The Capital Map. Where real estate money goes in 2027.

The repricing is over. Not every sector fell the same.

When rates rose in 2022, buyers and sellers stopped agreeing on price. Two years later the gap closed at a lower level: institutional property values fell 22 % from their March 2022 peak and then stopped falling.

The average hides the story. Apartments fell hard and are already recovering. Office fell the most and has not found a floor. Hotels barely moved. Pick a sector below to see how far it fell.

$1.53 trillion has to be refinanced in two years.

US commercial mortgages total about $5 trillion. Of that, $875 billion is scheduled to mature in 2026 and $652 billion in 2027. Many of these loans were written near the 2022 peak, when the average building was worth about 28 % more than now and borrowing was far cheaper.

An owner in that position cannot simply roll the loan. They need new equity, a preferred-equity partner, or a buyer. Banks hold the biggest share, and banks are cutting real estate exposure. Switch between lender and property type to see where the pressure sits.

$326 billion is waiting, and more of it is debt.

North American real estate funds ended 2025 with $326 billion of committed but uncalled capital. That is below the 2022 peak, but the mix changed: value-add and opportunistic money shrank, while debt and core funds grew.

The reason is simple. Over the past few years real estate debt funds outperformed equity funds, because a loan sits below the equity when prices fall. Drag through the years to see the shift.

Each kind of money holder reads this market differently. A pension fund wants steady income and moves slowly. A family office decides in a week but has a small team. A bank wants less real estate on its books, not more.

Where to buy: follow people, power and low new supply.

The best US markets for 2027 share three things: people and jobs keep arriving, power is available for new buildings, and developers did not overbuild in 2024–25. Apartment starts nationally fell from 547,000 in 2022 to 355,000 in 2024, but the cut was uneven.

Data centers show the power story most clearly: vacancy in the main markets is a record 1.4 %. Tap a state to see our view of its main metros.

Outside the US the logic is the same, adjusted for policy and currency. Japan offers cheap borrowing and a weak yen. India and the Gulf have people and capital arriving. Germany has a large market still working through its correction.

Lending and rescue capital pay more than their risk.

Every strategy trades return against what can go wrong. Normally the line is straight: more risk, more return. Today two strategies sit above it. Senior and mezzanine loans and preferred equity into good buildings with bad loans pay more than their risk suggests, because banks have stepped back and someone has to fill the gap.

Tap a dot to see what the strategy is and when it works.

Speed is now a return driver.

Every investment firm runs the same chain: screen, underwrite, check, write the memo, decide. At a mid-size firm the underwriting step alone takes one to three weeks. When AI does the reading and the modeling, the same team can look at 3–5× more deals and bid on the good ones first.

What to do with this

If you allocate capital: put new commitments into credit and rescue structures, keep money ready for Sun Belt apartments in the second half of 2027, and buy power, not land, in digital infrastructure.

If you run a fund or a lender: your limit for the next two years is how many deals your team can read and price well. Titleman can turn a data room into an underwriting package in under an hour, so your team spends its time deciding.