Reading Short-Selling Data: Short Interest, Short Volume & Cost to Borrow

Introduction

Whenever a stock gets a reputation for being "heavily shorted," three numbers get quoted to prove it: short interest, short volume, and cost to borrow. They sound interchangeable. They are not — and two of them are misread so routinely that the confusion has become its own source of bad decisions.

This lesson is a guide to reading short-selling data correctly: what each metric actually measures, how fresh it is, and the specific traps that trip people up — chiefly the belief that short volume is short interest, and that short interest above 100% of the float proves foul play. It follows the short selling and naked short selling lessons and assumes you know what a float is.

Short Interest: The Net Position (But Lagged)

Short interest is the total number of shares currently sold short and not yet bought back — the aggregate size of the bearish bet. In the US it is reported by FINRA roughly twice a month, based on settlement data, which means it reaches you with a lag of around two weeks.

It's usually expressed three ways:

• Raw shares — e.g. "15 million shares short."
• Percent of float — short interest ÷ free float. A high percentage means a large slice of the tradeable shares are shorted — crowded, and potential squeeze fuel.
• Days to cover (the short-interest ratio) — short interest ÷ average daily volume. It estimates how many days of normal trading it would take shorts to buy back everything. A high figure means the exit is narrow.

Its great weakness is the lag. By the time you read short interest, it describes positioning from up to two weeks ago; the real position may have changed completely. Treat it as a slow, backward-looking gauge of sentiment — useful, but never current.

A second point defuses a common myth: short interest can legitimately exceed 100% of the float. When a short seller sells borrowed shares, the buyer now holds real shares that can themselves be lent to another short seller. That re-lending chain means the same underlying shares get shorted more than once, so aggregate short interest can top the float without any naked shorting. It looks alarming; it is ordinary securities lending.

Short Volume: The Most Misread Number In Markets

Here is the one to get right. Short volume is a daily figure counting how many of that day's traded shares were marked as short sales. FINRA publishes it, and it is tempting to read a "60% short volume" day as "the stock is 60% shorted." That reading is wrong, and understanding why is the single most valuable thing in this lesson.

The reason is market makers. When a market maker fills your buy order but doesn't have the shares in inventory, it sells them to you short, then buys shares moments later to flatten its book. That sale is marked short in the data — even though the market maker ends the day with no net short position at all. On a liquid stock, a huge share of daily "short volume" is exactly this: routine, self-cancelling liquidity provision, not lasting bearish bets.

Short volume also only covers certain reporting venues, not the entire market, so it isn't even a complete count. The takeaway: short volume is a noisy, easily-misread figure. It is not short interest, and a high reading is not evidence of a heavy net short position.

Cost To Borrow: Often The Freshest Signal

The cost to borrow (CTB) is the annualised fee a short seller pays to borrow the shares, set moment-to-moment by supply and demand in the securities-lending market. Unlike short interest, it is essentially real-time.

That makes it, arguably, the most honest of the three:

• A low, stable CTB (a fraction of a percent) means shares are plentiful — "easy to borrow." Shorting is cheap and unremarkable.
• A high or rapidly-rising CTB — into the double or triple digits annually — means shares are scarce and short demand is intense: "hard to borrow." It's a timely sign of pressure that lagged short interest would miss for two weeks.
• Utilisation (the share of lendable stock actually out on loan) approaching 100% says the borrow pool is nearly exhausted — little room left to add shorts.

Because it reflects actual money changing hands now for the right to short, a spiking borrow fee is a genuine, current signal — where short interest is history and short volume is noise.

Putting Them Together

Read as a set, each answers a different question:

• Short interest → how large is the net short bet? (but two weeks stale)
• Days to cover → how hard would it be for shorts to exit? (squeeze potential)
• Cost to borrow → how much pressure is on the borrow, right now? (the timely one)
• Short volume → mostly noise; do not read it as net shorting

No single number tells the story, and none predicts a price. Together they sketch how crowded and stressed the short side is — a piece of context, never a trade signal on their own.

Common Misconceptions

• "High short volume means the stock is heavily shorted." The headline error. Most short volume is self-cancelling market-maker activity; it is not the net position.
• "Short interest over 100% of float proves naked shorting." No — re-lending lets the same shares be shorted repeatedly. It's normal securities lending.
• "Short interest is real-time." It lags about two weeks. For "now," watch the cost to borrow.
• "A high short interest means the price will drop." It's crowded bearish positioning — which is also stored-up forced buying if the stock rises. It cuts both ways.

Real-World Application

A trader hears a stock is "70% short volume" and about to collapse. Instead of reacting, they read the full picture. The 70% short volume they dismiss as mostly market-maker liquidity — not a net position. They pull the short interest: 8% of float, reported two weeks ago — moderate, and stale. Days to cover is 1.5 — shorts could exit in a day and a half, so little squeeze fuel. And the cost to borrow is 0.4% and flat — shares are plentiful, no borrowing stress at all. The composite picture is the opposite of the hype: this stock is not meaningfully squeezed or heavily net-shorted, and the scary "70%" was the most misleading number of the lot. They don't trade on a slogan; they read the data for what it actually measures. That discrimination — knowing which number is signal and which is noise — is the entire skill.

Key Takeaways

• Short interest = the net short position, reported ~twice a month with a ~two-week lag. Watch it as % of float and days to cover (squeeze potential), but know it's stale.
• Short interest can legitimately exceed 100% of float through re-lending — not proof of naked shorting.
• Short volume is the most misread metric: a daily count of trades marked short, much of it self-cancelling market-maker hedging, from only some venues. It is not short interest, and a high reading is not heavy net shorting.
• Cost to borrow is near real-time — a rising borrow fee (and high utilisation) is the freshest signal that shares are scarce and short demand is climbing.
• Read them as a set, as context — never as a standalone signal or a price prediction.