Data Centre S-REITs in 2026: AI Demand Narrative Meets Sparse Sector Data

Originally published on Finance Pulse Research. This Dev.to mirror is provided for the developer/data-analytics community; the full interactive analysis with live data tables lives on the original.

Sector Definition and Scope

A surprising gap sits at the center of this 2026 review. The sector label is Data Centre, yet the current dataset tracks 30 instruments and every one of them is classified as a REIT, while none are classified as stocks. More strikingly, the listed records do not provide a populated top-yield table for the sector, and the country distribution table is not yet covered in the source block. That makes this less a conventional leaderboard and more an examination of coverage quality, sector mapping, and what can still be extracted from the available REIT-level records.

In Asian income markets, data centre REITs typically sit at the intersection of infrastructure demand, digitalization, and income-seeking capital. The role is clear even when the dataset is incomplete: these vehicles represent real-asset exposure wrapped inside exchange-listed income structures. Finance Pulse Research tracks 30 instruments for this sector scope, with 0 stocks and 30 REITs. As of 2026-06-20, the freshness fields show a real-yield snapshot date of 2026-06-19 and a REIT snapshot date of 2026-06-06.

That scope matters. A sector with 30 instruments is not small in absolute terms, but the present data block does not deliver sector-specific aggregate yield fields, country counts, or a populated top-10 yield list. Analysis therefore depends on the REIT roster and on explicit acknowledgment of missing fields rather than assumption. Readers looking for broader REIT context can compare this piece with data centre REIT sector data, Singapore REIT analytics, REIT methodology notes, yield factor research, and cross-border flow coverage.

Aggregate Metrics Overview

The headline aggregate table is unusually sparse. Data shows the sector contains 30 instruments, all of them REITs, but average nominal yield and average real yield are data not available at the sector level in the source block.

Even with those gaps, a few structural points stand out. First, the entire sector universe in this file is REIT-only. There are 0 stocks and 30 REITs, so any interpretation of “data centre sreit” in this article is effectively a listed-property analysis rather than a blended equity-and-REIT sector study. Second, the blank aggregate yield fields contrast sharply with the broader comparison data supplied for other sectors. The all-REIT comparison bucket contains 65 instruments with an average nominal yield of 6.073553846153846 and an average real yield of 3.8411538461538464. Energy covers 16 instruments with average nominal yield of 4.5259375 and average real yield of 2.823. Finance covers 45 instruments with 4.365355555555555 nominal and 2.4590666666666667 real. Consumer includes 13 instruments with 4.475384615384615 nominal and 2.2786923076923076 real. Telecom spans 17 instruments with 4.398411764705882 nominal and 2.2545882352941176 real.

Those benchmarks do not fill the missing data centre sector averages, but they do establish the surrounding landscape. Broad REIT coverage in the database shows higher average income metrics than Energy, Finance, Consumer, and Telecom. The data centre sector entry, however, cannot be slotted cleanly into that ranking because its average nominal yield and average real yield are not yet covered.

A different pattern emerges when the reader shifts from missing top-line averages to the available constituent list. The sector tag says Data Centre, but the REIT roster spans Retail, Hospitality, Industrial, Office, Logistics, Diversified, Healthcare, and Data Center sub-sectors. That spread indicates a classification or coverage issue inside the current sector table rather than a pure-play data-centre-only universe. In other words, the aggregate row signals breadth, while the constituent detail signals heterogeneity.

The absence of populated yield-distribution statistics reinforces that point. The source block shows yield distribution count at 0 and real-yield distribution count at 0, with mean, median, quartiles, standard deviation, minimum, and maximum all data not available. Since datasets with fewer than 15 entries do not warrant summary statistics and this one technically has 30 instruments but no populated distribution values, the more useful approach is qualitative: assess how complete the coverage is, identify the valid rows, and flag anomalies explicitly.

Beyond the headline numbers, the freshness metadata is one of the most actionable fields in the entire file. The real-yield snapshot is dated 2026-06-19, the REIT snapshot is dated 2026-06-06, and the full fetch date is 2026-06-20. That timing gap suggests readers are looking at a recent pull, but not all sector-level calculations have been populated into the final table. In research terms, the sector frame exists, while several summary layers remain unfilled.

Top Performers Table and Analysis

This section presents the full top-yield table exactly as provided by the source data. In this case, the list is empty. That absence is itself a finding because the article topic centers on data centre S-REITs, yet the designated top_10_by_yield array contains no entries.

A conventional top-performer discussion is therefore not possible from the explicit top_10_by_yield table. Analysis instead has to infer ranking patterns from the broader REIT roster included under reits_in_sector. That roster is not a substitute for the missing top-10 table, but it does reveal how the coverage behaves once individual instruments are inspected.

