MultiScreen/Markets
Methodology · v0.5.0

A disciplined, repeatable
definition of a good market.

Every input is percentile-ranked within the universe, weighted within its category, and rolled into a 0–100 composite. Weights are versioned data — never hard-coded — and are validated against historical backtests before calibration.

Why the ranking matters

Buy in the model's top markets, and history says your rent growth outperforms.

The whole point of ranking markets is what happens after you pick from the top of the list. Markets the model scored in its top quartile went on to deliver about 2.1% more rent growth per year than the markets it ranked at the bottom. That's not a projection — it's a historical look-back: we re-ran the model as of every year since 2016, using only the data that existed at the time, then measured the rent growth each market actually delivered over the following 3 years.

7 of 7
Yearly tests won — the model's top picks out-grew its lowest-ranked markets in every test since 2016
+2.1%/yr
Extra rent growth per year, top-quartile markets vs bottom
77%
Top-quartile picks that beat the median market

+2.1% a year compounds. Over a five-year hold that's roughly 11% more cumulative rent growth — before any operational upside — simply from choosing the right market. Model weights were locked before the validation window, so every future quarter is a live, unretouched test.

Unlock the full rankings with Pro →

What the model measures

Demand
The growth engine. Rising employment, incomes, and net in-migration create new renter households and the ability to pay more. Also tracks the own-vs-rent premium: when home prices outrun rents, would-be buyers stay renters.
Scores wellMore jobs, faster wage growth, stronger in-migration, and a widening cost gap between owning and renting score higher.
Supply Constraint
The differentiator, and the heart of the thesis. New construction competes existing rents down, so we favor markets where builders are adding the fewest units relative to market size — and where geography itself (water, steep slopes, wetlands) caps how much can ever be built.
Scores wellFewer multifamily permits per capita and more measured land unavailability score higher. Heavy building on open land scores lower.
Fundamentals
How much room rents have left to run. Rents still affordable relative to local incomes can keep climbing; rents already stretched against wages have less headroom.
Scores wellLower rent-to-income (more affordability headroom) plus a healthy recent rent trajectory score higher.
Momentum
Direction and acceleration — whether demand and rents are speeding up or cooling, how much rent growth each unit of new supply is actually producing (our proprietary supply-adjusted demand term), and the current 12-month rent trajectory, which catches turns a multi-year average hides.
Scores wellAccelerating jobs, migration, and rents — rents growing now, not just historically — score higher.

Deliberately supply-weighted: two markets with identical demand are very different investments when one has twice the permits in the ground. The interaction term — supply-adjusted demand — rewards exactly that spread. Two safeguards keep scarcity honest: supply-constraint credit is scaled back where demand is weak (a quiet permit pipeline in a market nobody is moving to is not an edge), and metros dependent on a single industry take a haircut to the composite.

The full model is a Pro feature

Pro members see the complete machinery: exact category and input-level weights, every underlying data series, grade cutoffs, watch-item rules, and the full year-by-year backtest — plus the live rankings, comparisons, and tearsheets for all 53 metros.

Upgrade to Pro →Preview the rankings →

Data sources

Employment & wages
Jobs, average pay, sector mix
A federal census of employers — near-population coverage, not a survey sample
Population & migration
Population growth & net migration
Official government estimates of components of change, metro level
Building permits
Multifamily permitting activity
The forward supply signal — permits lead deliveries by 18–30 months
Market rents
Asking rents, monthly
A national observed-rent index — market-rate, not contract rents
Home prices
For-sale pricing, quarterly
Drives the own-vs-rent premium — ownership cost outrunning rents keeps households renting
Income-weighted migration
Who is arriving, not just how many
Government migration flows with incomes attached — what arriving households earn
Terrain & land constraint
Structural supply barrier
In-house measure: share of nearby land blocked by water, wetlands, and steep slopes — geography that permits can't change

Every stream is institutional-grade — government statistical programs plus a national rent index, with no scraped listing data and no licensing exposure. The specific series, transformations, and weightings are proprietary to MultiScreen; provenance documentation is available to institutional subscribers on request. Phase 2 adds licensed submarket data (rents, occupancy, under-construction pipeline) plus additional demand-side inputs — the ingestion slots already exist in the engine.