Open interest data: audit quality before interpreting it

Five original data audits covering account counts, venue coverage, record availability, volume consistency and weighted holder ratios.

Course outline

1. 1

   Long/short ratio: compare account counts with position size

   Calculate long-holder share, the account ratio and open interest from one ledger. More long accounts do not prove greater long exposure in the market.

   8 min

   In short: A long/short account ratio counts holders, not contracts. With 8 long and 2 short accounts, 80% of holders are long and the ratio is 4, yet both sides hold eight contracts and one-sided OI is eight. More long accounts do not prove greater long exposure.

2. 2

   Open interest data gaps: compare a consistent venue panel

   Separate real open interest changes from missing, stale or restored venue data. Reconcile visible totals against the same exchange panel at each timestamp.

   8 min

   In short: Compare aggregate open interest only across the same venue panel at both timestamps. When venue C first appears with 80, the visible total rises from 300 to 400, but the constant A+B panel only from 300 to 320; the other 80 is new coverage. A venue dropping out is missing data, not liquidations.

3. 3

   Open interest backtests: handle delayed data and revisions

   Replay open interest with the values actually available at decision time. Track delayed reports and revisions without leaking future information into the test.

   8 min

   In short: An open interest backtest must replay only the records that had arrived by each decision time. Store measurement time, arrival time and every revision separately; selecting a value by its measurement label, or applying a later correction, leaks future information into the test.

4. 4

   Open interest vs volume: check each interval for consistency

   Compare open interest changes with traded volume in matching contract units. Find interval-level discrepancies that disappear in daily totals.

   8 min

   In short: Under fixed contract units and one-sided OI, each interval must satisfy |ΔOI| ≤ V, where V is matched contract volume. Test every interval, not only the window total: an OI drop of eight on six traded contracts is a two-contract discrepancy even when the three-minute total passes.

5. 5

   Aggregate long/short ratios: weight account counts correctly

   Pool compatible long and short account counts across venues. See how a combined long share can rise while every single exchange's share falls.

   8 min

   In short: To aggregate long/short ratios across exchanges, sum long holders and total holders, then divide; do not average venue percentages. Weights matter: when each venue's long share falls ten points, the pooled share can still rise from 29/110 to 81/110 because the larger population is now at the high-share venue.

Open full-size diagram

1. Long/short ratio: compare account counts with position size
2. Open interest data gaps: compare a consistent venue panel
3. Open interest backtests: handle delayed data and revisions
4. Open interest vs volume: check each interval for consistency
5. Aggregate long/short ratios: weight account counts correctly

Educational material. Examples do not establish a profitable strategy. Trading costs, gaps and liquidation can produce losses beyond a planned stop.