Hey there —

Say a cheap market out loud and you will get the warning inside a week. You'll never be able to sell it. Delivered with total confidence, usually by someone who has never owned there.

22. That is the median number of days a house takes to go under contract in the county containing Detroit, against a national county median of 54 (🟢 county-level transaction data, month ending 2026-08-31). It also clears about 98.4% of asking, better than the typical American county manages. Cleveland's county: 25 days, 99.1%.

So the warning is pointing at something real and naming the wrong variable.

Sort all 2,698 counties by price and the sale times produce no pattern at all — 57 days in the cheapest fifth, 52 in the priciest, wobbling in between. Sort the same counties by how many houses actually trade and you get a clean ladder, every step: about 70 days at the thin end, 43 at the deep end. Depth predicts how fast you get out. Price barely does. I put both sorts side by side here, including the honest limits — the relationship is real but weak at the level of any individual county, and Birmingham's county is the counterexample sitting inside our own list.

Here is what made the week interesting to me, though. That is the same error twice.

The second one is on a line you have almost certainly set and never revisited. Vacancy, around five percent. Bad debt, three or four. Both percentages of a year. But the event those lines exist to absorb does not arrive as a percentage — an eviction arrives as months, and the months are set by statute where the property sits, not by your building or your screening. On an $1,850 unit that lands near $8,390, against the $1,998 both lines reserve together. The arithmetic is short, and it is not an argument that your percentages are too small. They are the wrong shape.

Cheap is not thin. A rate is not a duration. In both cases the label got priced and the mechanism did not.

One thing worth doing this week. Pick the market at the top of your shortlist and find out how many houses closed there last month. Your agent can pull it, the county recorder publishes deed transfers, and the public housing-market data centers carry county files directly. It takes about ten minutes, and you will know more about your exit than the price told you.

Then ask the second question, because the count cannot answer it: is your property the kind of thing those buyers were buying? A county trading 400 single-family houses a month is deep for a single-family house. It may be bathtub-thin for a twelve-unit.

I keep coming back to one question this week, and I would genuinely like to know your answer: have you ever passed on a market because of something you heard rather than something you looked up? No judgement — most of us have, and the warnings usually sound authoritative. Hit reply and tell me which one. I read every single response.

This week's scenario puts the same choice in front of you with real money on it: one house sells in six weeks, the other is $35,000 cheaper and takes ten. There is a third path, and it is a number.

— Martin