Apartment Development Is a Long Game, and Everything Follows From That

Building apartments and building condominiums can look like the same job from the outside. The same sites, the same architects, the same construction crews. But once I started looking at the business logic underneath each one, they stopped looking like variations on a theme.

The difference runs deeper than just "for rent" versus "for sale." It changes who you compete against, how you think about design, what your market analysis needs to show, and how the project earns the right to keep existing years after construction ends.

Most people treat the rent-or-sell split as a financing preference.
That framing undersells it.
The choice rewrites who your competition is, what your buildings need to do, and how your business compounds — or doesn't — over time.

The rent-or-sell decision looks like a financing choice. It's actually a business model choice — and the two don't optimize for the same things. Most developers who treat it as the former end up surprised by what the latter demands: competing against homeownership itself, designing for retention rather than buyer appeal, and treating building height as an economics problem before it's an architecture one.

That's the thread I want to pull on in this post.

The rent-or-sell decision changes the whole operating logic

Multifamily housing sits at a fork: you can build condominiums and sell them, or you can build apartments and keep them.

If you sell, you're in a for-sale product business. Your horizon is short — build, market, close sales, exit. The economics look closer to homebuilding than to income property development. You need buyers, a market that supports purchase prices, and a timeline that lets you move on to the next project quickly.

If you rent, you've entered the investment property category. You're not looking for buyers. You're looking for tenants — and then you're planning to keep the asset, use the cash flow, and eventually use the property's equity to fund future deals.

That second path — apartments as long-term held investment — is where things get structurally different. Holding isn't just a preference. It's a compounding strategy.

The leverage flywheel is why developers hold apartments

Here's the argument that explains why apartment developers don't just sell and move on.

When your first project is complete and stabilized — occupied, generating rent, covering expenses, throwing off cash — you now have a real income-producing asset on your books. That asset can be pledged as collateral. Its cash flow can support additional borrowing. Its value is no longer theoretical; it's established by actual market performance.

So when you're ready to do the next project, you're not starting from scratch. You're using the equity and income from your existing portfolio to improve the terms on the next deal. The portfolio finances the pipeline.

This is why developers often say the first deal is the hardest — not because it's the most complex, but because you have almost no portfolio collateral going in. Everything has to stand on the project's own economics. Once you've built and stabilized even one property, the dynamic shifts.

Apartment cap rate s — typically 4.6% to 6% — run lower than office (6.2% to 7.8%) or retail (6% to 7.2%). That might read as a disadvantage at first. But a lower cap rate signals that the market treats these as higher-value, lower-risk assets. Lenders prefer them. Institutional investors favor them. The income stream from apartment rent has historically proven more resilient than commercial leases during economic disruptions. Short 12-month residential leases, paradoxically, create more stable aggregate cash flow than long-term commercial leases, because the re-pricing cycle is faster and tenant concentration risk is spread across many households rather than a few business tenants.

That combination — low cap rate, high lender preference, resilient cash flow — is why apartment debt tends to be cheaper and more available than commercial debt.
The boring asset earns the better terms.

That stability is part of why apartment development became a serious institutional investment class in the 1990s, drawing REITs and high-net-worth investors in ways it hadn't before.

The obvious objection: if you can sell as condos and take the certain exit, why take on the operating risk of holding for decades?
The answer is that the compounding only works if you hold.
Selling the first project gives you cash for one next deal.
Holding it gives you collateral, cash flow, and proof-of-performance — and that combination improves every subsequent deal's terms.

Your competition includes more than other apartment buildings

Here's the thing that reframed how I think about apartment markets: apartment developers compete with homeownership itself.

When a household buys a home, they leave the rental market entirely. Every point of homeownership rate is a household that won't be renting from you. So the competitive analysis for an apartment project has to start with a wider picture than just the complexes going up nearby.

In the United States, roughly 65% of households owned their homes in 2019. In San Diego, that number was lower — about 54%. The gap matters. San Diego's lower ownership rate creates a proportionally deeper renter pool. More households are in the rental market for longer periods of their lives than in markets where ownership is more accessible or culturally expected.

That "gray market" category is the least obvious of the three and probably the most consequential to model correctly.

Deed-restricted affordable housing is the most direct constraint. In San Diego, a standard market-rate development must make 10% of its units deed-restricted affordable — renting at rates accessible to households earning 60% of area median income. Those units cost exactly the same to build as the market-rate units. The 90% market-rate units have to carry the cross-subsidy.