Zooming into the individual entries, the most relevant name to the article’s target keyword is AJBU.SI, Keppel DC REIT, the only row in the provided roster with sub-sector listed as Data Center. Its current yield is 4.52, its five-year average yield is 4.181, its NAV premium/discount is 34.07, its Distribution Safety Score is 25, and its five-year distribution growth is -14.254. Distribution Safety Score refers here to a 0-100 scale where higher indicates stronger payout coverage; this dataset shows observed values of 0 or 25, which implies limited dispersion within the current file rather than a full sector spectrum. NAV premium/discount measures how far market pricing sits above or below reported net asset value, with positive values indicating a premium and negative values indicating a discount.

That single pure data-centre-tagged entry creates an immediate contrast with the rest of the roster. The highest stated current yields in the file belong to non-data-centre sub-sectors: CRPU.SI Sasseur REIT at 9.23, A7RU.SI ARA Hospitality Trust at 7.73, M1GU.SI Sabana Industrial REIT at 7.63, A17U.SI CapitaLand Ascendas REIT at 7.59, and UD1U.SI IREIT Global at 7.23. Those figures cannot be called the official top 10 for the Data Centre sector because the top_10_by_yield array is empty, but they do show that the available constituent universe is led by retail, hospitality, industrial, and office records rather than the one data-centre-labeled REIT.

The data shifts when viewed through valuation rather than payout. AJBU.SI’s 34.07 NAV premium stands out against several deeply discounted names in the same roster, including UD1U.SI at -55.09, Q5T.SI at -35.23, TS0U.SI at -36.61, J85.SI at -45.0, and OXMU.SI at -69.52. OXMU.SI also carries an anomaly note on both valuation and growth. The file explicitly flags an extreme NAV discount of -69.5% and an extreme five-year distribution growth reading of -48.0%, stating these may reflect stale NAV data, illiquid market, or structural factors, and that the growth swing may also reflect one-time events or base effects. Those warnings matter because extreme values can distort pattern recognition if treated at face value.

That pattern breaks down when continuity is examined. Several names carry long distribution histories: A17U.SI shows 22 years of continuous distributions, T82U.SI shows 20, F34.SI shows 20, and a large cluster including A7RU.SI, C38U.SI, HMN.SI, P40U.SI, J85.SI, M44U.SI, K71U.SI, and C2PU.SI shows 19. By contrast, AJBU.SI has 12 years of continuous distributions. The shorter record does not imply weakness by itself; it simply places the data-centre entry in the middle of the continuity range found in this roster.

Cross-referencing with safety metrics reveals another layer. Among the available rows, AJBU.SI carries a Safety Score of 25, matching M1GU.SI, A17U.SI, C38U.SI, HMN.SI, O5RU.SI, K71U.SI, OXMU.SI, C2PU.SI, and F34.SI. Many others show 0, including CRPU.SI, A7RU.SI, UD1U.SI, ME8U.SI, Q5T.SI, TS0U.SI, N2IU.SI, J85.SI, M44U.SI, BUOU.SI, and T82U.SI. Since the scale is 0-100 and only two levels appear in this file, the score is more useful as a coarse filter than as a precision ranking.

Viewed through a five-year lens, the roster also mixes positive and negative distribution growth paths. A17U.SI records 12.875, HMN.SI records 7.345, Q5T.SI records 14.95, K71U.SI records 5.055, and J85.SI records 0.69, while AJBU.SI posts -14.254. Several others register moderate to steep declines, including UD1U.SI at -13.689, T82U.SI at -6.876, C2PU.SI at -6.934, and F34.SI at -9.094. In that context, the lone data-centre-tagged REIT combines a below-roster-top current yield with a relatively rich NAV premium and negative five-year distribution growth.

Country Distribution Within Sector

The country breakdown table in the sector-level source field is not yet covered. The explicit country_distribution object is empty, so a formal country-count table from that field cannot be reproduced with values.

Still, the constituent roster offers one directional observation: every listed entry under reits_in_sector carries country name Singapore. That does not replace the missing country_distribution table, but it indicates that the current sector mapping in this file is Singapore-centric at the instrument listing level. The concentration is total within the provided roster, spanning names such as Keppel DC REIT, CapitaLand Ascendas REIT, Mapletree Industrial Trust, Keppel REIT, and many others, all listed with Singapore as country_name.

Stepping back to the aggregate level, that concentration has structural implications. Singapore REITs often serve as regional wrappers for property exposure that extends beyond the domestic market, and the geography_focus fields in this roster support that point. The file includes China-focused, US-focused, Europe-focused, Pan-Asian, Global, Singapore-focused, Singapore/US, Singapore/Japan, and North-Asia exposure labels. So while the listing country appears singular in the available rows, the underlying asset focus is geographically mixed.

