Where we're looking, why, and what the data actually says.

We get asked at least once a month why we don't have a single South Florida position in the pipeline, given how much capital is still chasing Miami-Dade and Broward. The honest answer is that we looked hard, twice, and the numbers talked us out of it both times. Not the vibes — the actual filings, the insurance data, and the migration counts. This is what we found.

The carrying cost gap is enormous, and it isn't closing

Florida's statewide average homeowners insurance premium is projected at roughly $8,458 a year in 2026 — about three times the national average. But "statewide average" hides the real story, because Florida isn't one insurance market, it's six or seven regional ones stacked on top of each other. On a comparable dwelling, Miami-Dade is running somewhere in the $12,000–$20,000 range depending on coastal proximity, Broward is close behind at roughly $9,750, and the Keys are north of $14,800. Orange and Seminole counties — the core of the Orlando market — are sitting around $2,180–$3,645. Northeast Florida, where Jacksonville sits, runs a similar $2,185–$4,005. That's not a modest discount. That's a three-to-five-times difference in annual carrying cost before a single dollar of purchase price is on the table, and it applies every year the asset is held.

To be fair to South Florida, 2026 is actually the first year in nearly a decade this is getting better, not worse. Citizens Property Insurance approved an average statewide rate cut of 8.7%, with Broward seeing cuts closer to 17% and Miami-Dade around 10–14%, as tort reform and new private carriers finally start working through the system. We think that's a genuinely good development for existing owners. It doesn't change our read on new positions, though — even after the cut, coastal South Florida premiums are still multiples of what they are eighty miles inland. The gap is narrowing. It isn't closing.

The condo math changed on January 1, and it isn't a small change

Following the 2021 Surfside collapse, Florida gave condo buildings a multi-year grace period before enforcing stricter structural and reserve requirements. That grace period ended on January 1, 2026. Any condo building three stories or taller now has to complete milestone structural inspections and fully fund its reserves — boards can no longer vote to waive or underfund them. The practical result has been a wave of special assessments, some running into six figures per unit in older coastal buildings, on top of master-policy insurance increases that have doubled HOA dues in some Miami-Dade and Tampa Bay buildings. None of that shows up in a listing price. It shows up eight months after closing.

The migration data flipped, and most people haven't noticed yet

Here's the part that actually moved us off South Florida entirely, not just cautious about condos. Miami-Dade posted the largest domestic outflow of any high-flood-risk county in the country in the most recent full year of Census data — roughly 67,000 more people left than arrived — and it's one of only a handful of major counties nationally, alongside Los Angeles, that's shrinking in absolute numeric terms. Multiple large coastal metros, Miami included, would already be losing population outright if not for international migration, which has itself slowed sharply over the past year.

Compare that to what's happening ninety minutes north. Orlando added the tenth-most residents of any U.S. region last year and grew jobs at 1.3%, ahead of both the state and national rate — and the region's own long-range forecast has it adding more total jobs over the next decade than Miami, a metro with twice its population. Jacksonville is on pace for roughly 18,000 to 22,000 net new residents this year, and critically, the drivers behind that number aren't a single employer announcement or a pandemic-era remote-work spike. They're structural: no state income tax, strong school districts pulling in St. Johns County specifically, and a durable cost gap versus the metros people are leaving. Structural reasons for growth tend to persist. Vibes-based growth doesn't.

Don't buy the hype. Follow the data.

None of this means South Florida is broken or that every deal down there is a bad one. It means the market has gotten expensive to carry, more legally complicated to build in, and demographically softer than the headlines still suggest, all at the same time investors keep pricing it like nothing changed. We'd rather be early to the correction than late to the party. For now, that means Jacksonville and Orlando get our attention, and South Florida gets a pass — not because we don't like it there, but because that's what the permits, the premiums, and the population counts are actually telling us.

People ask us for our "secret markets" more than they ask us almost anything else, usually expecting a clever answer. The honest one is less exciting and more useful: we watch for a handful of leading indicators clustering in the same place at the same time — permits, employer filings, migration, land-use changes — and then we go find out in person whether the story holds up. Three markets are currently high enough on that list that we're actively spending time on the ground in each of them. None of these are opportunities on our pipeline yet. They're the step before that.

What happens next with all three is the same thing that happens with every market on this list: someone from our team gets on a plane, walks the parcels, sits down with the local planning office, and calls people who actually operate there. If the story holds up in person the way it holds up in the data, one or more of these could show up on our opportunities page in the next couple of quarters. If it doesn't, we say so and move to the next one.

People are sometimes surprised that a firm running a proprietary data model spends most of its time talking about site visits, county planning meetings, and phone calls with people we've known for years. Here's why: the model is very good at telling us where to look. It has never once told us whether a county commissioner is going to fight an annexation, whether a landowner actually wants to sell or is just fielding calls, or whether the "shovel-ready" site everyone's excited about has a wetlands issue nobody bothered to check.

Our system reads permitting activity, employer filings, infrastructure spend, and migration data continuously, and it's genuinely useful for narrowing a very large map down to a short list worth our time. But a signal is not a deal. Between a flagged market and an actual position, there's a long list of things that only get resolved by a person standing on the parcel, in the room with the seller, or on the phone with someone at the county who isn't going to put the real story in writing.

What the model can't do

• Tell you whether a landowner is emotionally ready to sell, or just curious what their land is worth
• Read the room in a zoning hearing and know which objection is going to actually stick
• Notice that the "available" utility capacity everyone's citing was allocated to another project last quarter
• Build the kind of relationship with a broker that gets you the call before a site is marketed at all

Every one of those is a real reason a deal that looked perfect on paper has died in our diligence process — and every one of those was caught by a person, not a dashboard. That's not a knock on the technology. It's the reason we built the technology the way we did: as a research tool that narrows the map, not a black box that makes the call. The decision still has to be made by somebody who's walked the site.

Why this matters more, not less, as the tools get better

The easier it becomes to generate a slick-looking market report, the less that report is actually worth. Everyone will have access to similar data eventually — some version of what we've built is going to become table stakes across the industry, and that's fine. What doesn't scale the same way is a team that's walked hundreds of these sites, that county planning offices actually take calls from, and that landowners trust enough to negotiate honestly instead of running a bidding war. That's slower to build and impossible to copy quickly, which is exactly why we've stayed small and stayed close to the work instead of trying to scale the research team faster than we can develop that kind of judgment.

So yes, we use the technology, and we'll keep improving it. But if you're evaluating us, evaluate the team that shows up to walk your site, not the dashboard that told us to call you. The dashboard is the easy part.