Setting the Asking Rent When a Long-Term Rental Is Vacant
Last week’s post was about pricing a short-term rental. This one is its companion for long-term rentals: what to ask when a unit comes vacant. It is a 9-page tutorial, and you can download the PDF.
How it was made
As with the short-term one, I had an AI coding assistant (Claude) work it out and write it up. Every number comes from a simulated market whose truth is known: a made-up apartment where comparable units rent for $2,400 a month. No real listing or tenant data went into it.
Why it is a different problem
A short-term rental sets a price every night. A long-term rental sets one number, and lives with it for the whole lease. Ask too much and the unit sits empty: one empty month costs about 8% of a year’s rent. Ask too little and the shortfall repeats every month of a tenancy that, in the simulation, lasts about 27 months.
What it finds
- The best ask is a little under the comps. The simple formula says ask $62 above the market rent. Tenants who pay more tend to leave sooner, and defaults rise with the rent. Both pull the best ask down to $2,378, which is $22 below the comps.
- Overpricing costs more than underpricing. Asking $100 too much costs $632 a year. Asking $100 too little costs $355.
- A rule of thumb for cuts. A $50 cut pays for itself if it saves 18 days of vacancy. If you would raise the rent to market at renewal anyway, 7 days is enough.
- Rent caps make the starting rent matter more. Under a cap, underpricing costs about twice as much, because a low start cannot be fixed at renewal.
- Comps mislead about price sensitivity. Nicer units ask more and still lease fast, so fitting rent against days on market across 400 simulated comps makes demand look 3.3 times less sensitive to price than it is.
- The vacancy itself is the best data. Counting inquiries every week, against what the comps predict, narrows the estimate of the market rent from ±$119 to ±$51 in four weeks.
Here is how four ways of setting the rent did over 3,000 simulated vacancies. The loss is measured against someone who knew the true market rent:
| Approach | Lost per vacancy | Days empty |
|---|---|---|
| Cut based on the weekly inquiries | $536 | 28 |
| Cut 2.5% every two weeks | $1,181 | 37 |
| Best fixed ask | $1,461 | 30 |
| Ask what the comps ask | $1,541 | 39 |
Cutting based on the inquiries saves about $418 a year over asking what the comps ask. Season matters for how long a unit sits, 18 days in June against 29 in November, but it barely moves the best ask.
The routine
Start a little above the best ask for the comps. Count the inquiries every week. Cut when they fall short of what the comps predicted, by as much as your estimate has moved.
Limits
It is a model, checked on a simulation. The demand, tenure and default curves are assumptions. It assumes renters accept an 18-month lease as readily as a 12-month one. The summary of California’s rent cap (AB 1482) is not legal advice.
As always: this is what worked (and didn’t) for us, not legal, tax, or investment advice.