Cboe U.S. Equities Historical Market Volume Data 2016

Rows: 3,622
Columns: 14

Daily historical market volume, notional value, and trade counts for U.S. equities exchanges and TRFs for 2016.

AI analysis

Dataset Analysis: Cboe U.S. Equities Historical Market Volume Data 2016

1. Dataset Overview & Analytical Value

This dataset, sourced from Cboe's official statistics portal (https://cdn.cboe.com/resources/us/equities/market-statistics/historical-market-volume/markethistory2016.csv), captures daily U.S. equity market activity across 16 distinct market participants throughout calendar year 2016. The data spans three reporting tapes — Tape A (NYSE-listed securities), Tape B (NYSE American and regional exchange listings), and Tape C (Nasdaq-listed securities) — plus aggregate totals, covering share volume, notional dollar value, and trade counts. With 3,622 rows representing 252 trading days × 16 participants, this dataset is structurally well-suited for longitudinal market microstructure analysis, exchange competition studies, and volume-driven correlation work. The Cboe provenance signals institutional-grade reliability, and the 2016 vintage captures a notable period including the U.S. presidential election volatility spike, making it particularly valuable for event-driven research.

2. Data Quality Observations

The dataset exhibits exceptional structural cleanliness: zero null cells across all 14 columns and all 3,622 rows ingested successfully, which is rare for financial time-series data of this granularity. The Total Shares and Total Trade Count columns each achieve 3,622 distinct values — one per row — confirming no accidental row-level aggregation collapse. The duplicate-row count is noted as pending Phase D recomputation, which warrants a follow-up check before any participant-level analysis, since the 16-participant × 252-day panel structure means even a single duplicate market participant entry on a given date could silently inflate aggregate calculations by up to 6.25%. The minimum values of zero across all Tape-level share, notional, and trade-count columns are worth scrutinizing: they likely reflect legitimate non-participation by certain market participants on specific tapes (e.g., a TRF reporting no Tape B activity), but should be validated against the participant taxonomy before being treated as missing data in disguise.

3. Key Column Distributions & Notable Patterns

The Total Shares column (mean: 511M, median: 270M, σ: 645M, skew: 2.18, outliers: 261) reveals the classic heavy-right-tail signature of equity volume data — the mean is nearly twice the median, indicating a relatively small number of high-volume days or participants dramatically pulling the distribution upward. The interquartile range spans 72M to 722M shares, a 10× spread that reflects the vast size differences between large exchanges and smaller TRFs. Tape A Shares carries the highest outlier count at 511 flagged observations, suggesting the NYSE tape is the primary driver of extreme volume days. Trade count columns tell a contrasting story: Tape A Trade Count (skew: 0.983, outliers: only 52) is the most normally distributed metric in the dataset, implying that while share volumes spike dramatically, the number of transactions on Tape A is comparatively stable — a pattern consistent with larger average trade sizes during high-volatility periods. Tape C Trade Count (skew: 1.66, Q1: 13,569, Q3: 1,027,480) shows a startling Q1-to-Q3 ratio of roughly 75×, reflecting extreme heterogeneity likely driven by the mix of high-frequency Nasdaq participants versus minimal-activity TRFs sharing the same tape.

4. Recommended Join Key Columns

The Date column (252 distinct values, zero nulls, typed as Date) is the primary and most reliable join key for cross-dataset correlation — it maps cleanly to any standard U.S. trading calendar for 2016. The Market Participant column (16 distinct string values, zero nulls) serves as the secondary join dimension, enabling participant-level matching against FINRA registration data, exchange fee schedules, or maker-taker rebate tables. Together, the composite key (Date, Market Participant) should yield a unique row identifier — which the 3,622 distinct Total Trade Count values implicitly support (252 × 16 = 4,032 theoretical maximum, with the ~410-row gap likely reflecting non-reporting days or participant absences). For time-series joins with macro datasets, Date alone is the recommended key, aggregating participant rows to daily totals first using Total Shares or Total Notional as the aggregation target.

5. Recommended Companion Datasets for Correlation Discovery

Several dataset categories would pair powerfully with this data. CBOE VIX daily closing values for 2016 would enable a direct test of the volume-volatility relationship — the presidential election period in November 2016 likely appears as a simultaneous spike in both VIX and Total Shares. Federal Reserve FRED economic indicators (e.g., daily Fed Funds rate, 10-year Treasury yield) would allow exploration of interest-rate sensitivity in notional trading values, particularly relevant given the December 2016 Fed rate hike. SEC market concentration or HHI data by exchange would complement the participant-level breakdown to assess whether volume fragmentation across Tape A/B/C correlates with broader market liquidity conditions. Finally, earnings announcement calendars (e.g., from Compustat or Bloomberg) would allow event-study overlays to determine whether Tape C volume spikes cluster around Nasdaq-heavy earnings seasons, given the dramatic Q1–Q3 spread observed in Tape C metrics.

Columns

Date (date)
Market Participant (string)
Tape A Shares (integer)
Tape B Shares (integer)
Tape C Shares (integer)
Total Shares (integer)
Tape A Notional (decimal)
Tape B Notional (decimal)
Tape C Notional (decimal)
Total Notional (decimal)
Tape A Trade Count (integer)
Tape B Trade Count (integer)
Tape C Trade Count (integer)
Total Trade Count (integer)