The self storage market spent the first half of 2026 convincing itself that rates had settled into something workable. Lenders were quoting, buyers were closing, and the phrase “elevated but stable” showed up in nearly every quarterly summary. That framing did not survive September. The 10-year Treasury yield, which sat in the mid 4.4% range in early summer, touched 5% on September 14, held it after the Federal Reserve raised rates two days later, and has traded between roughly 5.2% and 5.3% since. It is the highest sustained level since 2007. The 30-year bond is near 5.6%, a level not seen in nearly two decades. The 2-year, which carries the market’s view of the Fed, is at 4.84%.

For a sector that prices assets off long-term debt costs and builds them with floating-rate construction loans, an 80 basis point move in the benchmark in three months changes what a buyer can pay, what a lender will size, and whether a project that penciled in June still pencils today.

Key takeaways

• The 10-year Treasury crossed 5% in mid-September and has held between roughly 5.2% and 5.3% since the Fed’s first hike since 2023. A weak September jobs report lowered the odds of an October hike but did not bring the long end back down.
• Institutional storage cap rates in the low 5s now sit at or below the risk-free rate. Quoted five-year storage loans moved from 6.96% to 7.46% in three weeks, putting any deal priced below a 7.5% cap rate in negative leverage.
• Expect retrades and broken contracts through the fourth quarter, a reopened bid-ask spread that closes when loan maturities force sellers to move, and the REITs gaining share as the private market’s lender.
• Development was already contracting to the smallest delivery slate since 2016. The rate move stalls most unstarted projects, which is bad for pricing now and good for the fundamentals of existing facilities in balanced markets through 2028.

Why yields are rising

Several forces are pushing in the same direction at once, which is why strategists are treating this as a shift in the trading range rather than a spike to be faded.

Inflation has stopped cooperating. Core PCE, the Fed’s preferred measure, is running at 3.3%, and the committee has pushed its own timeline for reaching 2% out to 2029. Energy is the immediate culprit: the standoff over the Strait of Hormuz has kept Brent crude above $100 a barrel, with the price at $102 on October 2 and up about 7% over the past month despite talk of coordinated reserve releases.

The Fed is hiking again. On September 16 the committee raised the federal funds target range by 25 basis points to 3.75% to 4.00%, its first increase since 2023, on a unanimous 12 to 0 vote. Chairman Warsh said inflation “is too high and has been for too long,” and a majority of the committee pencilled in one more increase before year end. Markets briefly priced two more by spring 2027.

Then the labour market blinked. The September jobs report, released October 2, showed just 29,000 jobs added against a forecast of 84,000, with the prior two months revised down by a combined 60,000 and July now showing a loss. Unemployment rose to 4.2%. Yields dipped on the day and the odds of an October hike fell sharply. The honest reading is that the Fed is now caught between inflation it has not beaten and a job market that is weakening, and that the bond market has not decided which problem wins. What it has decided is that the long end of the curve is not coming back to the low 4s on the strength of one soft payroll print.

Supply of bonds is heavy and demand is thinner. The federal deficit is running near 6% of GDP. Technology companies raised more than $350 billion in U.S. credit markets through August to fund data centers, more than double what they raised in all of 2025, and that paper competes with Treasuries for the same institutional capital. Japanese 10-year yields have crossed 3% for the first time in three decades, giving Japanese institutions less reason to send money abroad. The selloff is global: U.K. 30-year gilts hit their highest yield since 1998 and German 10-year bunds their highest since 2009. This is a repricing of the term premium, the extra compensation investors demand to lend for long periods, and term premium moves tend to be sticky. The practical point for real estate investors is that a 10-year in the low 5s is now the base case for underwriting, not a stress scenario.

What it means for transactions

Storage came into this move with more momentum than most property types. Marcus & Millichap’s second half 2026 national report had transaction counts up nearly 20% year over year through June and dollar volume up almost 50%, although activity remained about 30% below the 2022 peak and deals above $20 million were down 65% from it. Trades between $1 million and $10 million make up roughly 80% of transactions. The average cap rate across the sector was 6.59%, and pricing per square foot sat about 20% below 2022 levels.

