What "Best Places to Invest" Lists Don't Tell You

These lists aren't wrong. They're answering a question you didn't ask. They rank where short-term rental returns looked good over some recent period, which is not the same as where you should put money next. The distance between those two things is most of what decides whether a purchase works.

The format only knows what already happened

A ranking has to be built on data that exists, and data that exists is history. So every list of this kind is a description of the recent past presented in the grammar of a recommendation.

That isn't dishonest. It's arithmetic. But the consequence is easy to miss: by the time a market is ranked highly enough to publish, its numbers have already been good for long enough to be measured, written up and edited. Whatever advantage existed in noticing it early is gone. You are reading a public thesis.

Sometimes that's fine, because plenty of markets stay good for years. The point is only that a list can't distinguish a market that's still climbing from one that peaked eighteen months ago and hasn't finished rolling over yet. Both look identical in trailing data.

Two dates, and they're rarely the same

Every list has a publication date. Every list also has a data window: the period its numbers cover. These get conflated constantly, and the gap between them is usually months.

A piece published in August might be running on data through March. That's not sloppy; there's a lag in how this data becomes available. But it means "2026's best markets" can be describing a market as it stood three quarters ago.

Then there's the update problem. Lists are frequently refreshed in place: same URL, same headline, new year in the title, some of the entries swapped. The publication date moves. It's not always clear how much of the underlying analysis moved with it. Before you take a ranking seriously, find the data window, not the byline date.

What the format structurally cannot include

Some omissions aren't oversights. They're things a ranking has no room for.

The price you'd actually pay

This is the big one. A market can be ranked on revenue, occupancy or growth without any reference to what property costs there. Two markets with identical revenue and a 40% difference in entry price are not remotely the same investment, and a revenue-ranked list will place them next to each other.

Your return is a relationship between what you earn and what you put in. Rankings almost always measure only the first half. If you want the number that settles it, that's cash-on-cash return, and no list can compute it for you because it doesn't know your deal.

Supply that's coming but isn't here

A list counts listings that exist. It doesn't count the units under construction, the permits issued last quarter, or the conversions already financed. In a market that made a list last year, some of that pipeline exists precisely because the market made a list.

Regulation mid-process

Ordinances take months to move through committees, hearings and votes. A market with a night cap under active consideration and a market with no proposal at all look the same in any dataset, right up until one of them doesn't. You can check this yourself. City agendas are public, and an afternoon is enough.

Which part of the market the numbers describe

A list gives each market one row. There is no room in that format for the possibility that a market's strong showing is coming almost entirely from large properties while the one-bedroom segment has been flat for two years. The row can only hold one verdict, so it holds the flattering one. Nothing about the ranking is wrong; it just can't say which part of the market it's describing. Reading the slice you'd actually own is a separate step you have to take yourself.

Use one as a shortlist, never as a decision

So here is the position worth holding: these lists are useful for one job and bad at the job people use them for.

As a filter, a list is excellent. Thousands of markets exist; you can seriously evaluate maybe five. Something has to do the first cut, and a reasonable ranking does it faster than you can. Take the shortlist gratefully.

Then throw the ranking away and start those five from scratch, because none of what settles the outcome was in it. Some of that is legwork nobody can do for you: reading council agendas, counting cranes, working out what a comparable property last sold for. Some of it is the standard analysis process. Either way it begins where the list ended, and the list ended earlier than it looked like it did.

Takeaways

• A ranking describes the recent past, so it can't tell a market still climbing from one that already peaked.
• Find the data window, not the publication date. They're often months apart, and lists get refreshed in place.
• Revenue rankings measure what a market earns and never what it costs to get in. Half a ratio.
• Supply already permitted, and regulation still sitting in committee, are invisible to every list and checkable by you in an afternoon.
• Use a list to make the shortlist. Nothing else.

Make the list argue with a number

The quickest way to stop treating a ranking as a verdict is to put something concrete next to it. Take the market a list just recommended, take a real property in it at a real asking price, and see what falls out.

Often the answer is that the market is fine and the deal isn't, which is a distinction no ranking is built to draw. Our Revenue Calculator runs in the browser, free, without an account, so almost nothing stands between reading a list and testing it. Treat what it gives you as an estimate, because that's what it is. But it's an estimate you can argue with, which puts it ahead of a position on a list.