Loss-to-Lease in Multifamily Underwriting: The Gap Interviewers Plant on Purpose

If you build multifamily models for interviews, loss-to-lease is the line that separates people who have actually underwritten a deal from people who memorized a GPR formula. Brokers love quoting market rents. Your model has to start from what tenants are paying today, then earn its way to market over turnover. Get that wrong and every DSCR, going-in cap, and Year 1 cash flow after that is garbage.

What loss-to-lease actually is

Loss-to-Lease ($) = Market Rent − In-Place Rent   (when market > in-place)

Loss-to-Lease (%) = (Market Rent − In-Place Rent) ÷ Market Rent

In-place rent is what the rent roll shows for occupied units right now. Market rent is what a newly leased identical unit would fetch today. When in-place sits below market, you have positive loss-to-lease and a value-add story: mark rents to market as leases turn.

When in-place sits above market (common in office renewals, sometimes in multifamily after a boom), renewals step down. Model that explicitly. Do not assume flat renewals.

Why brokers blur this

An OM that leads with market GPR makes the going-in yield look better and the upside look automatic. Your job is to rebuild collections from the rent roll. If you cannot explain the gap, you cannot underwrite the business plan.

Where it lives in the cap rate bridge

Gross Potential Rent (at market)
− Loss-to-Lease (in-place discount)
− Vacancy / credit loss
+ Other income
= Effective Gross Income
− Operating expenses
= NOI

Some models start GPR at in-place and grow into market. Others start at market and subtract a loss-to-lease line. Either works if you are consistent. Mixing both is how you double-count.

The annualization trap

This is the interview plant. Broker OM says:

• 200 units at $1,500 market rent → $3.6M market GPR
• Current occupancy 92%
• Average in-place rent $1,275 (15% below market)

If you take $3.6M × 92% as Year 1 collections, you just assumed every occupied unit is already at market. They are not. Year 1 collections look more like:

Occupied units × in-place rent × 12
+ new leases / renewals that mark to market during the year
− turnover vacancy during the burnoff

A cleaner interview approach:

1. Start Year 1 with in-place economic occupancy math
2. Assume a turnover rate (say 40% to 55% of units turn per year)
3. New leases and renewals step to market (or a % of market if you are conservative)
4. Show loss-to-lease shrink as the rent roll marks up

Physical occupancy of 95% with a fat loss-to-lease is still an economic problem. Physical and economic occupancy are not the same thing.

Worked intuition

Suppose 100 occupied units:

• Market rent $1,500 → market monthly rent roll $150,000
• In-place $1,275 → in-place monthly rent roll $127,500
• Monthly loss-to-lease = $22,500 (~15%)

If 50% of units turn this year and capture full market on turn, you do not jump to market overnight. You earn it through the year, with vacancy between move-out and move-in. Aggressive OM cases that "stabilize next quarter" deserve skepticism.

How fast does loss-to-lease burn off?

There is no universal answer. Interview convention for multifamily value-add:

Aggressive OM: full mark-to-market in 12 months with 95% occupancy the whole way. Your IC case should haircut that. Ask what happens if turnover slows or renewals only capture half the gap.

Loss-to-lease vs vacancy

Vacancy is empty units. Loss-to-lease is occupied units paying below market. Both reduce EGI relative to market GPR. Interviewers love asking you to explain the difference in one sentence.

"Vacancy is physical emptiness. Loss-to-lease is an economic discount on occupied units. You can be 96% occupied and still have a double-digit loss-to-lease problem."

Cap rates and valuation

Going-in cap should use in-place or trailing NOI, not market NOI, unless the asset is already marked to market. Using market NOI in the denominator of a going-in cap makes a value-add deal look like a core deal.

Exit cap usually capitalizes forward stabilized NOI after the business plan is done. That is a different question from going-in. The business plan story is: buy on in-place economics, create NOI by burning off loss-to-lease and fixing occupancy, exit on stabilized forward NOI.

Debt sizing interaction

Lenders often size on in-place or stressed Year-1 CFBD, not the fully marked market NOI. If your model sizes debt on market NOI while the lender sizes on in-place, you will understate equity need. Always ask which NOI the debt test uses.

Quick interview walkthrough

"I start from the rent roll. In-place rent times occupied units sets current collections. Market rent sets the upside. Loss-to-lease is the gap. I burn it off with a turnover schedule, not a flat vacancy haircut on market GPR. Year 1 NOI stays honest, DSCR stays honest, and the value-add story shows up as NOI growth over the hold instead of as a fake Day 1 yield."

Common mistakes

Interview prompts

"What is loss-to-lease?"
"The gap between market rent and in-place rent on occupied units. It is an economic discount, not vacancy."

"How would you underwrite a 15% loss-to-lease OM?"
"I rebuild Year 1 from the rent roll, assume a turnover rate, mark turns to market with a renewal haircut if needed, and show NOI ramping as the gap burns off. I do not take market GPR times occupancy as Year 1."

"How does this affect going-in cap?"
"Going-in should use in-place or trailing NOI. Using market NOI makes a value-add asset look core."

"What if renewals cannot take full market?"
"Then burnoff takes longer and Year 1 to Year 3 NOI is lower. I sensitize capture rate."

Practice

Build a multifamily pro forma where in-place and market rents are different inputs, then force yourself to get Year 1 EGI right before you touch exit. That habit shows up on every REPE modeling test that uses a real rent roll. Sunbelt-style cases are a good place to make loss-to-lease a first-class assumption instead of a buried plug.