Skip to content
All library documents

Estimating Hard-to-Borrow Risk in Equity Short Backtests

Article Quant Q&A · Author: siegel
Read the original

Summary

The document asks whether a low-frequency US equity statistical arbitrage strategy can assume that liquid stocks are available to borrow. The strategy holds roughly equal-sized long and short stock baskets, trades at the open, and replaces much of its portfolio daily. Its liquidity screen uses minimum share price and dollar-volume thresholds, while its slippage model estimates trading costs; neither measure directly confirms stock-borrow availability.

The response emphasizes that hard-to-borrow status depends on the brokerage relationship, and that locate access may be available through third parties. It offers an informal estimate that most of the described universe may be borrowable, while flagging recent IPOs as a group that can remain costly or unavailable regardless of trading volume. The answer gives no supporting data for that estimate and identifies no source for historical borrow-status records. For a realistic backtest, borrow availability and fees would need broker-specific historical data, such as archived status feeds; liquidity alone is not sufficient evidence.

Key ideas

  • Borrow availability depends on a trader's broker and locate relationships.
  • High share price and dollar volume do not guarantee that a stock can be borrowed.
  • Recent IPOs may be difficult or expensive to locate even when they trade actively.
  • The response's estimate that most screened stocks are borrowable is informal and unsupported by data.
  • Historical borrow-status testing may require records collected from broker feeds.

Tags

EquitiesUS marketsBacktestingExecution
Full text
# possible to estimate if hard-to-borrow?


# possible to estimate if hard-to-borrow?












I'm building a low frequency US equity stat arb system. On any given day the system is long ~100 stocks and short ~100 stocks. It trades once a day at the open, and on average 4/5 of the portfolio gets replaced each day.

I am using the formula on page 21 of http://www.courant.nyu.edu/~almgren/papers/costestim.pdf as my slippage cost model.

As a liquidity filter, I'm only trading stocks with minimum price of 5 dollars and minimum daily dollar volume of 30 million dollars.

Is it safe to assume that these fairly liquid stocks are all possible to borrow for shorting? If not, when you backtest a system, is there a way to estimate if a stock could have been shorted, or is there a place to buy that information?

## Answer by Nathan S. (score 0, accepted)

https://quant.stackexchange.com/a/17040

The list will differ depending on your brokerage relationship. Not every broker's hard-to-borrow list will be the same. And you may be able to use 3rd party relationships purely for locates (finding hard-to-borrows).

Your liquidity requirements look reasonable. I'm just going to pull a number out of the air (although I see it, this isn't something I track) and say that more than 90% of this universe will not be hard-to-borrow. You might also add recent IPOs, because they're usually unavailable or expensive to get locates for a while regardless of volume.

If you have the right relationships then there will be quite a few hard-to-borrow cases, but for a price you can still get most shares.

I don't know where you would get historical data on this to test, because I think you would just cache it off some status feed from your brokers.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.