Denver multifamily investment guide

Everything else in this series turns on what an owner may charge. St. Paul caps increases at 3 percent. Texas permits a municipality to impose a cap after a hurricane. The Carolinas, Georgia, Tennessee and Virginia preempt the question entirely. Colorado is the one market we cover that went the other way, and an underwriting built on the usual checklist will miss it completely.²

On rent, Colorado is permissive and has not moved. Section 38-12-301 of the Colorado Revised Statutes declares that "the imposition of rent control on private residential housing units is a matter of statewide concern; therefore, no county or municipality may enact any ordinance or resolution that would control rent on either private residential property or a private residential housing unit." The source note on that section records its last amendment in 2010. Bills to repeal the preemption have been introduced more than once and have not become law. Subsection (2) preserves genuinely voluntary rent-limiting agreements tied to a permit, and subsection (4) makes clear those must stay voluntary: a local government "may not deny an application for a development permit ... because an applicant for such a permit declines to enter into an agreement to limit rent."⁵

One further statewide item belongs in any Colorado model and is set out in our Colorado Springs guide rather than repeated here. Colorado does not give an owner a fixed assessment rate: the state recomputes the residential rate annually, and a single building carries two of them at once, 6.80 percent against local government levies and 7.05 percent against school district levies for 2026.⁴ Do not hold the assessment rate constant across a Denver hold period either.

On possession, Colorado changed the rules underneath the entire value-add model. House Bill 24-1098 created Part 13 of Article 12, and its operative sentence is Section 38-12-1303(1): a landlord "shall not serve a notice to terminate tenancy or a demand for possession or otherwise proceed with an action for unlawful detainer ... unless there is cause for the eviction." No-cause nonrenewal is gone. The lease simply ending is no longer a reason a tenant has to move.

Two boundaries decide how much this matters to a given asset, and both cut against the institutional buyer.

The owner-occupied carve-out at Section 38-12-1302(1)(b) reaches a single-family home, a duplex or a triplex, and only where the owner lives there, and it expressly does not apply where the property "is not a multifamily property of four or more dwelling units." Read the double negative carefully: at four units, the exemption is gone. Every conventional multifamily asset in this metro is inside Part 13.²

The other boundary is the useful one. Section 38-12-1302(1)(e) provides that Part 13 does not apply to "a residential tenant who has not been a tenant of a residential premises for at least twelve months." Protection attaches at the twelve month mark. A tenant nine months into a lease is outside the statute. That single fact converts average length of tenancy from a soft operating metric into the most important underwriting variable in a Colorado repositioning plan, and it is not a number that appears on any rent roll summary we have ever been handed.²

• Rent control remains preempted statewide, and that section has not been amended since 2010.²
• Cause is now required before a notice to terminate tenancy may be served.
• The owner-occupied exemption stops at a triplex. Four units and up are fully covered.
• Protection attaches once a tenant has been in place twelve months.
• A rental agreement cannot waive any of it. Section 38-12-1305 makes such a provision void and unenforceable.

Ask a Colorado sponsor for the distribution of tenancy lengths, not the average. Every resident past twelve months is a resident you cannot ask to leave without a statutory ground.

The policy analysis would matter less if the economy were carrying the market. It is not, and this is the softest employment picture we have looked at in a metro of this size.

Total nonfarm employment was 1,657,200 in July 2026, up just 0.3 percent over twelve months, on a labor force of 1,759,400. Unemployment rose to 4.1 percent from 3.9 percent in June, one of only two markets in our coverage where the rate moved up rather than down.¹

Six of the ten reported sectors contracted. Information fell 7.7 percent, financial activities 3.5 percent, other services 2.8 percent, manufacturing 2.3 percent and trade, transportation and utilities 1.2 percent. Professional and business services, the largest sector in the metro at 318,300 jobs, fell 0.2 percent. The growth was concentrated in education and health services at 3.7 percent, leisure and hospitality at 3.4 percent, construction at 2.7 percent and government at 1.4 percent.¹

We would flag the composition rather than the headline. Denver’s identity as a market rests on a professional and technical employment base, and both of the sectors that define it, professional services and information, are flat or falling. Education, health, hospitality and government are doing the work. That is a serviceable floor under rents and it is not the growth thesis that Denver is usually sold on.

Population is still rising, which matters more for housing demand than payroll does in any single year. Denver County reached 740,613, up 3.5 percent from its 2020 base, and Arapahoe County 673,820, up 2.9 percent.¹²

We would buy here, and we would underwrite it differently from anywhere else in this series. The affordability gap gives this market a durable renter base that does not need the labor market to cooperate. What it does not give you is the operational freedom that most value-add pro formas quietly assume. Our method is set out in how we evaluate a market, and the full ranked list is on the markets index.

• What is the distribution of tenancy lengths, and what share of residents are past twelve months?
• Does the business plan require vacating occupied units, and on what statutory ground?
• If the plan relies on the substantial repairs ground, is the scope over or under a hundred and eighty days, and has the first right of refusal been modeled?
• Are any of the planned repairs ones that cure a habitability condition, which would remove the ground entirely?
• Has ninety days of notice at the existing rent been carried in the downtime assumption for every unit?
• How much of the submarket employment sits in professional services or information, both of which are contracting?
• What rent growth is assumed given that six of ten sectors shrank and unemployment rose?
• What share of the projected return comes from operations rather than the exit?

The Colorado question is not what you may charge. It is whether you can get the unit back, on what ground, and how many days of notice at the old rent that costs you.