Academic Alpha

Short-Selling Risk: A Hidden Limit to Arbitrage

Short sellers face unique risks, including loan recalls and rising borrow fees. This paper shows that higher short-selling risk predicts **lower future returns**, **more mispricing**, and **less efficient prices**. These effects are stronger for trades with long expected horizons.

đź’ˇ Takeaway:
Short-selling risk—especially the variance of future loan fees—is a strong return predictor and source of persistent mispricing. It’s a true limit to arbitrage, particularly when trades require long horizons.

Key Idea: What Is This Paper About?

This paper introduces and tests a dynamic measure of short-selling risk based on the forecasted variance of equity loan fees. Unlike static proxies (like current borrow fees), this forward-looking risk constrains arbitrage and limits short-seller participation. Stocks with higher short-selling risk have lower returns, higher pricing errors, and lower trading volume, even after controlling for short interest. The effect is stronger when trades take longer to resolve.

Economic Rationale: Why Should This Work?

đź“Ś Relevant Economic Theories and Justifications:

  • Limits to Arbitrage (Shleifer & Vishny, 1997): Risk of being bought in or facing fee spikes deters arbitrage.
  • Loan Recall and Fee Volatility: Borrow costs change unexpectedly, eroding profits.
  • Holding Horizon Risk: Mispricing that requires long horizons is especially unattractive to short sellers.
  • Price Efficiency Impairment: Higher short-selling risk leads to slower information incorporation.

đź“Ś Why It Matters:
Explains why short interest remains a strong return predictor despite being public—because many can’t or won’t act on it due to dynamic borrowing risks.

Data, Model, and Strategy Implementation

Data Used

  • Period: July 2006 – Dec 2011
  • Equity Lending Data: Markit
  • Market Data: CRSP, Compustat, OptionMetrics, TAQ
  • Options Data: Used for put-call parity tests
  • Sample Size: 220,000 firm-months, ~4,500 US stocks

Model / Methodology

  • ShortRisk: Forecasted variance of loan fees using lagged lending + firm data
  • Forecasting Inputs:
    • Variance of new loan fees and utilization
    • Tail risk proxies (99th percentile)
    • Fails-to-deliver, IPO flag, option presence, volatility
  • Portfolio Sorts: Quintiles on ShortRisk
  • Tests:
    • Fama-MacBeth regressions
    • Five-factor alphas
    • Price delay (Hou & Moskowitz, 2005)
    • Put-call parity arbitrage (long horizon test)

Trading Strategy (From ShortRisk Signal)

  • Long: Low ShortRisk stocks (more liquid to short)
  • Short: High ShortRisk stocks (costlier/riskier to short)
  • Rebalancing: Monthly
  • Filters: Can overlay with short interest, size, or mispricing signals
  • Enhancement: Focus on Micro/Small caps and long-horizon mispricings (e.g., put-call parity deviations)

Key Table or Figure from the Paper

📊 Reference: [Figure 1] – Long-Short Portfolio Returns by Short-Selling Risk

đź“Ś Explanation:

  • Shows monthly and cumulative returns to a strategy that buys low ShortRisk stocks and shorts high ShortRisk ones.
  • Annualized alpha: 9.6%
  • FF5-adjusted alpha: 0.80%/month
  • Effect is not subsumed by short interest
  • Impact is strongest for small caps and long-duration arbitrage setups

Final Thought

💡 Short-selling risk is the friction that keeps mispricing alive—even when it’s visible to all. 🔍📉

Paper Details (For Further Reading)

  • Title: Short-Selling Risk
  • Authors: Joseph Engelberg, Adam Reed, Matthew Ringgenberg
  • Publication Year: 2018
  • Journal/Source: Journal of Finance
  • Link: https://doi.org/10.1111/jofi.12601
Share Tweet Share Share Email

Read next