ShortAtlas methodology
Short selling data limitations
Public short-selling datasets are useful reference data, but they are delayed and limited. Learn the main limitations before interpreting them.
Public data is not real-time
Most public short-selling datasets are published after the trading activity or settlement date they describe. They should not be treated as live market data.
ShortAtlas shows generated-at timestamps and source dates so users can distinguish the data cycle from the current market session.
Coverage and definitions can differ
Source files can differ in coverage, reporting context, market center, instrument eligibility, and publication timing. A missing record does not automatically mean there was no activity.
Daily short-volume data, Short Interest, fails-to-deliver data, borrow fees, and lending availability answer different questions and should not be merged into one signal.
No investment advice
ShortAtlas avoids squeeze scores, bearish labels, bullish labels, target prices, and predictive wording. The goal is to make reference data easier to inspect.
Users should verify important decisions against original sources and professional advice where appropriate.
Questions this data cannot answer
Public short-selling files do not identify who holds a position, when it was opened, at what price, or whether it is hedged against something else. They also do not show borrow fees, lending availability, or recall risk.
Because of that, no combination of the fields on a ticker page can establish that a position is profitable, crowded, or about to be closed. Treat the data as a description of what was reported, not as a window into intent.