Anant Raj’s Data Centre value unlocking: Fact or fiction?

For most of its life, Anant Raj was a cautionary tale. Founded in 1969 and listed out of a clay-products business incorporated in 1985, it spent five decades assembling one of the largest land banks in Delhi-NCR, and a reputation for not doing very much with it.

Special Thanks to my dear friend and mentor Sushant B for supporting this article’s research 🙏🙏

Its corporate history is riddled with major strategic moves gone awry: a slide into BIFR “sick company” status by 1999, a mid-2000s hospitality push that announced twelve hotels and delivered five, and a commercial portfolio that by 2016 was only about 30% leased.

Then, the story turned. The same idle IT parks and half-empty buildings that were once the evidence for the shareholder disappointment became, almost overnight, the cheapest ready-to-convert data-centre real estate in the country.

Anant Raj began repurposing them, with land and shell already paid for, into data centres at a claimed capital cost of roughly ₹26 crore per megawatt, against an industry norm nearer ₹30-55 crore.

The stock followed the reinvention & a major upturn in branded developers prospects:

On 21 July 2026, the board decided to split these businesses. It approved a composite scheme to hive off the data-centre and cloud business out into a separately listed company, Ashok Cloud Private Limited, and leave the real estate and infrastructure business in the listed parent.

For a special-situations investor, that is exactly the kind of moment worth stopping for. But this is not a clean, obvious value-unlock. It is a genuinely two-sided situation: a real growth engine wrapped in real DC execution and governance doubts, and that is what makes it interesting.

Section 1 | Why demergers create value

When one listed company houses two very different businesses (a fast-growing, capital-hungry infrastructure bet sitting next to a cyclical property developer), the market rarely values it well. The investors who want the growth business do not want the cyclicality attached and the analysts who cover property have no framework on how to value a data centre. So the whole trades at a blended, compromise multiple, and the better half is dragged down by the worse.

All the above is underpinned by the management’s / shareholders’ natural desire to get a “better valuation”.

That is the conglomerate discount. A demerger breaks it: each business starts trading on its own, attracts its own investors and analysts, gets its own capital allocation, and is freed from the other’s cycle.

India’s recent record is a long list of the pattern paying off, and the demerger tracker behind this series bears it out: Reliance carved out Jio Financial Services in 2023, ITC listed ITC Hotels in early 2025 to free a cleaner, higher-return FMCG parent, and Tata Motors split its commercial-vehicle and passenger/JLR businesses 1:1 in late 2025.

Different sectors, same logic, the parts usually worth more than the whole. Anant Raj is a textbook candidate: the two halves differ in growth rate, margin, capital intensity and investor base about as sharply as two businesses can. Whether the discount actually closes here, though, depends less on the theory than on the execution.

Section 2 | The two businesses, and why they don’t belong together

Anant Raj today is really three things stacked in one wrapper: a residential real estate developer, a small annuity portfolio of leased commercial and hospitality assets throwing off about ₹91 crore of rent a year, and a young but rapidly scaling data-centre and cloud operation.

The scheme cleaves this into two: the property and infrastructure businesses stay in Anant Raj Limited (ARL); the data-centre and cloud undertaking leaves for Ashok Cloud.

2A. The base that stays: real estate and infrastructure

The remainco is not a dud, and that matters, because it is what separates this deal from a “dump the bad business” demerger. Anant Raj’s core is residential development in the strongest micro-market in the country: Gurugram, and specifically its roughly 240-acre, largely fully-paid legacy land bank in and around Sector 63A. That land was bought decades ago, so the developer’s margins on it are structurally higher than a peer buying land at today’s prices.

And the cycle is with it. NCR has been the standout of India’s housing upcycle, with Gurugram absorption outrunning new launches for years and inventory overhang collapsing from the mid-70-months range in 2017 to roughly 10 months.

Anant Raj’s own numbers rode that wave: consolidated revenue grew from ₹249.66 crore in FY21 to ₹2,059.97 crore in FY25, and net profit from ₹0.23 crore to ₹421.54 crore over the same four years. FY26 added another leg: revenue ₹2,511.60 crore, up 21.9%, and PAT ₹557.02 crore, up 30.8%.

The development pipeline behind that growth is large and freshly disclosed. The flagship Sector 63A township carries an estimated development potential of about 9 million sq ft and roughly ₹15,000 crore of revenue potential, and the company reports about 11.41 million sq ft of ongoing and planned residential projects plus 83.43 acres of fully-paid Delhi land held for future development. The two largest named launches anchor the near-term pipeline:

But there’s a catch. Anant Raj is a comparatively weak brand next to DLF, Godrej Properties, Lodha, Prestige or Sobha, which leaves it more exposed if the NCR cycle turns or supply floods back.

