Real Estate's Rate Selloff Was a Pricing-Power Test: Why Host Rose While CBRE and BXP Fell

When the 10-year Treasury yield climbed back above 4.8% this week, real estate took the hit — but not as one market. The scoreboard split down the middle: CBRE and BXP on the losing side, Host Hotels & Resorts and Fermi in the gainers. If you read that as "rates are bad for real estate," you have caught the headline and missed the signal. The split is the signal. It is a test of which businesses can reprice for a higher-rate world and which are quietly hostage to a cap rate.

Rates hit real estate through a mechanism most investors only half-remember. A commercial property is valued as its annual net operating income divided by a capitalization rate, and that cap rate tracks long-term yields. When the 10-year rises, cap rates rise, and the value of a standing building falls even if its rent never changes. That is why the worst damage this week landed on the two names with the longest, most rate-exposed cash flows.

The office and the servicer take the long end

BXP, the office landlord concentrated in gateway markets, is the textbook case of duration. Its tenants sign leases that run for years, so its income is locked in at today's rents while its asset values trade against a moving cap rate. It is also the most leveraged name here — about two dollars of debt for every dollar of equity and $15.6 billion of net debt — so a higher discount rate pulls on both sides of the balance sheet at once. Over the past week it fell roughly 5%. Its fundamentals were already thin: return on equity near 5.7% and return on invested capital near 4.2%. In sector terms, this is a weak report card on top of a rate squeeze.

CBRE is a different animal wearing the same loser sticker. It is a real-estate services business, not a landlord, but its earnings are a function of deal flow — transactions, leasing, and debt placement that go quiet when financing gets expensive and cap rates stay sticky. The company's actual operating strengths look fine: revenue up about 15% year over year, roughly 15% return on equity. Yet the stock trades at about 54 times forward earnings, because the "E" in that multiple is depressed by a weak transaction market. AInvest's aggregate signal still labels CBRE a Buy with a strong fundamental score. So on fundamentals this is a stretched grind, not a broken franchise — but neither is it a stock that earns its keep from collected rent. Both "losers" fell for the same reason, and only one of them resents it.

Host: the short-duration rebuttal

Now the name that shouldn't have been a gainer if you believe the lazy version of the story. Host Hotels & Resorts owns luxury and upper-upscale hotels, and hotels are the anti-office: no multi-year leases, no locked-in rent. Room rates are set nightly against today's demand, which means the business reprices with the market instead of against the market. Rate hikes and a strong economy are not a threat to its earnings — they show up as higher room rates.

And that is precisely what the numbers show. Host reported second-quarter comparable RevPAR up 7.0%, then in early September raised its full-year 2026 RevPAR growth guidance to 4.75%–5.25%. It trades at roughly 15 times earnings with a dividend yield above 7%, return on equity around 15.5%, and free cash flow growing. It is up about 25% year to date and roughly 1% just this week while the rate-sensitive names fell. In one stock you get the collective strength most investors think a REIT can't have: real earnings, a fat yield, and short-duration growth that a higher-rate backdrop doesn't break. Among these four, Host is the one that checks value, growth, and profitability together — and it is also the one that held up.

Fermi: not a rate story at all

That leaves the odd one out. Fermi, the AI-data-center developer, is not rising because rates helped it; rates are essentially irrelevant to it because there are no earnings to discount. It burned through about $1.2 billion of free cash flow over the trailing year, and returns on invested capital are deeply negative. It is a development story running on a specific catalyst — a first binding 15-year, roughly $6.5 billion lease at its Texas data-center campus — and on momentum after the stock collapsed about 78% over the trailing year from a post-IPO high near $37 to under $6. Its 5% swing on a given day is normal operating temperature, not a signal. Treating Fermi as the same "real estate" as Host or BXP is how a diverge-into-the-sunrise momentum chart ends up in the same sentence as a 7% dividend.

What this week's split should tell you is simple. "Real estate is getting hit by rates" is a label, not an analysis. The factor question the market was answering is which names earn their income from repricing power — Host, which reprises every night — versus which are priced off cap-rate compression and deal flow — BXP and CBRE. If rising long-term yields are your working assumption, Host belongs in the income-and-repricing sleeve of a portfolio as a partial hedge against the rate-sensitive names; BXP and CBRE are the ones you want patience on, not conviction, until the 10-year and the transaction market give them a reason. Fermi, whatever its lease pipeline, is a speculation with no earnings to anchor it — the part of this diverging sector that belongs to a sleeve only if you can afford to lose the whole bet.