Short Volume

One of the Market Radar screens. It shows short selling three ways, on three tabs of one screen.

Selling a stock short means borrowing shares you do not own, selling them, and buying them back later. A short seller makes money if the price falls, and loses money if it rises.

Open Market Radar

At a glance

People mean several things by "how shorted is this stock". This screen keeps them apart, on separate tabs, measured by different rules on different clocks.

• Who uses this. Short sellers, squeeze hunters, and options traders pricing calls and puts on a crowded name.
• Squeeze watch, the tab the pane opens on. Reported short interest, joined to the options board's own figures for the same names. And to the model's signal on each one, the same one Trade Cards prints.
• Daily tab, labelled Daily marking on screen. A file published every day. It shows how many of the trades in that file were marked short, read against that name's own last 60 trading days.
• Positions tab, labelled Reported positions on screen. A count taken twice a month of how many shares are held short. The change since the last count and the days to cover are both stated and drawn.
• Two clocks. The daily figure, and Squeeze watch's options and signal columns, are one trading day old. Short interest and days to cover can be three weeks past.

Key difference: short volume is trading activity, measured every day. Short interest is a position, measured every two weeks.

Squeeze watch

The reason an options trader opens this pane at all. It publishes no new reading. It places the short interest this screen already holds against the options figures the desk already serves for the same names. And against the model's own signal, on one screen.

• What is plotted. Short interest across, IV rank up. A dot is sized by tonight's options volume against its own average. Its colour is the model's side: solid for a Trade Cards signal, faint for an outlook short of one, grey for no signal.
• Squeeze risk. A crowded short that starts covering can push the price up fast. Buying under that kind of pressure tends to lift call prices and implied volatility with it.
• Borrowing cost. A name that is crowded and hard to borrow often prices its puts higher too. A scarce borrow already makes shorting the stock itself expensive.
• Days to cover. The more days of ordinary trading it would take to close the whole reported position, the further a covering rally can run once one starts.
• IV rank, on this tab, is the share of a name's own weekly implied volatility readings that sit below tonight's, 0 to 100. It reads against that name's own history, never against another name's.
• Put/call, on this tab, is puts over calls by open interest. Above 1 means more puts are open than calls.
• Model signal is the same bullish or bearish signal Trade Cards and the Market Scan print, over the same time frame. A signal is the top of the ranking with a measured edge. An outlook is a measured side short of a signal. No signal means no measured edge on that name.
• Two ages, one screen. Short interest and days to cover are the same fortnightly settlement the positions tab states below. IV rank, put/call, options volume and the model's signal are the latest published figures. Neither age is guessed from the other.
• Nothing here is ranked by size. The table opens in ticker order, same as every other table on this screen. Reordering by a column is yours.

The daily tab

Daily marking

A high daily figure is not a crowd betting against the company. Most of it is market makers doing their job. A market maker is the dealer who takes the other side of your order.

• Most short marked volume comes from dealers covering their own risk. When you buy, the dealer on the other side sells short and buys it back later. That trade is recorded as a short sale in exactly the same way a bearish one is. Bearish means betting the price will fall. A dealer doing this is hedging, which means taking one position to offset the risk of another.
• Nothing cancels out during the day. A share sold short and bought back an hour later still counts in the file.
• Since its own last 60 days, the figure the tab now leads with. The share of this name's own last 60 trading days, tonight included, at or below tonight's reading, 0 to 100. It says whether tonight is ordinary or unusual for this name, not for the market.
• The five day average. The column next to today's figure. It is the same marked short share, averaged over the last five trading days. That smooths out one noisy trading day so it does not look like a change in the name's usual level.
• Compare a name against itself. Compare today's figure to what this same stock usually reads over the past year. The screen shows that usual level as the median, the middle reading of the past year. That comparison tells you something.
• Comparing two names tells you little. How much dealer hedging a stock attracts depends on the stock itself. Across names, the figure mostly shows which of them options traders use.

The figure is a fact about how trades were marked. Two names showing the same number can be doing entirely different things.

What this tab is, and is not. It reads each name against its own recent record, not against other companies. A reading of 40 to 50 percent here is ordinary, mostly market maker hedging rather than a bet against the company.

