How many stocks are under short sale restriction? 261 a day

There is a rule that quietly decides whether you can short a stock today, and most screeners do not show it. It is not obscure and it is not rare. On an ordinary session it applies to a few hundred US stocks, and roughly three quarters of them look perfectly ordinary in the closing table. This is a measurement of how often it fires, on 180 consecutive session pairs.

1. What the rule actually says

Rule 201 of Regulation SHO, usually called the short sale restriction or SSR, is a circuit breaker. The SEC's own description of the trigger is a decline "of at least 10 percent in one day", measured against "the covered security's closing price as determined by the listing market for the covered security as of the end of regular trading hours on the prior day". Once it fires, trading centres must "prevent the execution or display of a short sale order of a covered security at a price that is less than or equal to the current national best bid". In plain terms: you cannot hit the bid to get short. You have to post above it and wait for someone to come to you. Long sales are untouched, and orders marked short exempt are untouched. Two details matter for counting it, and both come from the SEC's staff guidance rather than from the rule text. The first is that the determination "is limited to regular trading hours" — a stock can be down 30% in the pre-market and still not be restricted, because the circuit breaker cannot be triggered before 09:30. The second is duration: once triggered, the restriction "will apply to short sale orders in that security for the remainder of the day and the following day". That second detail is why the number of stocks restricted today is not the same as the number of stocks that fell 10% today.

2. How many names trigger on an ordinary session

We took every US symbol with a daily bar, required a prior close of at least $1 and at least $1,000,000 of turnover on the day so the name is actually tradable, and compared the session low against 90% of the prior close. A median of 6,186 symbols qualified per session. There was no session in the sample without a trigger. The rule is not an exceptional-day event; it is a daily feature of the market, and on the busiest session in the window it touched 874 names.

3. The restricted list is twice the trigger count

Because the restriction runs through the following session, the set of names you cannot freely short on a given day is the union of today's triggers and yesterday's. So the practical answer is 261, not 130. About one name in five triggers again on the carry-over day. Counted properly, a median of 103 names on the list — 42.7% of it — did not trigger anything today at all: they are serving out the second day of a rule that fired yesterday, and nothing in today's price action will tell you they are there.

4. Half of them do not look restricted at the close

This is the part that catches people out. The trigger is an intraday touch. Nothing requires the stock to still be down when the bell rings. On the median session 50.9% of the day's triggers close back above the line that restricted them, and the middle half of sessions runs from 44.2% to 58.0%. A median of 7 names per session finish the day higher than they started it while still being under a short sale restriction. If you build your list of restricted names by scanning an end-of-day table for the biggest losers, you will miss about half of them, and the seven green ones will never occur to you at all. Put the two effects together — triggers that recover, and yesterday's names still serving their second day — and the closing table becomes a poor guide to the restricted list.

5. Cheap stocks trigger far more often

Splitting the qualifying universe by the prior closing price gives a clean gradient.

$1 to $59.45

$5 to $204.14

$20 to $1001.45

$100 and up1.1

A stock under $5 triggers the restriction 8.6 times as often as a stock over $100. That is not a statement about which is a better trade. It is a statement about where the rule lives: the low-price end of the market, which is also the end where the day trading screeners point. We have measured before that half the names on a 09:35 watchlist are gone by 10:00 and that the bar to be a top gainer is far higher than most people assume. This is the same neighbourhood, viewed from the short side.

6. What we had to check before trusting the count

The whole measurement rests on one assumption: that the daily bar we are reading reflects regular trading hours, since the rule ignores everything outside them. A daily bar that quietly included pre-market prints would inflate every number above, because the pre-market is exactly where a falling stock prints its worst price. So we checked it rather than assuming it. We pulled minute bars for 30 sampled sessions, 6,257 symbol-days in total, and put three values side by side: the daily bar's low, the low computed from minute bars between 09:30 and 16:00, and the low across the whole day including extended hours. The third row is the one that settles it. In about one symbol-day in six the extended-hours session went lower than regular hours did — and the daily bar did not follow it down. The bar is already on the basis the rule uses. We also recomputed the headline counts through a second, independently written pass that reads the files in the opposite order and stores them in a different structure. It produced the same session count, the same date range, and the same medians, with zero mismatches. One earlier check pointed the other way and was wrong, so it is worth naming. We first tried to verify the regular-hours low using our own one-second recordings, and it appeared that the daily low sat far below the recorded regular-hours low. The recordings begin when a symbol enters the scanner's universe, not at 09:30, so for a name added late in the morning the recording simply never saw the early low. The minute-bar comparison replaced it.

7. What this changes at the screen

Three things follow from the numbers, and none of them are a strategy. The restriction is not a prohibition. It is a change in how your order has to be worked: passive instead of aggressive, which means you may not get filled at all on the move you were trying to catch. That cost does not show up in a backtest that assumes you can sell at the bid. Limitations The trigger condition is evaluated from a daily bar's low against the prior daily close. We verified that this bar is on a regular-hours basis, but the real circuit breaker is evaluated by each listing market on its own tape in real time, and a print that is later cancelled or corrected can differ from what settles into a daily bar. Treat these counts as a close estimate of the official list, not as the official list itself. The universe is every US symbol carrying a daily bar, filtered only by a $1 prior close and $1,000,000 of turnover. That includes ETFs and other non-operating-company listings. Of 2,173,992 symbol-days in the window, 1,115,148 were kept: 909,386 fell below the turnover floor, 116,864 below the price floor, 31,938 had no prior close to compare against, and 656 were exchange test symbols. Rule 201 also contains exceptions we did not model, including short exempt marking, and the count does not tell you anything about whether shares were locatable or borrowable, which is a separate constraint entirely. The window is 180 session pairs from December 2025 to August 2026. It is one stretch of one market regime, and a quieter or more violent period would move every count here.