The LOI isn't where submarket selection begins. By the time you're signing a letter of intent, the submarket decision is already locked. The question — whether this specific submarket has the fundamentals to support your underwriting assumptions over a 5-7 year hold — should be answered before you spend three days on a rent roll and two days on comps. In our experience building the data layer for CRE acquisitions, the firms that consistently underwrite well start their market intelligence work at the top of the funnel, not midway through diligence.
What "Data-Driven" Actually Means for Submarket Selection
There's a version of "data-driven submarket selection" that means pulling a CoStar vacancy report and drawing a circle on a map. That's a starting point, not a method. The firms doing this well are running three to five signal categories simultaneously, cross-referencing them against each other, and using the intersection to define a buy box — not just a geography.
The five categories worth tracking systematically:
- Absorption trends. Net absorption over trailing 8 quarters tells you whether occupancy is structurally improving or just holding steady due to low new supply. A submarket with 2% net absorption and zero under-construction pipeline is a different risk than 2% absorption with 1.5M SF of new product coming in the next 18 months.
- Effective rent trajectory. Asking rents versus effective rents (after concessions) tell you whether face-rate increases are real or whether landlords are giving back the headline gains in free rent. In several Sunbelt industrial submarkets, asking rents rose 18-22% over 2022-2024 while effective rents rose only 11-14% due to increasing concession packages. That's a meaningful difference for a model assuming market-rate rent growth on rollover.
- Tenant credit concentration. A submarket dominated by a single industry sector (logistics, financial services, government) carries concentration risk that doesn't show in vacancy or rent data. We've seen submarkets with 4% industrial vacancy where 40% of the occupied space is leased to a single sector facing cyclical headwinds. The 4% vacancy figure looked healthy. The tenant composition didn't.
- Capital markets activity. Recent transaction volume and cap rate movement in the submarket tells you whether institutional capital is entering or exiting. Falling transaction volume in a submarket — even if fundamentals look stable — often precedes a cap rate decompression event by 12-18 months.
- Supply pipeline. Under-construction and proposed development provides the forward-looking view that trailing data can't give you. A submarket with strong current fundamentals but 15% of existing inventory under development is a different underwriting thesis than one with tight fundamentals and a constrained land supply.
When to Run Submarket Analysis in the Deal Timeline
Most deal teams run submarket analysis concurrently with rent roll analysis and comp research — all three happen in the first week of diligence after the PSA is executed. The problem with this timing is that submarket analysis at that stage is confirmatory, not decisional. You've already committed to the deal; you're looking for reasons to get comfortable, not reasons to walk.
The firms that use market intelligence most effectively in acquisitions run a lighter version of submarket analysis at the top-of-funnel screening stage — before the LOI, and in some cases before the first broker conversation. They maintain a standing data layer for their target markets: quarterly absorption, effective rent trend, supply pipeline, and recent comparable sales. When an opportunity surfaces, the first question is whether it falls in a submarket that's already on their buy list based on these signals. If it doesn't, they skip it or underwrite it with a heavier risk haircut before deciding whether it's worth pursuing.
This approach requires maintaining the data layer continuously, which is the investment. The return is that diligence becomes confirmatory rather than exploratory, and deal teams spend fewer analyst days on opportunities that weren't viable from a submarket perspective to begin with.
The Comp Research Problem Within Submarket Selection
Submarket selection and comp research are often treated as separate activities. They're not. The comparable lease transactions you use to validate rent assumptions are drawn from the same submarket you're analyzing for market conditions — and the way you define submarket boundaries dramatically affects which comps are in scope.
In dense urban markets, a "submarket" defined by a broker might be a 10-block radius. An analyst using that same definition for comps might include transactions from meaningfully different micro-locations — different transit access, different floor plate sizes, different vintage. In our experience, submarket definition is one of the most frequently debated items in deal team post-mortems when a deal underperforms. "Our comps were from a different micro-market" is a phrase we hear often.
The practical solution is to define submarket tightly and then check whether you have enough comp volume to support that definition. If a tight submarket definition produces fewer than 8-10 comparable lease transactions in the trailing 24 months, you're in a low-liquidity market where comps need to be supplemented with broker intelligence rather than treated as statistically reliable. That's not a reason not to invest — but it's a reason to widen your cap rate range and discount rent growth assumptions accordingly.
