Compare

The same question, asked of three operators.
59 per cent, 50, and 21.

Three companies build data centres in India and no two of them use the same words. One publishes built, installed and sold capacity. One publishes a headline that mixes operational with advance stage, and an operational figure two pages away from it. The third publishes neither an estate nor a headline, only three campuses and the phasing inside one of them. Putting those levels side by side produces a table that looks like a comparison and is not one.

Two measures survive. What earns divided by what the company headlines, asked of all three, and what each company's data centres actually earned, asked of the two whose statements are cited here. Every figure in them is printed in a filed document. On the first, Sify Infinit Spaces Limited earns on 59.2 per cent of its headline and Anant Raj on 21.4, and Techno Electric & Engineering Company on 50.0. Of 7 measures below, 2 travel between them.

Exhibit 1

Three operators earn on 59, 50, 21 per cent of what each of them headlines

Megawatts, all three ladders on one scale. Each column is that company's own rungs in its own words, and no bar in one column is the same measurement as the bar beside it.

Sify Infinit Spaces LimitedSify Infinit Spaces LimitedFY2025

59.2per cent of its headline earns

111.37 MW of 188.04, both printed in the same document.

  • Builtpage 142188.04 MW
  • Installedpage 142127.56 MW
  • Soldpage 142111.37 MW
Anant RajAnant RajFY2024-25

21.4per cent of its headline earns

6 MW of 28, both printed in the same document.

  • Operational and advance stage to operationalisepage 528 MW
  • Operationalisedpage 36 MW
  • Operationalised colocationpage 75.5 MW
Techno Electric & Engineering CompanyTechno Electric & Engineering CompanyFY2025-26

50.0per cent of its headline earns

24 MW of 48, both printed in the same document.

  • Three campuses added togetherpage 7248 MW
  • Commissioned and livepage 7224 MW

One scale, so relative size is visible as well as relative shape. Sify Infinit Spaces Limited headlines 6.7 times the megawatts the smallest estate here does, and every ladder is drawn against the widest rung in the exhibit rather than each against its own.

Sify Infinit Spaces Limited

59earning41not earning

Anant Raj

21earning79not earning

Techno Electric & Engineering Company

50earning50not earning

The rung names do not correspond. Sify Infinit Spaces Limited descends from built to installed to sold, three rungs of one estate defined in one document. Anant Raj descends from a headline printed with the words “operational and advance stage to operationalise” to an operational figure printed elsewhere in the same report, and then to the part of that figure which is colocation rather than cloud. Techno Electric & Engineering Company descends from three campuses added together, a total the report never prints, to the one of them it calls “commissioned and live”. Reading any ladder against another rung by rung would set one definition beside a different one.

The number above each ladder is the one that travels, because it is built from that company's own two figures. 59.2 per cent against 21.4 is not a difference in disclosure quality. Every company here disclosed enough to be measured. It is a difference in how much of what reaches the market is a built asset with a customer on it.

The third ladder carries the same gap inside a single address. Techno Electric & Engineering Company describes Noida as a 16 MW campus whose first phase is 500 kW, both in one sentence at printed page 72. That is 32 to one between the number attached to the site and the first increment due to run on it, without leaving one campus or one page.

The rung above it carries the warning the other two ladders carry. The words on Techno Electric & Engineering Company's live campus are “Commissioned and live. Receiving strong demand signals from global and domestic cloud operators. Phase II capex planning is underway.” A campus is called commissioned and live in the same breath as a second phase of it still being planned, which is what makes 24 MW the campus rather than what earns on it. Set against the 250 MW targeted by FY 2029-30 at printed page 73, it is 9.6 per cent.

Source: Sify Infinit Spaces DRHP, Basis for Offer Price, key performance indicators, printed page 142; capacity by data centre, printed pages 49 and 301; return on capital employed, printed page 287 Anant Raj Limited annual report FY2024-25, corporate overview highlights and the management discussion, which state the same three parts in the same words in two places. Techno Electric & Engineering Company Limited annual report FY2025-26, management discussion and analysis, hyperscale campuses

Exhibit 2

Sify Infinit Spaces Limited earns 66 times what Anant Raj's data centre arm earns

Millions of rupees on both sides, and the same twelve months to 31 March 2025. Lakhs are restated as millions, which is a change of scale inside one currency rather than a conversion between two.

Two data centre operators compared. One row here is like for like; the rest are levels, which sit side by side without asking the same question.
MeasureSify Infinit Spaces LimitedSify Infinit Spaces LimitedAnant RajAnant Raj
Revenue from data centreslike for likeThe second row that is like for like. One of these companies is a data centre operator entire, so its revenue is the answer. The other prints its data centre arm as one column in a statement of subsidiaries, and that column is the answer. Same measure, same period end, both filed.14,284Rs mn215.14Rs mn
Revenue, whole groupNot a comparison, and it is here to stop the row above being read as one. For one company the group is the data centre. For the other the group is a property developer, and the data centre is about one per cent of it.14,284Rs mn20,600Rs mn
Operating cash flow, whole groupFiled by both, at group level by both. For one that is cash from selling megawatts. For the other it is mostly cash from selling homes, so the levels sit side by side without asking the same question.5,474Rs mn966.1Rs mn
Return on capital, as each publishes itBoth print the ratio and neither prints the same one. One divides by average capital employed, the other by capital employed at the close. A closing denominator in a year of rising equity flatters the result, so these two numbers are not a ranking however alike they look.7.92per cent, average capital11per cent, closing capital

Techno Electric & Engineering CompanyTechno Electric & Engineering Company is not in this table. Its megawatts are filed and cited, so it stands in the ladders. The financial statements in the same report are not drawn on here, so it appears in none of the financial rows.

Netweb TechnologiesNetweb Technologies is not in this table. Owns no megawatts. It builds the machines that fill other people's estates, so every row below would compare a megawatt against a rupee of backlog.

Both sides are filed figures for the same year now. The row that travels is the first one, because it asks each company what its data centres earned: Sify Infinit Spaces Limited is a data centre operator entire, and Anant Raj prints its arm as one column in a statement of subsidiaries. On that measure they are 66 times apart.

The three rows under it are levels rather than comparisons, and the last is the sharpest of them. Both companies print the words return on capital employed, and neither prints the same arithmetic: one divides by average capital employed, the other by capital employed at the close. In a year when equity rose, the closing denominator is the kinder one. The higher number on that row is not the better return, it is the different formula.

Source: Sify Infinit Spaces from its draft red herring prospectus. Anant Raj from its annual report for FY2024-25, filed with the exchange. Both pinned by checksum and cited by printed page in the sources document.

Where this widens

A third operator, not a fourth metric

Both figures on this page come from a document filed with a regulator or an exchange. Several more names on the coverage matrix publish megawatts and file the same way, so the ratio above can be asked of them on the same terms. Adding one of those widens this comparison. Adding another metric does not, because a metric only travels when both companies print both of its numbers.