Most real estate capital chases the same twenty metros. The logic is understandable — depth of market, liquidity, brand recognition — but it produces crowded fields and compressed margins. We've spent the past year building our first development in a different kind of market, and the reasoning is worth sharing.

Cambria County sits in southwestern Pennsylvania, anchored by Johnstown. It is not a growth market in the way analysts typically use that word. Population has been flat-to-declining for decades, and the median household income is well below the national average. On paper, it is exactly the kind of place capital overlooks.

What the top-line numbers miss is the composition of demand. Three fundamentals shape our thesis.

Aging housing stock. A significant share of the county's rental and owner-occupied housing predates 1970. Energy inefficiency, deferred maintenance, and floor plans designed for a different era all push the better-earning cohort of renters toward anything new. New construction in a stock this old does not need to out-compete the overall market — it only needs to out-compete what exists on the newer end of the curve.

Employer anchors that are not going anywhere. Conemaugh Health System, regional manufacturing, and the education sector provide a base of wage-earning households that need reliable housing near work. These are not speculative jobs tied to a single industry cycle.

Supply that effectively cannot respond. Land pricing, construction costs, and the risk appetite of national homebuilders make Cambria County economically unappealing for large-scale new-build activity. Local builders are active but small. When we add 32 units of purpose-built townhome rental product, the market does not absorb that and then demand the next 32 within a quarter. It absorbs slowly — which, counter-intuitively, is a feature for a staggered-delivery sponsor.

The flip side is real. A small market has thinner comp sets, fewer active buyers at exit, and less liquidity than a Tier 1 metro. We underwrite that reality directly: longer marketing periods, conservative rent growth, a baseline of owner-carry financing as an exit option rather than a reliance on institutional buyers. We would rather build a plan that works in a flat market than one that depends on a rising one.

Small-market investing is not for everyone. It requires patience, local relationships, and a willingness to do the work of understanding a place rather than importing a coastal playbook. What it offers in return is a setting where disciplined sponsorship can still generate meaningful basis advantage — because very few others are competing for the same dirt.

That's the short version of why we're in Cambria County. In future pieces, we'll walk through how we underwrite against these conditions, how we structured the project timeline to match the market's absorption profile, and what we track to confirm the thesis is holding as we go.

A proforma is a forecast, and forecasts are confident in ways the world is not. The purpose of stress-testing is to remove that false confidence before capital goes to work. Our rule on every deal is simple: the base case is never the whole story.

Here is how we approach it, in the order we actually apply it.

Start with what we control, then break it. Hard costs, soft costs, and timeline are the variables most under a sponsor's influence. We build the base case with current contractor pricing and realistic durations, then stress each: materials cost up 10%, labor cost up 10%, timeline extended by three to six months per phase. A plan that works only if costs don't move is not a plan; it is a hope. The real question is whether the project still delivers an acceptable outcome to investors when two of those three variables move the wrong way together.

Next, stress what we don't control. Exit pricing, rent growth, absorption speed, and interest rates sit outside any sponsor's influence. For each, we model a conservative case that assumes no rent growth, longer lease-up, and exit pricing flat to today's comps. If the model still clears hurdles with zero market tailwind, the deal has been stress-tested against a world that cooperates only minimally.

Look for correlation, not just magnitude. The dangerous scenarios are rarely one variable moving in isolation. A construction delay often happens alongside rising rates, because both respond to the same macro pressure. Soft rent growth often shows up with longer absorption. We specifically model paired moves — delay plus rate increase, slow absorption plus flat exit pricing — because those are the realistic shapes of adverse scenarios.

Stress the capital stack, not just the asset. A deal that pencils on an unlevered basis can still fail on a levered one if debt service assumptions break. We test interest rate scenarios 150 to 300 basis points above today's quotes, confirm debt service coverage holds, and check what happens if a construction loan extension becomes necessary. The asset and the capital stack are two different risk profiles, and both deserve their own stress tests.

Ask: what breaks it? Every stress test should end with an honest inventory of the scenarios under which the deal does not work. For Skyview, that list is short but real: a sustained regional employment shock, a capital markets environment where owner-carry financing also closes off, and a construction cost spike combined with a hard-stop timeline. We don't pretend these can't happen. We document them, explain what mitigants we'd use, and let investors read our work.

The output of stress-testing is not a single "right" answer. It is a map of outcomes — some excellent, some acceptable, some poor — along with the probability-weighted sense of where the plan sits. That map is the underwriting. The proforma is just the centerline through it.

When we say Precision in Underwriting, this is what we mean: not elaborate models, but disciplined interrogation of the assumptions every model depends on.

Most townhome developments of our size break ground on everything at once. It is the default answer for a reason — one mobilization, one subcontractor cycle, one finish line. We deliberately chose a different path for Skyview, and the reasoning connects directly to how investor capital is exposed during the build.

Our plan develops 32 units (16 lots of two units each) in four sequential phases of four units. Each phase runs about twelve months from land prep through certificate of occupancy, with phases overlapping such that total construction time is forty-eight months across the project. Two features follow from that structure, and both matter to investors.

Peak capital is materially lower. A monolithic build requires the full cost of land, site work, and vertical construction to be funded before the first unit generates income. A staggered build funds phase-by-phase. When Phase 2 breaks ground, Phase 1's first units are already leasing. By the time Phase 4 mobilizes, the earliest units can be approaching stabilization or sale. Investor capital is therefore exposed more briefly against a smaller peak balance than an all-at-once build would demand. The same project requires less money outstanding at the high-water mark.

