The most boring chart in real estate. On purpose.

Source: Nareit REIT Industry Tracker, Q1:2026 (May 2026); underlying data S&P Capital IQ Pro and Nareit. Same-store NOI, percent change over four quarters, U.S. listed equity REITs in the FTSE Nareit All Equity REITs Index; series begins Q1 2000. reit.com — the workbook is free, public, and downloadable.
Required disclosure: index compounded by the sponsor from Nareit’s published growth rates; 2000 = 100. Nareit does not publish a cumulative NOI level index — this is an illustrative approximation, not a publisher’s series. Sector aggregates reflect a limited number of constituent REITs (manufactured homes rests on two to three companies across the period) and are listed-REIT results, not private-market results. Nareit requires written permission to reproduce; it is free and routinely granted.
Disclose alongside: self storage is normalising after a 2021 that grew 18%. In twenty-six years the two sectors’ soft patches have not coincided — which is the argument for a fund that holds both.

Source: computed from the Nareit REIT Industry Tracker, Q1:2026, same-store NOI block — 105 quarters, Q1 2000 – Q1 2026, no gaps. Every bar is derived at runtime from Figure 1.1’s index rather than typed in, so no count here can drift from the series drawn above it. The most recent quarter, Q1 2026, was +5.64% — a fourth consecutive quarter above 5.6%.
How a year is counted: twenty-six index points give 25 year-over-year steps, 2001–2025, and a step counts only if the index rose. One step is flat at the precision this article publishes — all equity REITs, 2009 to 2010, both 111.4 — and flat is not growth, so that row reads 20 rather than 21; the ranking is the same either way. The twenty-sixth year, 2000 against 1999, is Nareit’s quarterly record and not something the index can draw from its own base year, so no bar here claims it. Figure 1.3 counts the same 25 steps and carries the twenty-sixth on the same terms.
Phrasing note: the right-hand figures are simple means of quarterly year-over-year rates, which the compounded one-decimal index cannot reproduce at that precision — they stay as published literals rather than being derived. Write “averaged 4.76% year-over-year growth per quarter,” never “compounded at 4.76%.” Counted by quarter instead of by year, the same block reads 103 of 105 for manufactured homes — retail 97, all equity REITs 87, apartments 83, self storage 83, office 70 — which is the count the guess above is scored against. This figure replaces the retired “25 consecutive years · NCREIF” claim; NCREIF does not track manufactured housing, and the 25 drawn here is a Nareit-derived number that only looks like the one it retired.

Source: computed from the Nareit REIT Industry Tracker, Q1:2026, same-store NOI block — the same workbook as Figures 1.1 and 1.2. Annual rates compound the four quarterly year-over-year rates into calendar years (the Figure 1.1 construction), and every bar is derived at runtime from Figure 1.1’s index — one source, so the two figures cannot disagree.
How a year is counted: the chart plots the 25 year-over-year steps, 2001–2025. The 26th year is 2000 itself, which grew against 1999 in all four of its quarters. A reviewer who prefers the claim with zero inference writes “every year for 25 straight years (2001–2025)” — the chart supports either. Figure 1.2 counts these same 25 steps across every sector, so the two figures cannot disagree on a count.
Phrasing note: write “computed from Nareit’s quarterly data,” never “Nareit publishes annual growth” — it does not. A level-weighted annual calculation may differ by a few hundredths of a point; the sign of every year is robust.

Source: Fannie Mae, “Multifamily Product Overview: Manufactured Housing Communities,” July 2026; data as of 31 December 2025. The 3,215-loan / $24.4 billion table covers 2005–2025 vintages.
The denominator rule: $11M net loss ÷ $24.4B acquired is 0.045%; Fannie’s published 0.11% is net loss against liquidated acquisition UPB of $10.0B. Say either — never the two halves crossed. And “returned” applies only to principal that has actually run off (~$10.0B), not the full acquisition balance.
Scope: agency loan performance describes the asset class and is not attributable to, or indicative of, any Elevation offering. Past performance does not guarantee future results.

Source: StockAnalysis.com statistics pages, “Beta (5Y),” retrieved 17 August 2026; each sector average cap-weighted using market capitalisations as of the same date — manufactured housing from ELS, SUI and UMH ($29.7bn combined), self storage from PSA, EXR, CUBE and NSA ($109.3bn). Independent anchor: NYU Stern (Damodaran), “Betas by Sector,” January 2026 — R.E.I.T., 190 firms, levered beta 0.64.
Three caveats that must survive editing: (1) the indication rests on three manufactured-housing companies and four storage ones — seven in total, not a market; (2) listed REIT equity beta reflects leverage, index membership and public-market sentiment, and may differ materially from directly held property that is not marked to market; (3) providers report materially different betas for the same company on the same day, so the gap below should be read as a direction, not a decimal.
What changed, and why: until August 2026 this figure showed no beta for self storage and said in print that the two sectors were indistinguishable, on the strength of a 0.61-vs-0.60 pair attributed to Nareit’s sector indices. That pair was carried without a citation and was never measured on this figure’s method. Measured here — one source, one method, cap-weighted the same way for both — they are not close: 0.73 against 1.03, with every storage constituent at or above the highest manufactured-housing one. The comparison is now made rather than withheld, and it is made on numbers a reader can check in the same twenty seconds as the rest.

Sources: value — Green Street Commercial Property Price Index, 4 June 2026 release, the same table published in full as Figure 1.2 of “Nobody is making more of these.” Twelve property sectors plus the all-property benchmark — thirteen rows; this figure draws six of them, and the full ranking is one click away rather than reprinted here. Beta — the Figure 1.5 set immediately above, unchanged: StockAnalysis.com “Beta (5Y),” retrieved 17 August 2026, anchored by NYU Stern (Damodaran). The right-hand bars are read off Figure 1.5’s own numbers at draw time, so the two figures cannot disagree.
The fourth caveat — this one is the pairing’s own: Figure 1.5’s three caveats travel with the right-hand panel unchanged. This figure adds one. The left panel measures property value in the private market; the right measures the equity beta of one listed company per sector (three, cap-weighted, for manufactured housing). Putting them side by side is a join of two lenses on a sector, not one measurement of it — which is why they are drawn as two panels with two scales, two tick rows and two captions, and why nothing joins a row across the gap.
Value, not return: the CPPI is a price index with no income component. “Index level” and “value” are the correct words for the left panel; “return” and “total return” are not, in any caption, tooltip or summary.
What is claimed, and what is not: that on one published five-year equity measure, cap-weighted identically for both, manufactured housing is lower-beta than self storage — 0.73 against 1.03. That is a statement about listed equities over five years, not about private property values, and not a forecast. This figure previously left storage’s row blank and said the two were indistinguishable, citing an uncited 0.61-vs-0.60 pair from a different method; that has been withdrawn. What has not changed is the rule that produced it: no bar on this panel may be filled from a method other than the one in its own caption, and no row may be deleted to tidy the picture — storage stays because at 244.7 it is the closest peer manufactured housing has, and removing it would overstate the lead.
The superlative rule, carried over: highest index level is the defensible value claim. Nearest to its 2022 peak is not — mall (−1%) sits closer than manufactured housing (−8%). The drawdown column travels beside the level column here and in the data table so the flattering half never ships alone.
Permission: Green Street’s licence is internal-use by default; redistribution rights must be confirmed before external publication — unlike Nareit, this one is not routinely granted.