Rent control operates differently, but the supply effect runs in the same direction. California's Costa-Hawkins Act protects buildings constructed after 1986 from local rent control measures, which is why most new apartment construction in the state falls outside rent-controlled territory. But the political pressure is real, and developers watch every ballot cycle. The economic argument against rent control isn't that existing tenants are better off without it — often they aren't, in the short term. The problem is the supply side. A 2019 study published in the American Economic Review found that San Francisco's rent control expansion reduced rental housing supply by 15% — units converted to owner-occupied housing or redeveloped entirely.

If construction costs stay the same but the revenue ceiling drops, developers walk away. Not because they want to, but because the math stops working. The result is that the units already in the market become cheaper for the people currently in them, while new supply dries up. The households who don't already hold a controlled unit — typically younger residents, newcomers, and anyone entering the market after the freeze — are the ones the supply shortage hits hardest.
Both effects are real.
A good market analysis accounts for both.

Short-term vacation rentals distort the data in a subtler way. A property owner choosing between $2,300 per month on a standard lease or $2,300 per week on a vacation rental platform isn't making a political statement. They're making a rational economic choice. When that choice plays out across thousands of units in a market, it removes supply from the long-term rental pool — not dramatically enough to explain entire housing crises, but enough to skew the market signals that developers rely on.

Understanding the competitive landscape tells you whether a market supports a project.
The building code tells you what you can actually afford to build in it.

Building height is really a building code economics story

Walk through any growing city and you'll notice that almost every new apartment complex is a variation on the same basic form: three or four stories of wood-frame apartments sitting on top of a concrete parking podium. That pattern isn't aesthetic inertia. It's code and economics working together.

The building code classifies construction by fire resistance — five types, numbered in Roman numerals. Type 5 is wood-frame construction: combustible, but cheap. Once you go above four stories with wood framing, the code steps up to Type 3, which requires fire-resistive structural systems and exterior skin. Above eight stories, the entire building requires Type 1 construction — the most resistant and most expensive classification, typically steel and concrete throughout.

Here's the part that stuck with me: you can build four stories efficiently, or you can build eighteen-plus stories efficiently, but almost nothing pencils in between. Type 1 construction is expensive enough that the cost per unit is too high until you're at a scale — roughly 18+ stories — where you can spread those costs across enough units for the revenue to cover them. That's why city skylines have this strange gap. Lots of four-story podium buildings, lots of towers, almost nothing between 9 and 17 stories.

The low-rise podium product dominates for exactly that reason. Lenders know how to underwrite it. Contractors know what it costs. Everyone knows what the rents should look like. Architects sometimes find it repetitive — and many do — but the economics are difficult to argue with when the alternatives are more expensive or structurally impractical.

The concrete podium level (parking below, apartments above) also gets a code break. The garage structure below is Type 3, which is allowed to support the lighter Type 5 wood frame above. That combination is why "sticks over podium" is the phrase you keep hearing in multifamily development circles.

Site and design decisions start with a defensible target market

Once the product type and height are determined, every design decision downstream becomes a question of who the target tenant is. And that question cannot be answered abstractly.

The opening of the material that sourced much of this post made a sharp point about assumptions. Every developer makes them. The problem isn't the assumption — it's an assumption with no supporting rationale. "We'll charge $3,500 a month" is not a supported assumption. It's a wish. "We'll charge $3,500 a month because every comparable unit in the submarket is leasing at $3,200 to $3,500, we'll be the only new product coming online in the next 24 months, and our target demographic's income levels support that rent-to-income ratio" — that's a supported assumption.

The target market question drives everything downstream:

• Unit mix — more studios and one-bedrooms for young singles and young professionals; more two- and three-bedrooms for families
• Unit type — dual master suites work for roommate scenarios; defined bedrooms matter more for families
• Amenities — young professionals want co-working space, fitness centers, and bike storage; families want playgrounds and school proximity
• Parking — in transit priority areas, some projects provide zero assigned parking; suburban family projects may need 1.5 to 2 spaces per unit

San Diego has made a notable move on parking: developers in transit priority areas (within half a mile of high-frequency transit — buses running every 15 minutes, light rail, commuter rail) are no longer required to provide any minimum parking. The market decides. Parking can also now be "unbundled" — rented separately from units, so a tenant without a car isn't subsidizing spaces they'll never use. In a market where structured parking can cost $40,000 to $60,000 per space to build, this flexibility has real impact on whether a project pencils in dense, transit-served locations.

Every one of those decisions — unit mix, amenities, parking strategy — is a derivation, not a preference.
The target market is the input.
Everything else is output.

Designing for the right tenant gets them in the door.
Keeping them there is where the economics play out over time.

