35p, raised €375m of debt, and the share-price discount to NAV has compressed from 46.7% (March 2024) to ~10% today as operating performance and the AI-driven data-centre narrative have taken hold. The single most important valuation point is that the gap between the discounted share price and the independently-reviewed NAV has now largely closed, so the upside from here depends on NAV growth (driven by EBITDA roll-forward, the Prague Gateway 26MW build, and potential discount-rate compression) rather than a re-rating.

Fair value estimate

Methodology: NAV-based, cross-checked against EV/EBITDA for digital-infrastructure peers. The Investment Manager's DCF NAV at 31 March 2025 was 129.6p per share, derived using a 9.32% weighted average discount rate and independently reviewed by a Big-4 valuer. EBITDA has grown 7-9% on a constant-currency basis through the financial year (Q3 trading update 2026-03-04), so March 2026 NAV — to be reported on 19 June 2026 — should plausibly land in the 135–142p range absent significant FX/discount-rate moves. The Prague Gateway data centre is currently held at nil value but will receive a third-party valuation at the March 2026 NAV (potentially a positive uplift).

Fair value range: 120p – 142p per share (implied market cap £919m – £1,087m). Midpoint ~131p implies market cap ~£1,003m vs. the disclosed £895.9m — upside of approximately 12%. Even a low-case 120p only implies modest upside; a high-case relies on Prague Gateway being marked at meaningful value.

Sector context

Confirmed as Financial Services (ICB), but in substance this is digital-infrastructure equity exposure delivered through a closed-end fund wrapper. On infrastructure-platform metrics — 4.7x net leverage, ~50% EBITDA margins, inflation-linked long-tenure contracts — the underlying portfolio is in line with tower/data-centre peers; the differential is the fund-wrapper structure and the mid-market Central European weighting. Listed peers include Helios Towers (African towers), Cellnex Telecom (European towers), Digital 9 Infrastructure (a UK closed-end peer currently in wind-down — a clear cautionary precedent). CORD trades at lower EV/EBITDA than listed tower comps which often sit at 18–25x.

Investment thesis

• AI-relevant data-centre pipeline at fair carrying value. The 26MW Prague Gateway data centre on CRA-owned land has been nominated by the Czech government as a potential EU AI Gigafactory candidate and is currently held at nil book value pending an independent valuation at the March 2026 NAV 2026-03-04 Q3 trading update; 2025-09-18 Q1 trading update. DCU (Belgium) has 13MW operating with 11MW expansion potential, including a recent 17MW Antwerp power-secure addition. Hudson is expanding by 2MW. Speed Fibre's BTCIL acquisition was framed as supporting "the rapid uptake in artificial intelligence applications" in Ireland 2025-09-01 BT Ireland completion.
• Multi-year compounding from a diversified, contract-backed base. Portfolio EBITDA grew 7.1% constant-currency over the nine months to Dec 2025 to £125.1m, with the dividend covered 1.8x by AFFO. Index-linked contracts, blue-chip customers (TVP, Polskie Radio, Czech Radio, Orange, T-Mobile, Proximus, Pfizer, Atos) and 70%+ fixed/hedged debt mean operating cash flows are insulated from rates and FX swings 2026-03-04 Q3 trading update.
• Discount narrowing supported by tangible catalysts. Migration to the Official List / FTSE 250 eligibility is targeted via an April 2026 EGM, the management fee is market-cap based (aligning the manager with shareholders), insiders (including Steven Marshall, ex-American Tower President) own 2.3% and have been buying 2026-03-04 Q3 trading update; 2025-06-19 annual results.

Key risks

• Hudson chronically underperforms. EBITDA losses persist; £18.4m write-down in FY24; LTM EBITDA loss still negative through Dec 2025 with management noting it is "unlikely to show positive EBITDA in the next twelve months" 2025-06-19 full-year results. Carrying value reduced to £36m but it remains a value drag.
• CRA legal dispute. A long-running squeeze-out valuation claim resulted in a first-instance ruling against CRA in February 2025 (now under appeal); the size of any liability is to be determined and remains a contingent overhang 2025-06-19 full-year results; 2026-03-04 Q3 update.
• Discount-rate sensitivity and FX exposure. Weighted-average discount rate is 9.32%; a +100bps WACC move cuts NAV by ~£270m (~35p/share) on management's published sensitivity. The portfolio is heavily PLN/CZK/EUR-denominated and the Company does not hedge balance-sheet FX 2025-06-19 annual results, note 6 and FX section.

Operating leverage

The portfolio combines capacity-constrained services with operational gearing (towers, fibre, data centres) — the fixed cost base is significant: tower operating costs scale slowly with new tenants; data centre power costs are largely passed through to customers; broadcast networks have minimal incremental cost per added channel. Aggregate EBITDA margin is ~46% (£168m EBITDA on £364m revenue, LTM Dec 2025). Specific high-leverage inflection points: (1) Prague Gateway adds 26MW into a strained European power market where comparable build-cost-to-EBITDA multiples can be 5-7x — if it fills, EBITDA could lift £15–25m/year; (2) DCU has 11MW unbuilt; (3) Hudson's new 2MW data hall, if filled, swings the asset from EBITDA loss to positive contribution. On a 10–20% portfolio revenue surprise above plan, I would expect operating profit to grow roughly 25–50% above plan, given pass-through power, index-linked revenue and modest incremental opex. This is moderate-to-strong operating leverage but not pure-software-like.

Value-trap signals

• Closed-end-fund structure historically out of favour in UK; sector precedent (Digital 9 in wind-down) is unhelpful.
• Multi-year persistent discount to NAV despite operating delivery suggests structural rather than transient market scepticism.
• Hudson has missed expectations consistently since 2022 acquisition.
• 4.7x consolidated net leverage is at the upper end of comfort for an investment company.
• CRA legal dispute is unresolved.

Earnings vs. expectations

Trading updates and results across 2022–2026 show CORD has consistently met or modestly beaten its annual NAV total-return target of 9% (11.6% in FY25, 9.3% in FY24); portfolio EBITDA growth guidance of "high single digits" has been delivered each year (Emitel 6.7%, CRA EBITDA after asset sales 3.0%, Speed Fibre adj. EBITDA 11.0% in latest period); the dividend has been raised three times since IPO ahead of the original schedule. The pattern is deliver-to-promise rather than blockbuster beats — there is no equity-research consensus referenced in filings, so beats/misses are measured against management's own guidance.

Conviction

Conviction: 3 — moderate. Anchors: (1) NAV is independently valued by a Big-4 firm semi-annually using DCF cross-checked against multiples; (2) operating performance is verifiable and consistent across multiple reporting periods; (3) listed-fund structure means share-price-to-NAV gap is observable. Limits: (1) underlying assets are private/illiquid mid-market platforms where a different valuer could land 10-15% in either direction; (2) Hudson and CRA legal dispute are real uncertainties; (3) Prague Gateway's value is genuinely unknown until the independent valuation lands at March 2026 NAV.

Driver scoring rationale (0-1000 overall: 530)

A partial fit. Data-centre and Prague Gateway exposure plus DCU/Hudson interconnect assets give meaningful AI-receiver characteristics, and operating leverage is real, but: (1) most revenue is still broadcast/tower (not AI), (2) the closed-end fund wrapper limits the upside vs. owning a pure-play picks-and-shovels name, (3) valuation is now near rather than below fair value, (4) 4.7x leverage and Hudson/CRA legacy issues take it out of the top band.