Submarket Signals That Predict Rent Growth Versus Those That Don't
Not all submarket data predicts rent growth with equal reliability. In our data analysis across several hundred CRE transactions, we've identified the signals with the strongest forward correlation to effective rent movement over 2-4 year horizons:
| Signal | Predictive for Rent Growth? | Notes |
|---|---|---|
| Current vacancy rate | Weak alone | Only meaningful in context of trend direction and supply pipeline |
| Net absorption trend (8 quarters) | Moderate-strong | Consistent positive absorption predicts tightening better than point-in-time vacancy |
| Concession trend (free rent months) | Strong | Declining concessions often lead effective rent improvement by 1-2 quarters |
| Supply pipeline vs. trailing demand | Strong | Pipeline/demand ratio above 1.5x is a reliable headwind signal |
| Transaction cap rate movement | Moderate | Capital market cycle doesn't always align with operating fundamentals |
| Tenant credit composition | Often overlooked | Sector cyclicality affects absorption before it shows in vacancy statistics |
The key insight from our analysis: vacancy rate alone is a lagging indicator. Firms underwriting purely from vacancy data are looking at where the market was, not where it's going. Absorption trend and concession direction are worth building directly into your buy-box criteria as forward-looking signals.
Applying Market Intelligence to Cap Rate and Rent Growth Assumptions
The output of a rigorous submarket analysis should feed directly into your underwriting assumptions — not just serve as a narrative backdrop for the IC memo. Specifically:
- Vacancy assumption: Use the trailing 8-quarter trend, not point-in-time data. If vacancy has improved from 11% to 8% over 8 quarters with positive absorption, a 7% stabilized vacancy assumption is defensible. If it's held flat at 8% with concessions rising, 10% is more appropriate.
- Rent growth rate: Differentiate between in-place lease escalations (contractual and certain) and renewal/rollover rent assumptions (market-dependent). Use submarket effective rent trajectory for rollover, not asking rent or prior-cycle peak. In 2024 industrial markets with softening concessions, a 3-4% annual effective rent growth assumption on rollover was defensible in well-absorbed submarkets and aggressive in oversupplied ones.
- Exit cap rate: Cap rate assumptions need to incorporate where the submarket is in its capital cycle at the projected exit year, not just today. A submarket with rising transaction volume and institutional capital entering currently is a different exit environment thesis than one where volumes are declining. We model exit cap as a range — base/bull/bear — with the spread sized to the confidence level in the submarket's capital market liquidity.
Building a Repeatable Submarket Intelligence Practice
The operational challenge for most deal teams is that this level of submarket analysis requires significant ongoing data maintenance. You can't run it fresh on every deal; the lag between data collection and deal pipeline is too long. The teams that do it well maintain a living market intelligence database for their target submarkets — updated quarterly at minimum, monthly in active markets — and use it as a standing input to deal screening rather than a deal-level research project.
In practice, this means building data feeds from primary sources — CoStar, Yardi Matrix, local broker market reports, municipality permit data for supply tracking — and normalizing them into a consistent format that allows quarter-over-quarter comparison. The normalization step is where manual processes break down. Broker market reports use different definitions of submarket boundaries, different vintage classifications, and different treatment of partial buildings. Getting apples-to-apples comparisons requires a data cleaning layer that most deal teams don't have the bandwidth to maintain by hand.
Automated market intelligence aggregation — pulling from standardized data sources and applying consistent normalization rules — is what makes a standing submarket intelligence practice operationally feasible for a mid-market deal team. It's not a luxury. For teams running 50-100 acquisition screens per year across multiple markets, it's the difference between a buy-box that is calibrated to current conditions and one that's running on data from the last time someone had time to do a deep submarket pull.
What to Watch in 2025 and Beyond
The submarket trends worth paying particular attention to across major US asset classes:
Industrial submarkets adjacent to major port facilities are showing absorption bifurcation — Class A near-port is holding tight while Class B and C mid-county space is softening as occupiers right-size from the pandemic-era over-leasing. A blended submarket vacancy figure masks this divergence. Underwriting Class B industrial in a submarket with strong Class A metrics requires its own comp set and its own vacancy assumption.
Office submarkets in 18-hour cities (Nashville, Austin, Denver, Raleigh) are outperforming gateway market Class B significantly. Effective rent recovery in those submarkets has been faster and more consistent than the headline office market narrative suggests. For acquirers willing to underwrite the office risk, those submarkets show absorption trends that support more aggressive rent growth assumptions than most models are currently using.
Multifamily in oversupplied Sunbelt submarkets is a 2024-2025 vintage opportunity for buyers who underwrite to 2027-2028 normalized conditions rather than current vacancy. The supply cycle in those markets peaked in late 2024; absorption is expected to catch up to deliveries by mid-2026. That's a thesis that requires submarket-specific supply pipeline data to defend credibly — generic "Sunbelt multifamily" framing doesn't hold up at IC.
Submarket selection isn't a binary call. It's a structured analysis that shapes every assumption in your model. The LOI deadline doesn't change that — it just means you need to have done the work before you get there.