And its cash conversion has historically lagged its reported profits: operating cash flow was actually negative in FY24, around minus ₹16 crore, the kind of gap that points to revenue booked ahead of collections.

2B. The jewel that leaves: data centre and cloud

At Anant Raj the high-growth, high-margin, high-multiple business is the one walking out the door into Ashok Cloud.

Anant Raj has 28 MW of IT load operational (21 MW at Manesar and 7 MW at Panchkula), built brownfield inside its existing Haryana campuses. In H1 FY26 the data-centre business did just ₹58.42 crore of revenue, well under 5% of the group’s top line, yet it contributed about 75% of absolute EBITDA and 43% of PAT, at a segment PAT margin north of 43%.

The demerger filing itself pegs data-centre turnover at ₹145.90 crore against consolidated turnover of ₹1,627.72 crore, roughly 9% of revenue. Either way the point holds: a sliver of revenue, a majority of the profit engine, and a business the market simply cannot value correctly while it is buried inside a “realty” ticker.

The runway is large and the ambition larger. The stated roadmap is 63 MW by FY27, 117 MW by FY28, and 357 MW by FY32 across Manesar, Panchkula and Rai, a target that has actually been raised from an earlier 307 MW even as near-term timelines slipped, with about a quarter of the capacity earmarked for higher-value cloud rather than plain colocation.

Management guides to roughly ₹1,200 crore of data-centre revenue by FY27 and about ₹9,000 crore by FY32, though brokers model it lower and later: Emkay pencils in about ₹700 crore of DC revenue in FY27E and a roughly 20% IRR on the DC book only by FY45E.

Two very large state MoUs sit on top: ₹25,000 crore with Haryana signed 1 June 2026 and ₹4,500 crore with Andhra Pradesh via the subsidiary in November 2025.

CAPACITY AND REVENUE ROADMAP

On the cloud side, “Ashok Cloud” is a sovereign public-cloud platform designed, built and operated for Anant Raj by Orange Business, with deployment complete and customers onboarding. The company has stacked up the institutional credentials that actually gate government and enterprise work: empanelment as a MeitY sovereign cloud provider and a BSNL data-centre provider, PSU alliances with TCIL, RailTel, CSC and BSNL, TIA/Tier-III and ISO certifications, and a liquid-cooling tie-up with Spain’s Submer for AI-ready racks.

2C. The bear case

None of the above is the whole story, and a fair special-situations piece must clearly outline the risks too.

First, differentiation. India’s DC market splits into four archetypes: established operators (NTT, STT, Yotta) with about 950 MW live, financial-investor platforms (Princeton, CapitaLand), global entrants (Equinix, Colt, Digital Connexion), and domestic players.

Anant Raj sits in that last and least-advantaged bucket, arguably a “me-too” colocation landlord without a clear right to win.

Second, credibility of the economics: roughly ₹26 crore/MW capex is 15 to 25% below even efficient retrofit norms, and that the 75 to 80% EBITDA margins the small 6 MW to 28 MW base has shown are a landlord-style artefact that will compress toward the 40 to 55% range that established operators like NTT, Nxtra and STT actually earn as staffing, security and service scale in.

Third, the gap between operational, handed-over and billed capacity: as of Q2 FY26 only about 8 MW of the 28 MW had been fully handed over, and public conference calls had gone quiet after that quarter, thin disclosure for a story about this complex.

And then there is the track record, which is the crux. Beyond the 1999 BIFR episode and the hospitality retreat noted at the top, the other issues have persisted that include a court halting a 12-acre project in 2023-24 over an allegedly fraudulently obtained licence, RERA orders to refund homebuyers (about ₹9.5 crore) for delays, and, most relevant for the reader, the fact that Anant Raj has done this before: it spun off its earlier real estate arm as TARC (renamed from Anant Raj Global in April 2021), and TARC was subsequently penalised by SEBI in 2022 over dozens of disclosure lapses.

These specific legal and regulatory claims are unverified here and should be checked against the underlying filings and orders. Which brings us to the one precedent that matters most.

Section 3 | How the demerger will work

This is a two-step composite scheme, not a single carve-out, so it pays to be precise.