The public REITs have been buying at much tighter yields. Public Storage put its year-to-date acquisitions in the low 5% cap rate range. CubeSmart valued a 15-property portfolio contributed to a joint venture at a mid 5% cap rate. SmartStop bought three Spartanburg properties at 5.9% and describes its stabilized U.S. targets as mid 5%.

Set those numbers against a 10-year above 5.2% and the problem is stark. A low 5% cap rate on an institutional-quality storage asset now sits at or below the risk-free rate, where the spread has typically run well above 100 basis points. Either the risk-free rate comes back down, or cap rates on the best assets move up, or buyers accept a negative spread because they believe in NOI growth that the REITs’ own numbers are not delivering. Public Storage’s same-store NOI fell 1.9% in the second quarter and CubeSmart’s fell 0.7%. Extra Space and NSA grew, but largely through expense discipline rather than revenue. And the street rate the private owner competes on is still falling: Yardi Matrix has national advertised rates down 1.9% year over year in August, a steeper decline than June or July, with only four of the top 30 metros showing increases.

For leveraged private buyers the math is harder still. Best-case quoted rates on stabilized self storage permanent loans were 7.46% for five-year terms and 7.66% for ten-year terms as of October 3, with bridge money at 9%. Three weeks earlier the same lender sheet showed 6.96% and 7.19%. A buyer paying a 6.5% cap rate and borrowing at 7.5% is in negative leverage from day one, and a buyer at the sector average of 6.59% is barely better. That was true at the margin in the spring. The September move made it structural for anything priced below a 7.5% cap rate.

Colliers’ Steig Seaward noted in June, when the 10-year was still around 4.5%, that transaction cap rates already exceeded modeled levels and that “cap rates may have further to run if Treasury yields remain elevated.” Yields have since risen another 70 to 80 basis points.

The effects will show up in three places.

Deals under contract get retraded or die. A buyer who locked a rate in July will close. A buyer with a term sheet but no lock is now looking at a coupon 50 to 70 basis points higher than modeled. Some will ask for a price adjustment, some sellers will grant it, and others will let the deal fall over and remarket in the spring. Expect an uptick in busted contracts through the fourth quarter and a corresponding flow of re-listed assets.

The bid-ask spread reopens. Sellers anchor to second-quarter broker opinions of value. Buyers reset to today’s debt. Colliers described the dynamic as one that “widens bid-ask spreads by increasing financing uncertainty, which is slowing deal volume and delaying price alignment.” Historically the gap closes when sellers move, not buyers, and the sellers who move first are the ones with loan maturities. Loans written in 2021 at coupons in the 3s are rolling into a 7.5% refinance market, and those owners face a cash-in refinance or a sale. That is where the next round of realistic pricing will come from.

Cash and low-leverage buyers gain share. The REITs buy with equity and unsecured debt priced off their own credit, not a bank’s storage loan sheet. Extra Space originated $141 million of bridge loans in the second quarter and carries roughly $1.5 billion in outstanding bridge balances; Public Storage’s lending platform has grown to $173 million. When bank debt is expensive and scarce, the operators with the deepest balance sheets become the private market’s lender, and they get paid twice: once on the loan and again on the option to buy the asset when it matures.

For due diligence, the immediate implication is that underwriting assumptions from three months ago are stale. Exit cap rates should be widened, not held flat. Debt assumptions should reflect quoted coupons, not a rate a lender mentioned in May. Rent growth should be stress-tested against street rates that are still declining in most Sunbelt markets, because a buyer paying a thin or negative spread over Treasuries has no room for a revenue miss. And the property tax line should reflect the buyer’s purchase price, not the seller’s bill, because a reset that was survivable at a 6% cap rate is not survivable at 7.5% debt.

What it means for development

The supply story going into this move was already one of contraction, and the move accelerates it.