What the daily figure is measured against. The daily file counts only the volume that went through the reporting facilities it covers. That is mainly off exchange volume. Trades on the exchanges themselves are not in it. So the figure is short marked volume divided by reported volume, not by everything that traded. The number you divide by is called the denominator. This one is the file's own, not the whole market's.

One line on screen says what share of all the trading that day the daily file actually covered. It gives that share for the middle name of the ones measured. It uses the latest trading day both files share. That line names its own date only when it differs from the date already shown at the top of the tab. The lower that line reads, the smaller the slice of the day's trading these figures are measured against.

Sort order. The table arrives ordered by ticker. You can sort by the figures, but the screen never arrives ranked by them.

The positions tab

Reported positions

This tab shows the short position reported in each covered name, as most recently published. A covered name is a company that appears in the file we hold.

• The settlement date. Firms count their short positions on a set date. That is the settlement date, the day those trades officially completed. The tab states that date once at the top. Publication runs about a week behind, so a reading here can easily be three weeks old.
• Shares short. The number of shares currently sold short and not yet bought back, as firms reported it.
• Percent of shares outstanding. The reported short shares divided by the company's own most recently filed share count. The date of that filing is on the row.
• Days to cover. The reported short shares divided by average daily volume. Roughly how many normal trading days of buying it would take to close the whole reported position. Drawn as a short bar beside the number.
• The change column and the small chart on each row. Both are worked out from the reported share counts, not from the percentages. So if a company issues more shares, the percentage moves but the change column does not. The change also draws as a small bar either side of a centre line. Which side fills says the direction.

Why our percentage is smaller than other sites'. Shares outstanding is every share the company has issued, as filed with the regulator. A float is somebody else's estimate of how many of those shares are actually available to trade, by rules they do not publish. We divide by shares outstanding, which is the larger number. So the percentage here reads lower than the float based figure on most other sites.

The percentage is not carried on every row. The tab counts how many rows carry it and how many are blank.

Days to cover is arithmetic on two figures, not a measure of what anyone intends to do. It moves when volume moves, even if nobody's position changed.

The three measurements against each other

A name can have very high short volume and very few shares actually held short. That is ordinary rather than contradictory. It usually means market makers are hedging customer buying, not that anyone is building a bearish position.

What this screen will never show

• No trade signals. This screen shows reported short volume, short interest and the options market's own figures. That is not a trade idea, and no name is flagged as one to take.
• Borrowing costs vary by broker. Selling short means borrowing shares first, and the lender charges for that. Some names cost a lot to borrow. Some cannot be borrowed at all. Only your own broker can tell you the cost.
• No threshold for "heavily shorted". The screen prints the reported count and the days to cover. It names no level as a trade trigger.

Dark pools

Dark pools are private venues where large trades are recorded away from the public exchanges. Roughly forty percent of US volume goes through venues of that kind. This screen shows short sale volume only. It does not show dark pool trades.

Using it with discipline

• Read the age before the figure, on every tab. The daily figures, and Squeeze watch's options and book columns, are one trading day old. Short interest and days to cover describe a settlement weeks past.
• A high figure is a fact about the past. It says nothing about what comes next. Names sit with a large short position for years.
• Options activity can move the daily figure on its own. A name that has just become popular with options traders shows more dealer hedging. That means more short marked volume, and not one person in that sequence is betting against the company.

Frequently asked questions

Is a high figure a sign that people are betting against the company? On its own, no. That is the single most common misreading of this data, on any tab.

Why does the positions figure not change for days at a time? Because it genuinely does not change. It is republished twice a month, and the last figure stands until the next one.

Why does the percentage here read lower than another site's? Almost always because we divide by a different number. We divide by filed shares outstanding. Most sites divide by an estimated float, which is smaller. On the daily tab, sites also differ on which venues they include. So do not compare our percentage to another site's percentage. Compare the direction each one is moving instead. That comparison holds up better than the level does.

Does Squeeze watch tell me a squeeze is coming? No. It places a reported short position beside how the options market prices the same name tonight, and beside the model's signal on it. The signal is the model's published view on the stock over its time frame. It does not say whether a squeeze happens. No name is flagged as likely to squeeze.

Can I see the history? Yes. The daily tab draws each name's own recent record as a small chart. The positions tab draws the last twelve published share counts the same way.

Related panes

• Options Snapshot, where the hedging behind much of the daily figure, and the options figures Squeeze watch joins in, actually show up.
• Public Filings, what insiders and funds disclosed.