Short selling: asymmetric risk and the locate / borrow fee structure.

[01]The mechanics of a short sale

To sell short, you must borrow shares from someone willing to lend them, sell those borrowed shares in the market, and at some future point either return identical shares (covering the short) or have your broker close the position when the lender recalls. The cycle:

1. Locate: your broker confirms availability of shares to borrow before allowing the short. For widely-held large-caps, this is typically automatic. For thinly-held small-caps or stocks already heavily shorted, the “hard-to-borrow” or “HTB” flag applies, and the short may be unavailable or carry elevated fees.
2. Borrow: your broker transfers shares from a lender (typically a long-term holder via a securities-lending program) into your account. The lender retains economic ownership; you have a delivery obligation back.
3. Sale: you sell the borrowed shares in the market at the current bid. Proceeds are credited to your margin account as collateral.
4. Hold: while short, you pay the borrow fee (the “short interest rebate”) to the lender, typically expressed as an annualised rate. Easy-to-borrow stocks: 0.25–1 % annual. HTB: 5–100 %+ annual.
5. Cover: at any time, you buy identical shares in the market and return them to the lender, closing the short.
6. Recall: the lender may demand return of shares at any time (“recall”); if your broker cannot find a replacement lender, your position is forced-closed, called a “buy-in.”

[02]The asymmetric risk profile

A long position has bounded downside (the stock can only go to zero, capping loss at 100 % of position value) and unlimited upside. A short position has the opposite: bounded upside (the stock can only go to zero, capping gain at 100 % of short proceeds) and theoretically unlimited downside (the stock can rise without bound).

The implications are operationally severe:

• Position sizing: a short that triples in price produces a loss of 200 % of the original short proceeds, eliminating substantial portions of account equity.
• Margin requirements: brokers require higher initial margin and tighter maintenance on shorts to compensate for the asymmetric risk.
• Position aging: long-only positions can be held indefinitely at zero ongoing cost. Shorts accrue daily borrow fees that compound over the holding period.

[03]Borrow fees and the cost of carry

The annualised borrow fee on a short position is set by the securities-lending market based on supply and demand. For large-cap easy-to-borrow stocks, fees are minimal (often 0.25–1 %). For stocks heavily shorted, with limited float, or with restricted lendable supply, fees can be very high — routine HTB rates are 5–25 %, and crisis-period rates can spike above 100 %.

Worked example: short 500 shares of a small-cap at $40 per share, total proceeds $20,000. Borrow fee: 25 % annualised (typical HTB). Daily fee: $20,000 × 0.25 / 365 = $13.70. Over a 30-day hold, the borrow fee alone is $411 — substantial relative to a typical price-target gain.

The 100 %+ borrow fee scenario is rare but real, particularly during meme-stock-style episodes or short-squeeze peaks. At 200 % annualised borrow fee, every day held costs the trader 0.55 % of the position value. Holding for a week consumes 4 % of position value in borrow fees alone, before any price move.

[04]The short-squeeze risk

A short squeeze occurs when a stock heavily shorted by retail or institutional traders begins rising rapidly. Rising price triggers margin calls on existing short positions, forcing those traders to cover — which means buying the stock, which adds further upward pressure, which triggers more margin calls. The feedback loop produces extraordinarily rapid price increases that are largely disconnected from underlying fundamentals.

Notable historical squeezes: Volkswagen 2008 (briefly the world’s most valuable company by market cap during the squeeze; subsequently reverted), GameStop January 2021 (rose from ~$20 to over $480 in two weeks; many short traders catastrophically liquidated), AMC Entertainment May–June 2021. The pattern: extreme price increases over days-to-weeks, sometimes 5–25× the pre-squeeze price, sometimes a partial reversion afterward.

Indicators of squeeze vulnerability: high short interest as a percentage of float (above 20 % is elevated, above 40 % is dangerous), low float (limited supply for short covers), retail-investor enthusiasm or coordinated buying campaigns, recent positive catalyst that broke the short thesis. Avoid shorting names that exhibit two or more of these characteristics regardless of how attractive the fundamental short case looks.

[05]The buy-in risk

If a stock loan is recalled by the lender and your broker cannot find a replacement lender, your short position is force-closed via buy-in: the broker buys shares in the market at any prevailing price and returns them to the lender. You lose any pricing discretion. In an active short squeeze with limited lendable shares, buy-ins occur at the worst possible prices; the short is closed at the peak of the price move, crystallising maximum loss.

Buy-in risk is highest in HTB names with concentrated short interest, exactly the names where short squeezes occur. If you are short an HTB name that is rising sharply, your position is at risk both from the margin-call mechanism and from the buy-in mechanism, often with the latter triggering before the former. The insight: for HTB shorts, your time-to-cover is shorter than the margin-call calculator suggests, because the broker may force-close on a separate trigger.

[06]The dividend-payment obligation

If a shorted stock pays a dividend during your short, you (the borrower) are obligated to pay the dividend to the original lender. The mechanic is automatic in most retail short accounts: the dividend amount is debited from your account on the ex-dividend date. The economic effect is reasonable (the lender shouldn’t lose the dividend by lending) but operationally it’s an additional cost-of-carry on top of the borrow fee.

The implication for shorting dividend-paying stocks: factor in the expected dividend stream over the hold period as an additional cost. For a 4 %-yielding stock held short for one year, the dividend obligation is 4 % of position value — a meaningful drag on short P/L. This is one of the structural reasons short interest concentrates in low- or no-dividend names; the cost-of-carry on dividend-payers is too high for many short theses.

[07]The Reg SHO uptick / locate-rule framework

SEC Regulation SHO, in force since 2005 (with amendments), governs short selling in US securities. Key provisions:

• Locate requirement: brokers must have reasonable grounds to believe shares can be borrowed before allowing a short sale. The pre-trade check eliminates most “naked” shorts.
• Threshold list: stocks with persistent fail-to-deliver positions are placed on a regulatory threshold list; brokers must close out fails within 13 days.
• Alternative uptick rule (Rule 201): when a stock falls 10 % or more from the prior day’s close, short sales for the rest of that day and the following day are restricted to prices above the national best bid — preventing piling-on shorts during sharp declines.

The framework does not prevent shorting; it constrains specific patterns associated with market-manipulation concerns. For most retail shorts on liquid names, the framework is invisible — the broker handles the locate automatically and the alternative uptick rule rarely binds.