Thread by @ReformedTrader

1/ Borrowing Fees and Expected Stock Returns (Hendrix, Crabb)

"Stocks with high borrowing fees tend to underperform their peers over the short term, but persistence of high borrowing fees is fast-decaying and not systematically predictable."

papers.ssrn.com/sol3/papers.cf…

2/ "Lending data are sourced from a proprietary database maintained by DFA that contains the global lending activity of its US mutual funds.

"Fee = fed funds rate (or OBFR) – rebate rate.

"We compute a weighted average (by number of shares in each loan) fee for each security."

3/ Small caps are the 10% lowest U.S. market cap, 12.5% in developed ex-U.S., and 15% smallest in emerging markets

Small caps tend to have higher borrowing fees than large caps, more stocks on loan, and a greater dispersion in fees required to borrow the stocks.

4/ High-fee = 50+ percentile (market cap weighted fees)

For a one-day lookback window, "stocks on loan with a high fee have annualized returns that are 8.85% lower than for stocks not on loan.

"We observe the same pattern using three-day and five-day lookback periods."

5/ Again for small caps, "across all countries in every year from 2011 through 2018, high-fee stocks, on average, underperform not-on-loan stocks."

6/ For the one-day lookback window, "in all countries, small cap high-fee stocks underperform small cap stocks not on loan.

"Results for the three-day and five-day periods are consistent."

7/ Again for small caps, "high-fee stocks not only underperform stocks not on loan on the next day but also continue to meaningfully underperform in the week following portfolio sorts. The magnitude of annualized average underperformance is similar."

8/ Again for small caps, "stocks that remain high fee tend to persist in underperformance, while high-fee stocks that migrate out of that category tend to perform in line with stocks not on loan."

NOTE: "We are forming portfolios with information not known at time t."

9/ "One year after we assign small cap stocks to high- and low-fee categories, only half of high-fee stocks remain in their starting category, on average."

10/ "Small cap stocks in the top fee decile and with a high utilization are more likely to also have a high fee over the subsequent five days, one month, six months, or one year than other high-fee stocks."

11/ "High-fee stocks with lower market cap and higher P/B have a higher likelihood of staying high-fee, but this explains only 3.6% of the variation.

"The model assesses point-to-point persistence; stocks are even less likely to stay high fee consistently throughout the year."

12/ "A daily-rebalanced strategy that sells small caps when they are high-fee and buys them back when they cross the threshold would have increased turnover of 1,153%/year.

"Excluding only the top fee decile, high-utilization stocks still has 120%/year additional turnover."

13/ "It would also forgo the high lending revenue from high-fee stocks and reduce diversification.

"Given these considerations, an efficient approach may be to temporarily refrain from purchasing the most expensive-to-borrow stocks."

Related research:

https://twitter.com/ReformedTrader/status/1323470487171706884

14/ "Among *large cap* stocks, stocks on loan with a high fee on average underperform stocks not on loan by 7.44% annualized, with a t-statistic of –2.56.

"Results are similar among *all cap* stocks, though slightly stronger due to the inclusion of small caps."

15/ Related reading:

Equity Factors: To Short Or Not To Short, That is the Question

https://twitter.com/ReformedTrader/status/1252013541487304707

The Loan Fee Anomaly: A Short Seller's Best Ideas

https://twitter.com/ReformedTrader/status/1324519224279064576

Shorting Costs and Profitability of Long–Short Strategies

https://twitter.com/ReformedTrader/status/1249502722253971457