What breaks at property #4

Nothing breaks at two properties.

Two properties is a shoebox and a decent memory. You know which unit the water heater went in. You know the Home Depot run in March was the Elm Street tub surround, because there was only one bathroom project that year and you were in it up to your elbows. Your CPA asks a question in February and you answer it from the driver's seat of your car.

Then you buy the fourth one, and somewhere in the following twelve months you discover that the system you had was not a system. It was you, remembering things.

This post is about the specific things that stop working, roughly in the order they stop.

1. The Schedule E page break

Here's the first hint the IRS gives you, and it's a literal one: Schedule E has room for three properties per page.

Property four goes onto a second page. Which is fine — that's what the form is for — but it means your per-property detail now lives on one page and your totals live on another, and the two only reconcile if the underlying records were clean. Nobody notices this until a preparer asks why column B on page two doesn't tie.

It's a trivial mechanical thing. I bring it up because it's the moment the form itself stops assuming you're a hobbyist, and it's a decent early warning that the rest of your setup is about to be tested.

2. Banking: decide before you need to

Two schools, both defensible:

One account per property. Cleanest possible audit trail. Every transaction is pre-tagged by the account it came from. Becomes genuinely annoying around eight properties, when you're managing eight balances and eight sets of monthly minimums.

One operating account with disciplined tagging. Scales further, requires actual discipline, and falls apart the first month you get busy.

I'm not going to tell you which is right — plenty of investors run either successfully. What I will say is that switching later is painful. Re-tagging a year of commingled history to satisfy a question about one property is a specific kind of misery. Pick your approach while you still have three properties and the stakes are low.

The one thing that isn't a school of thought: running rental money through your personal checking account. If you're doing that at four properties, fix it before anything else on this list.

3. The three things that stop being reconstructable

At some point each of these transitions from "I know that" to "I could probably figure that out" to "I have no idea."

Which property a purchase belonged to. The $340 Home Depot charge. At two properties you know. At five you're staring at a bank feed in February trying to remember what you were doing on a Tuesday in July.

Whether a job was a repair or an improvement. This one's worse, because the answer depends on facts you knew at the time and did not write down — what was actually wrong, what you actually did, whether you replaced a component or the whole system. A year later all you have is an invoice that says "plumbing work, $4,200," which supports neither position.

Where the receipt is. Bank feeds prove you spent money. They don't prove what you spent it on. For anything material, the invoice is the documentation and the bank line is just the shadow.

The common thread: all three are cheap to capture in the moment and expensive-to-impossible to reconstruct later. Same lesson as the assessor's land values, if you read my closing costs post — the information is free on the day and gone by the time you want it.

4. The safe harbors are per-property, and nobody tracks them that way

This is the one I'd most want a growing investor to know, because it's real money and it's genuinely obscure.

The repair-vs-improvement rules come with elective safe harbors that let you expense things you'd otherwise capitalize. Two of them matter most for residential rentals:

• The de minimis safe harbor, which lets you expense items under a per-item dollar threshold, provided you've got the election and a written policy in place.
• The small taxpayer safe harbor, which lets you expense a year's improvements on a building if they stay under a limit — and that limit is calculated as a percentage of the building's unadjusted basis, capped in absolute dollars.

Here's the part people miss: that second one is computed per building, per year. Not portfolio-wide. Each property has its own ceiling based on its own basis, and each property either fits under it or doesn't, independently.

Which means the useful question isn't "what did I spend on repairs this year." It's "what did I spend on this building, and where does that sit against this building's limit." If your books only total at the portfolio level, you cannot answer that, and you will find out in March that a $9,000 job on your cheapest property blew past a ceiling you never calculated — and now it's a capitalized improvement depreciating over 27.5 years instead of a deduction you take today.

The numbers move, so check the current thresholds with your CPA rather than trusting a blog post. The structural point doesn't move: these are per-property tests, and per-property tests require per-property books.

5. Suspended passive losses are tracked per property. Forever.

Most rental losses are passive. If you can't use them — no passive income to absorb them, no REPS, no qualifying STR — they don't vanish. They suspend and carry forward.

Per property. Indefinitely.

And they're not just a footnote, because when you fully dispose of a property in a taxable sale to an unrelated party, that property's suspended losses generally free up. Which can meaningfully change which property you sell, and when.

So the balance matters. And it lives in exactly one place: your accountant's software, in a schedule most investors have never asked to see.

Two things worth doing. Ask for that schedule every year and keep your own copy — because if you ever change preparers, that history is the thing most likely to get lost in the handoff, and reconstructing a decade of suspended losses from old returns is a project. And know the balances before you decide what to sell, not after.

6. Why January and not March

Everything above is fixable. It's just fixable at wildly different prices depending on when you do it.

In January you're setting up the year: accounts, tagging, a policy for how receipts get captured, a note to yourself about what the safe harbor thresholds are. That's an afternoon.

In March you're reconstructing the year, at the exact moment your CPA is least able to help you and most expensive per hour. And the answers you produce under deadline pressure are the ones you're least confident defending later.

The thing about property four is that it doesn't feel like a threshold when you cross it. It feels like buying another house. The strain shows up quietly, one unanswerable question at a time, and by the time it's obvious you've got a year of history to untangle rather than a system to set up.

The short version

At four properties, the form stops assuming you're casual and so should you. Separate the banking, capture the property tag and the receipt at the moment of the transaction, track repairs against per-building limits rather than portfolio totals, and get your suspended loss schedule in writing every year.

The repair-versus-improvement question is usually the first to bite, and it is rarely close. Which one lands first depends on whether you are scaling in one market or four.

This is general information, not tax advice. Thresholds, elections and dollar figures change, and the right treatment depends on your situation — check with a CPA before you act on any of it.