Moss v. Commissioner: no post-event ballpark guesstimate

By Max Medvedev · 6 Aug 2026

The sentence every hours case is decided against

Moss v. Commissioner, 135 T.C. 365 (2010), is where the phrase comes from: the regulations "do not allow a post-event ballpark guesstimate." Two holdings sank the participation claim — an hour count estimated after the year closed, and hours that were really time spent available rather than time spent working. Both still lose.

What happened

Moss is not a short-term rental case. It is foundational passive-activity doctrine, and it transfers directly, because a short-term rental that clears the average-stay gate is then tested for material participation under exactly the rules Moss was decided under.

The participation claim rested on two things that turned out to be the same weakness twice. The hours were not recorded as the work happened; they were estimated afterward. And a large share of what was counted was time spent on call — reachable if something came up — rather than time spent doing anything.

What the court held

The regulations "do not allow a post-event ballpark guesstimate." An estimate assembled after the fact is not proof of participation, however honestly it was arrived at.

And being on call is not participation. The seven tests in Reg. §1.469-5T(a) are written around hours of work performed in the activity. Availability is a state, not an act, and a state does not accumulate hours.

Both holdings are unhelpful to a certain kind of owner, and that is the point. An owner who is genuinely tethered to a property all summer — phone on, ready for the next call — has given up real time. §469 does not care. Only the work counts.

The tell

The tell is a total that could only have come from memory.

Look at how the two failures compound. Once availability is inside the count, the number gets large fast: an on-call day is 24 hours wide. And once the count is an estimate, nothing anchors it — no start and stop times, no artifact dated the same day, nothing to test it against the day job or the guest calendar. A big number with nothing under it is exactly the shape an examiner is trained to open.

The pattern repeats across the cases that came after:

• Mirch v. Commissioner, T.C. Memo 2025-128 — roughly 920 hours claimed, 744.5 of them site management and on-call. Credible hours came in under 100, and the entire deduction was lost. The log was undated, used a flat block per turnover regardless of a 1-night or 14-night stay, and claimed 168 cleaning hours against a deduction taken for professional cleaners.
• Penley v. Commissioner, T.C. Memo 2017-65 — about 2,520 hours claimed, more than a full-time job's 2,000-hour year, with no start or stop times. Implausible on its face, and discarded.
• Hairston v. Commissioner, T.C. Memo 2019-104 — padding: trivial one-hour entries, vague descriptions like supervising painting, logged weekly rather than daily. The padding tainted the whole log, not only the padded lines.
• Sezonov v. Commissioner, T.C. Memo 2022-40 — a short-term rental with no log kept as the work happened; the reconstructed hours were dismissed in the same words Moss supplied.

The lesson those four share is the one worth internalizing: over-claiming invites a court to discard everything. 800 hours that can each answer for themselves are worth more than 1,200 that cannot, because credibility is assessed on the log as a whole.

What this means for your records

Two practical rules fall out, and they pull in the same direction.

On-call time has no line in a defensible log. Not a discounted line, not a separate category — no line. Its value in every test is zero, and including it damages the entries that are real by putting the whole record in the same frame the court rejected in Moss and again in Mirch.

Every entry names an act and carries an anchor. The act is what you did — answered the guest thread, ran the turnover, met the plumber, updated pricing, reconciled the payout. The anchor is the thing that already existed for its own reasons: the message with its timestamp, the invoice, the receipt, the calendar entry, the bank line, the photo. Start and stop times come from when the work actually started and stopped, which is why real logs have untidy durations and estimates have round ones.

That is the difference a log kept as the work happens makes. It cannot produce a post-event estimate, because there is no post-event step — the entry is written the day the work is done, next to the record that proves it. It also converts the on-call problem instead of arguing with it: the hour you spent reachable does not count, but the twelve minutes you spent answering the guest at 9pm does, and only a real-time log ever captures those twelve minutes.

What your hours are actually worth

The estimator applies these rules the way the cases do — excluded categories dropped, then all seven tests run on what is left. Put your hours through it and it shows which test you pass or miss once on-call, travel and investor-type time come out. The full list of what counts and what never does is in how to prove your short-term rental hours.