Step one: Anant Raj Cloud Private Limited (ARCPL), the wholly-owned subsidiary that has housed the data-centre and cloud operations, is first amalgamated into Anant Raj Limited, consolidating everything under one roof.

Step two: that combined data-centre and cloud undertaking is then demerged out into Ashok Cloud Private Limited (ACPL), which will be listed separately.

What you receive. For every 1 equity share of Anant Raj (face value ₹2) you hold on the record date, you will be credited 1 equity share of Ashok Cloud (face value ₹2), fully paid: a 1:1 entitlement, with no cash changing hands.

The nuance that separates this from a clean carve-out. In a textbook demerger the parent ends up owning none of the spun-off company; shareholders (including promoters) hold it directly instead in their respective shareholding pattern pre-demerger.

Here that is only half true. After the scheme, Ashok Cloud remains a subsidiary of Anant Raj: the parent company holds roughly 51% and public shareholders hold about 49% directly. So an ARL shareholder ends up with data-centre exposure twice: once through the new Ashok Cloud shares in their demat, and again through ARL’s retained 51% stake.

That is a double-edged design: it keeps the parent tied to the growth engine (good, if you like the DC story and dislike a pure-property ARL), but it also means the value-unlock is partial, and it plants a potential holding-company discount on ARL’s 51% stake, since the market rarely gives a parent full credit for a listed subsidiary it controls.

As a housekeeping step, ARL had already acquired 37.43 crore ACPL shares for ₹74.86 crore (at ₹2 face value) on 21 July 2026 to set the resulting company’s pre-scheme capital.

The Demerger Timeline.

Board approval on 21 July 2026 is literally the starting gun and the final listing may be at least 18-24 months away.

From here the scheme must clear the stock exchanges and SEBI (a no-objection or observation letter comes before the tribunal), the National Company Law Tribunal (a first-motion order to convene meetings, then a second-motion sanction), and the shareholders and creditors, including the mandatory public-shareholder e-vote, before Ashok Cloud can list on BSE and NSE.

The precedent: TARC. Anant Raj has run this playbook once already. Its earlier real estate arm was demerged and listed as TARC (The Anant Raj Corporation) around 2021. The results are sobering. Years after listing, TARC carries a market cap of roughly ₹3,656 crore, trades near ₹133, and has posted a three-year return on equity of about minus 8% on high debt and weak interest coverage, though, tellingly, it has finally found operational traction lately, with Q1 FY27 pre-sales up around 300% year-on-year to ₹602 crore and collections up 80% to ₹305 crore. The lesson for an Ashok Cloud investor is double-sided: a demerger from this promoter can take years to reward holders and can carry governance baggage into the new entity, but the underlying assets, given time, can eventually deliver. Therefore, it would be naive to assume that the value-unlocking would be instant.

Section 4 | The valuation math: is there value?

Disclaimer: What follows is an illustrative if-then framework using peer multiples and the numbers on the table today. It is not a target price and not a recommendation.

The market values the whole of Anant Raj at around ₹22,000 crore of market cap at roughly ₹625 a share, on a trailing P/E in the high-30s. Inside that sits a data-centre business earning about 75% of the group’s EBITDA on under a tenth of its revenue, and property multiples alone might be the wrong lens for it. Infact, our valuation estimates derived in part from MOSL estimates already assign a reasonable valuation to the data centre part.

Sum-of-the-parts. On 22 July 2026, the day after the board approved the scheme, MOSL released a note on SoTP-based target of ₹710 a share (a BUY, roughly 15% above the price then), and split that value across the same three pieces:

How the numbers are arrived at: The residential and data-centre legs are discounted-cash-flow valuations (project and capacity cash flows discounted back, at a 12.4% cost of capital for residential), each carrying a 25% premium that credits growth not yet in the launched pipeline, while the commercial leg is simply the rental stream capitalised at an 8.0% yield. Per-share figures use MOSL’s roughly 360m share count.

Real estate (residential plus commercial) is worth about ₹354 a share and the data-centre arm about ₹341. This means about 48% of Anant Raj is already attributed to the DC & Cloud business.

Even if we go by the above valuation estimates and assume there is a 15% valuation upside, it’s hardly convincing since the actual value unlocking is at least 18 months away.

Meanwhile, we doubt the stock will be driven by value unlocking events but rather by how the DC story continues to play vs what the management has charted out. This demerger story requires tracking on-time execution on a quarterly basis rather than a near term value unlocking event.

Hope this was insightful.

Rahul Rao, CFA