The picture changes at the geography-focus level. Pan-Asian appears repeatedly across A17U.SI, HMN.SI, P40U.SI, N2IU.SI, J85.SI, M44U.SI, O5RU.SI, J91U.SI, AJBU.SI, and F34.SI. Singapore-focused appears in M1GU.SI, C38U.SI, TS0U.SI, T82U.SI, Q1P.SI, ACV.SI, and A68U.SI. China-focused appears in CRPU.SI, CWBU.SI, and RF7U.SI. US-focused appears in A7RU.SI, OXMU.SI, and SK6U.SI. Europe-focused appears in UD1U.SI and Q5T.SI. That means the sector’s country listing concentration does not equate to a single-market revenue profile.

For readers examining regional income structures, the current dataset therefore says two things at once: formal country distribution is data not available at the sector-summary level, and the constituent-level evidence points to Singapore listings acting as a regional conduit for multiple underlying markets.

REITs in This Sector

This is the most data-rich section because the source block supplies a full reits_in_sector list. The table below includes all entries exactly within the available fields. Aristocrat status is shown as a true/false flag from the dataset; when first mentioned, aristocrat status refers here to a derived label for sustained distribution history within Finance Pulse Research’s framework, and this file identifies only whether the status is present, not the full qualification rule.

Switching from yield to valuation, the spread is wide enough to overshadow the article’s AI-demand theme. A7RU.SI carries a 286.36 NAV premium, and the file explicitly attaches an anomaly note: extreme NAV premium of 286.4% may reflect stale NAV data, illiquid market, or structural factors. At the opposite end, UD1U.SI at -55.09, OXMU.SI at -69.52, and J85.SI at -45.0 show very deep discounts, with OXMU.SI also flagged for the growth anomaly already noted. C2PU.SI’s 56.58 premium is also marked as an extreme NAV premium anomaly, again with the caution that stale NAV data, illiquid market, or structural factors may be involved.

By contrast, the data-centre-specific row AJBU.SI is neither the highest-yielding nor the most discounted entry. Its defining combination is different: current yield of 4.52, five-year average yield of 4.181, NAV premium of 34.07, Safety Score of 25, 12 years of continuous distributions, and five-year distribution growth of -14.254. That places it closer to a premium-valued, lower-current-yield profile within this mixed roster.

Another useful divide appears between complete and incomplete records. Eight names show current yield, NAV, and growth fields as null: Q1P.SI, RW0U.SI, J91U.SI, CWBU.SI, RF7U.SI, SK6U.SI, ACV.SI, and A68U.SI. Each also shows 0 years of continuous distributions in this file. That pattern suggests coverage gaps or inactive history fields rather than meaningful zero-income interpretation. Analysis therefore treats them as incomplete records, not as evidence of nonexistent distributions.

From a sector-purity standpoint, the table also reveals that the “Data Centre” bucket is currently broader than its label. Only AJBU.SI is tagged Data Center, while the rest span Retail, Hospitality, Industrial, Office, Logistics, Diversified, and Healthcare. Readers seeking a cleaner comparison set may find it useful to cross-check this article against data centre REIT sector data, Singapore market dashboards, distribution safety framework, NAV discount explainer, and REIT income screens.

Data Sources and Methodology

This analysis uses the exact figures in the supplied Finance Pulse Research data block for the Data Centre sector. Data freshness fields show a real-yield snapshot date of 2026-06-19, a REIT snapshot date of 2026-06-06, and a fetch date of 2026-06-20. Those timestamps indicate recent retrieval, but they do not guarantee full population of all sector-level summary tables.

Known gaps are significant and have been handled explicitly. Average nominal yield and average real yield for the sector are data not available. The sector yield-distribution and real-yield-distribution fields show count of 0 and no summary values. The top_10_by_yield array is empty. The country_distribution object is empty. Rather than infer missing values from the REIT roster, the article marks those fields as not yet covered or data not available.

Anomaly handling follows the notes embedded in the source. A7RU.SI, UD1U.SI, OXMU.SI, and C2PU.SI carry explicit anomaly annotations on NAV, and OXMU.SI also carries an anomaly annotation on five-year distribution growth. These have been discussed with the same caution stated in the file: extreme values may reflect stale NAV data, illiquid market, structural factors, one-time events, or base effects. Methodological context is available through REIT methodology notes.

Related Analyses

Readers comparing digital infrastructure with broader listed property income can extend this review through data centre REIT sector data, Singapore REIT analytics, yield factor research, NAV discount explainer, and cross-border flow coverage. Together, those pages add valuation, payout, and market-structure context that the present sector file only partially captures.

This analysis is based on publicly available market data and derived
metrics calculated by Finance Pulse Research. Finance Pulse Research
is a data analytics publisher. Content is for informational and
educational purposes only. Nothing herein constitutes investment
advice, a recommendation to buy or sell any security, or an offer of
any kind. Data as of 2026-06-20.

Finance Pulse Research builds open data analytics for Asian dividend markets — real yields, REIT NAV discounts, and foreign-flow signals across 11 countries. Stack: FastAPI + Next.js + Postgres + Celery, with data from yfinance, FRED, World Bank, and direct exchange feeds. More at finance-pulse24.com.