Keeping good tenants is financial discipline, not customer service

One of the counterintuitive points worth keeping: good landlords work hard to keep tenants happy not because they're generous, but because churn is expensive.

When a lease turns over, the unit may sit vacant for 30 to 60 days. There are re-leasing costs, potential repairs, and the full qualification cycle for a new tenant. In California, if a non-paying tenant needs to be evicted, the process can take 90 or more days in practice. Then there's re-leasing again after that.

An operator who keeps a good tenant isn't being sentimental. They're making a rational financial choice. Amenities, concierge service, moving assistance when tenants transfer to a larger unit in the same complex — all of these cost money, and all of them reduce the churn that erodes cash flow. The large corporate apartment operators (Avalon, Camden, and others) invest heavily in resident experience systems precisely because tenant retention, at scale, is a meaningful yield driver.

This also connects to why you see renewals that come with small rent bumps. A tenant who renews at $2,400 instead of $2,200 — on the same apartment, same cost structure — produces $2,400 of revenue for the operator at essentially zero incremental cost. No vacancy gap, no turnover work, no new qualification process.

Case study: The Platform showed what a defensible market argument looks like

The Platform in San Jose is the clearest illustration I've seen of how all of this comes together in a real project.

It was part of the Market Park Planned Community — a 120-acre redevelopment of what had been a flea market — that included 1.5 to 2 million square feet of office and around 4,000 residential units. The Platform itself was 551 apartments across three buildings on 6.5 acres. Podium product: four-story wood frame over a concrete parking garage, opened in 2019, right on a new BART extension.

The developer's market-defining story was built on actual data.

San Jose added 240,000 jobs between 2009 and 2017 — roughly 30% growth, or 3.5% per year. Median family income was around $106,000. During that same eight-year period, the region added only 51,500 housing permits. That's a jobs-to-housing ratio of about 4.6 to 1. Workers were commuting in from further and further away, and Bay Area homeownership was out of reach for most of the people those tech firms were hiring.

The original rent target was $3.25 per square foot per month. To actually make the pro forma pencil — meaning generate enough return to justify the investment, not just break even — they needed to push that to $3.40. That's not a large adjustment, but it shows how thin the margin can be even on a well-located, well-analyzed project. The increase had to be justified by the data, not just assumed. And the data — BART extension, tight vacancy, median income levels, job growth pipeline — supported it.

The city of San Jose pushed the developer to increase density, requiring 551 units rather than the 450 originally planned. For a developer trying to amortize fixed costs — land, infrastructure, parking structure — more units across the same footprint generally improves the economics. The developer ended up with a project that was denser, higher-revenue, and ultimately more defensible.

Amenities were calibrated to the specific target market. Co-working space inside the complex was aimed at tech workers who might work remotely or run early-stage companies. Bike storage and repair, resort-style pools, rooftop deck, fitness center, pet grooming — each of these was an informed choice about what a young professional or young family earning tech-sector incomes would actually pay a premium to have.

None of it was a guess. Every assumption had a rationale, and every rationale had data behind it.

A few things I'm taking away

If I had to reduce this material to the ideas most worth carrying forward:

• The rent-or-sell decision is the first and most structural choice in multifamily development — everything else follows from it
• Holding apartments long-term isn't just patience; it's portfolio compounding, because stabilized assets become collateral for the next deal
• Homeownership is a legitimate competitor; lower local ownership rates directly expand the addressable renter pool
• The gray market — affordable housing mandates, rent control, short-term rentals — creates distortions in apartment economics that have to be modeled honestly, not ignored
• Rent control reduces costs for current tenants and reduces supply for future ones; both things are simultaneously true, and a good market analysis accounts for both
• The four-story wood frame over concrete podium is everywhere because code, cost, and lender preference all converge at that form factor
• Building between 9 and 17 stories almost never pencils, because you pay for Type 1 structural costs without enough unit count to amortize them
• Every design decision — unit mix, amenities, parking strategy — is downstream of a defensible target market analysis
• Keeping good tenants is a cash flow discipline, not a hospitality preference
• The Platform case study shows what a supported market argument actually looks like: specific data mapped to a specific thesis, with every assumption linked to evidence

"When we own portions of outstanding businesses with outstanding managements, our favorite holding period is forever." — Warren Buffett, Berkshire Hathaway 1988 shareholder letter

Apartment development isn't really a construction business.
It's a compounding business dressed up as a construction business.
The building is the asset, but the long game is the portfolio.

That last point is probably the one I'll keep thinking about after this series ends. Not "here's our assumption" — but "here's our assumption, here's the data that supports it, and here's what would have to be wrong about that data for the assumption to break." That discipline is what separates a development pitch from a development plan.