The Storage REIT Sector in One Number: My Weighted Big-3 Composite, Q2 2026

The sector's demand has stopped shrinking, but it is not yet earning its way back. Revenue turned positive on occupancy, not on rate. It's all small figures but directionally positive. Expenses are still growing almost 4x faster than revenue, so margins are still compressing even with NOI back above zero.

The three operators together publish full revenue, operating expense and NOI by market — PSA for 25 named markets (15 before 2025), EXR and CUBE for roughly 30 MSAs each. We normalized 70+ raw market names into canonical metros, summed sub-markets where definitions differ (e.g. CubeSmart's LA + Riverside vs Extra Space's combined LA–Riverside–Orange County), and weighted each metro's growth by each operator's square footage there.

The geography of the recovery is unambiguous. Gateway and coastal metros lead: Boston (+4.7%), Minneapolis (+5.2%), Honolulu (+4.4%), San Antonio (+4.5%), New York (+4.0%), San Francisco Bay Area (+3.5%) all posted positive weighted NOI growth in Q2 2026. The Sunbelt supply belt is still underwater: Houston (-5.8%), Tampa (-5.3%), Dallas-Fort Worth (-4.9%), Orlando (-3.4%). The pattern has been stable for four quarters: markets that never over-built are already growing again, while the 2021–22 development magnets keep absorbing new supply at the expense of rate.

The supply data says the same thing from the other side: the losers sit on heavy recent-supply bases (Phoenix carries 6.9% under construction, triple the national average; Cape Coral–Fort Myers -14.3% NOI with 5.4% UC), while the leaders are metros where construction barely registered (San Francisco 0.8%, Minneapolis 0.9%, Chicago 1.2%, Boston 1.1%). The heatmap below shows the full five-year arc — and the table adds the supply context that explains it.

• Weighted move-in volume growth: -3.2% YoY. Move-outs: -4.6%. Move-outs have fallen faster in each of the last three quarters. That gap is the entire occupancy recovery.
• Net absorption ratio (move-ins minus move-outs, divided by move-ins): 10.3% vs 8.7% a year ago. Strongest second quarter in the series.
• Why churn is slowing: existing-home sales of 4.06M in both 2024 and 2025, the slowest two years since 1995 (NAR). 30-year mortgage at 6.66% (Freddie Mac). Only 11.2% of households moved in 2024, the lowest ever recorded (Harvard JCHS). Quits rate parked at 2.0% (BLS).
• The tenant base has re-mixed toward stayers: EXR says length of stay is up about 1.5 months YoY, CUBE calls the base "particularly sticky," PSA flagged "a material reduction in churn."
• The caveat: a housing recovery lifts move-ins but also unlocks the locked-in movers. Churn rises with it.

• I track same-store NOI growth as the trigger for when the investment cycle turns. A trigger here means one pre-set, measurable signal that history says marks the turn, so the decision is made in advance instead of in the moment.
• The signal: PSA's print, the sector's longest continuous series, crossing back above +3% year over year. Every prior cross-up marked the front edge of the best buying window of its cycle (Q1 2011 is the canonical case).
• Q2 2026: composite +0.1%, PSA -2.2%. 5.2 points from the trigger.
• My April framework scores the cycle 5.99 out of 10, early-entry tier. Cross-up penciled in for the Q3 2026 print in late October, or Q1 2027 if it slips. Getting there needs about five points of acceleration in two quarters; the composite did 4.3 points in the two quarters ending Q2 2021.