The project teaches itself. Every construction project has flaws the first time through — subcontractor pairings that don't work, finish specifications that run over budget, design details that cost more in the field than they looked like in the plan. In a one-shot build, those flaws are baked into every unit. In a four-phase build, we find them in Phase 1 and correct them before Phase 2 mobilizes. Costs tighten as the project progresses. Schedules become more reliable. Quality improves. This is not a theoretical benefit — it is a measurable one that we track and report phase-over-phase.

There are costs to this approach, and we want to be clear about them. Each phase incurs its own mobilization, which is a real dollar amount. Total project duration is longer in calendar time than a single-shot build would be. And some efficiencies that come from bulk purchasing are partially given up — though we offset this by ordering long-lead items in staggered but overlapping batches to capture volume pricing where it matters most.

The trade we are making is straightforward: we accept somewhat higher per-unit mobilization costs and a longer overall timeline in exchange for a lower peak capital requirement and an operational feedback loop between phases. For a sponsor using outside capital, that is a favorable trade — because the cost of capital is a real number, and lowering the amount of capital exposed at peak is one of the few levers we have to reduce it.

This is what we mean by staggered development. It isn't a phrase we invented, but we've designed the full 48-month construction schedule around it — because in a market like Cambria County, the absorption profile rewards measured delivery, and the investor economics reward discipline about how much capital is outstanding at any given moment.

A lot of real estate sponsors do not build. They acquire, they plan, they raise capital, and they hire a general contractor to execute. The model works — right up until it doesn't. The point at which it stops working is almost always the same: when something in the field departs from the plan, and the sponsor has to negotiate with a GC whose incentives are no longer aligned with the investors.

At Northstar, we made a different decision early. Our construction arm — Duca Brother Developments, run by two of our three Managing Members — functions as the general contractor on our projects. That is not a coincidence of partnership. It is a design choice, and it shapes the investor experience in specific ways.

Change orders live inside the partnership. In an arm's-length GC relationship, a surprise in the field is a negotiation: the GC bills a change order, the sponsor pushes back, and the conflict is resolved somewhere on the spectrum between fair and expensive. When the GC's principals are two of the three partners building the project, that dynamic simply does not exist in the same form. A material departure from plan gets resolved around the same table, with the same people, whose upside is the same upside.

Schedules have real accountability. It is easy for an external GC to blame a delay on weather, subcontractor availability, or the sponsor's drawings. When the GC is inside the sponsor, excuses become problems to solve rather than reasons to extend the timeline. The buck stops somewhere identifiable.

Construction knowledge informs underwriting. The Duca family has been doing residential construction in the region for generations. That knowledge feeds backward into how we evaluate deals: site costs, labor availability, material lead times, permitting timelines, local inspector relationships. These are the details that turn a paper budget into an honest one, and they are not the kind of thing an acquisition team learns from a broker's offering memorandum.

The trade-off of this model is that we don't do thousands of units a year. Vertically integrated sponsors are inherently limited in scale by their construction capacity, and that shows up in how selective we can be about deals. We view that as a feature: we only take on projects where our team can actually do the work, because that is the version of sponsorship we believe produces reliable outcomes.

This structure is also why our reporting on construction is more granular than is typical — a subject we cover in a separate piece — because the same people doing the work are the people writing the update.

"Operators, not order-takers" is not a marketing line. It is a statement about where the accountability sits when something goes wrong. For investors trying to distinguish between sponsors who look similar on a pitch deck, the question of who actually builds is one of the more useful filters available.

Every sponsor claims to be transparent. The word is so common in pitch decks that it has nearly lost meaning. We would rather describe transparency by the specific reporting we commit to produce, because the difference between a transparent sponsor and an opaque one is mostly a matter of what shows up in the monthly update.

Here is what we consider the minimum useful content of an investor report on a development project like Skyview.

Where the money has gone. A running tally of capital calls, draws, and expenditures against the original budget — by phase, by category, not just an aggregate. Investors should be able to see at a glance whether the project is tracking within its contingency, consuming into it, or over it.

Where the project is on the critical path. A Gantt-style schedule showing planned vs. actual for each phase, with specific callouts where the project is ahead, on schedule, or behind. "On schedule" is a claim that should be testable against a number, not a sentence.

What changed from the last report. A standing section that names what is materially different from the prior month. This is the hardest section to write, because it requires honesty about setbacks. It is also the most useful for an investor trying to form a mental model of how the project is actually unfolding.

Leasing and sale activity, in the context of the underwriting. As units come online, we report unit-by-unit leasing or sales velocity, and we compare it to the absorption assumption used in the original underwriting. If we underwrote to 45-day average time-to-lease and we're tracking to 60, that belongs in the report — not buried, not spun.

Local market data. Rent comps, sale comps, and any relevant macro data for the county or submarket. This contextualizes the project's performance against the market it operates in, and it helps investors distinguish between project-specific issues and market-wide ones.

An honest "what we're watching" section. Every project has risk. A sponsor who reports no risk is reporting poorly. We maintain a short, specific list of items we're monitoring — a delayed permit, a subcontractor availability question, a macro data point worth noting — and we update it every month.

What you will not see in our reports is the kind of content that crowds out the signal: stock photography, glossy renderings of completed phases, boilerplate language about "our commitment to excellence." Reporting is for facts. Marketing is for somewhere else.

The test we apply internally is whether, if an investor reading the report had to make a decision that afternoon based solely on what we sent them, they could do it with confidence. If the answer is no, the report is incomplete. If the answer is yes, the report has done its job.

Transparency, by this definition, is not a feeling or a value. It is a document. And documents can be judged.