Deliveries in the first half of 2026 totaled 27.8 million square feet, the lowest first half since 2014. Full-year completions are expected near 53 million square feet, the smallest slate since 2016. Yardi Matrix reported first-quarter construction starts running 29% below the prior-year pace, and its forecast steps down to roughly 45 million square feet in 2027 and under 39 million in 2028. The national pipeline under construction stands at 54.7 million square feet.

Almost none of that pipeline was underwritten with a 10-year above 5% and a Fed that has started hiking. Construction loans float over SOFR, and the front end of the curve moved with the Fed on September 16. Every project under construction saw its interest carry rise within the month. Interest reserves sized in 2024 or 2025 on the assumption of rate cuts will run short, and developers will be asked to fund the gap from equity.

Three categories of projects are affected differently.

Projects already vertical will finish. A developer with steel up and a lender funded to 60% of budget does not walk away. The cost is a longer lease-up at lower rents than modeled and a stabilized value that may not support the takeout loan. Expect more of these to reach the market in 2027 as lease-up sales priced to the debt problem rather than the asset. For buyers, that is the opportunity: a half-leased project with a maturing construction loan will trade at a discount to replacement cost that was not available in a lower-rate environment.

Projects entitled but not started will mostly pause. This is the largest category and where the 2027 and 2028 supply forecasts are most likely to fall further. A developer facing an all-in construction coupon in the 9% to 10% range, a loan-to-cost cap around 60% from a bank that has already trimmed its storage exposure, and an untrended yield on cost that needs to clear 8% to justify the risk will shelve the project unless the site is exceptional. Land sellers holding out for 2022 pricing will find the buyer pool for storage sites has shrunk again.

Projects in markets with negative rent growth are done for this cycle. Tampa, Orlando, Atlanta, Charlotte, Cape Coral and parts of Texas still show declining street rates as they absorb the 2022 to 2024 delivery wave. Adding a rate shock to a market with falling rents and a multi-year absorption timeline takes feasibility from marginal to negative.

The counterintuitive point is that the rate move is good news for owners of existing stabilized storage in balanced markets. The next three years will bring the least new competition since the early 2010s. That supports occupancy, supports rate, and eventually supports value once the capital markets settle. The near-term pain is in pricing; the medium-term reward is in fundamentals.

The short to midterm picture

For the rest of 2026, the likely outcome is a slower fourth quarter than the first half suggested. Deal counts will hold up below $10 million, where buyers are often local operators using SBA or relationship bank debt and sellers are estates or retirements that transact regardless of the rate environment. Institutional volume will thin as bid-ask spreads reopen. Broken deals will recirculate. Lease-up projects with 2027 loan maturities will begin quietly testing the market.

Through 2027, the path depends on which of the Fed’s two problems wins. If the September jobs report is the start of a real slowdown, the hiking cycle ends at one or two moves, the front end of the curve comes down, and the 10-year drifts back toward 4.75% as growth fears replace inflation fears. Storage would resume the gradual thaw it enjoyed in the spring, though from a lower price base than sellers currently expect. If instead oil stays above $100, core inflation holds above 3% and the labour market merely cools rather than cracks, the committee delivers the additional hike it has signalled, the term premium keeps building, and a 10-year at or above 5% becomes the baseline. In that case cap rates for institutional-quality storage move toward 6% and private-market cap rates toward 7.5% and above, and the sector spends 2027 repricing rather than recovering.

Either way, the supply side is already decided. The projects that will deliver in 2027 and 2028 are the ones under construction now, and there are fewer of them than at any point in a decade. Development restarts when rates fall or rents rise enough to restore yields on cost, and neither is imminent.

For buyers, the discipline is simple to state and hard to practice: price to today’s debt, not the debt a lender quoted in the spring, and pay attention to sellers whose loan maturities give them a reason to meet the market. The best storage acquisitions of the last cycle were made when financing was scarce and sellers were tired. The conditions forming now are not identical, but they rhyme.