Bybit
One or more volume-pattern metrics deviate noticeably from the peer-basket reference. The per-metric breakdown below shows which components are driving the score.
Volume patterns sit in the middle of the peer-basket range. Some metrics show modest deviation from reference venues but none is in the atypical zone.
M01Atypical · Low
Organic trading produces trade sizes that follow predictable mathematical patterns across many orders of magnitude. We test both the leading and second leading digits of trade sizes against Benford's Law using a sample-size-invariant χ²/N statistic. A score above 70 indicates natural size diversity; below 40 suggests round-number clustering or uniform-size bot patterns. Requires a minimum of 1,000 trades.
M01Typical
Organic trading produces trade sizes that follow predictable mathematical patterns across many orders of magnitude. We test both the leading and second leading digits of trade sizes against Benford's Law using a sample-size-invariant χ²/N statistic. A score above 70 indicates natural size diversity; below 40 suggests round-number clustering or uniform-size bot patterns. Requires a minimum of 1,000 trades.
M02Typical
Reported volume should be proportionate to the observable liquidity that could plausibly absorb it. We compute the ratio of 24h trade volume to mean ±2% bid-ask depth across 1,440 daily snapshots; both extremes score lower than the middle in v1 scoring. High ratios suggest volume may exceed what the visible order book could plausibly absorb. Low ratios indicate deep liquidity relative to reported flow, a structural pattern common in market-making-dominant venues, not a quality failure in itself. The score does not yet distinguish direction; a forthcoming methodology revision will reflect the asymmetry directly. Requires 50% snapshot coverage.
M02Typical
Reported volume should be proportionate to the observable liquidity that could plausibly absorb it. We compute the ratio of 24h trade volume to mean ±2% bid-ask depth across 1,440 daily snapshots; both extremes score lower than the middle in v1 scoring. High ratios suggest volume may exceed what the visible order book could plausibly absorb. Low ratios indicate deep liquidity relative to reported flow, a structural pattern common in market-making-dominant venues, not a quality failure in itself. The score does not yet distinguish direction; a forthcoming methodology revision will reflect the asymmetry directly. Requires 50% snapshot coverage.
M03Atypical · Low
Genuine market activity arrives in irregular bursts driven by many independent participants. We measure Shannon entropy of inter-arrival time distribution across multiple time scales and lag-1 autocorrelation. High entropy with low autocorrelation indicates organic flow; bot activity produces rigid, regular spacing. A score above 70 indicates natural timing patterns; below 40 suggests mechanically-timed trade submission. Requires 5,000 trades.
M03Typical
Genuine market activity arrives in irregular bursts driven by many independent participants. We measure Shannon entropy of inter-arrival time distribution across multiple time scales and lag-1 autocorrelation. High entropy with low autocorrelation indicates organic flow; bot activity produces rigid, regular spacing. A score above 70 indicates natural timing patterns; below 40 suggests mechanically-timed trade submission. Requires 5,000 trades.
Order-book quality is broadly in line with the peer-basket reference. Effective spread and depth metrics are competitive; quote stability may sit at the edge of the typical range.
Order-book quality is broadly in line with the peer-basket reference. Effective spread and depth metrics are competitive; quote stability may sit at the edge of the typical range.
M04Typical
The effective spread captures the true cost of trading, accounting for where execution actually occurs relative to the mid-price. We compute a volume-weighted average of signed price deviation from mid across all trades using Lee-Ready direction inference where needed, expressed in basis points. Below 5 bps is excellent; 5-15 bps is typical; above 50 bps is atypical for major pairs. Requires 50% order book snapshot coverage.
M04Typical
The effective spread captures the true cost of trading, accounting for where execution actually occurs relative to the mid-price. We compute a volume-weighted average of signed price deviation from mid across all trades using Lee-Ready direction inference where needed, expressed in basis points. Below 5 bps is excellent; 5-15 bps is typical; above 50 bps is atypical for major pairs. Requires 50% order book snapshot coverage.
M05Typical
Market depth at the top of the book can be cosmetic. What matters is how much price moves when real size is executed. We simulate sweeping the book at standard sizes and measure the average slippage in basis points using 1-minute snapshots. A steep slope means large trades move price significantly; shallow venues absorb size with minimal impact. Returns insufficient data when snapshot depth fails to reach the target band.
M05Typical
Market depth at the top of the book can be cosmetic. What matters is how much price moves when real size is executed. We simulate sweeping the book at standard sizes and measure the average slippage in basis points using 1-minute snapshots. A steep slope means large trades move price significantly; shallow venues absorb size with minimal impact. Returns insufficient data when snapshot depth fails to reach the target band.
M06Atypical · Low
Healthy market-making produces a moderate ratio of order book updates to executed trades. Too few updates suggests a sleepy book; too many suggests quote stuffing without genuine trading intent. We compute book events per trade over 24h using WebSocket data and score the ratio against a non-monotonic curve peaking in the 50-1,000 range. Requires an active WebSocket feed; venues without WebSocket coverage return insufficient data for this metric.
M06Typical
Healthy market-making produces a moderate ratio of order book updates to executed trades. Too few updates suggests a sleepy book; too many suggests quote stuffing without genuine trading intent. We compute book events per trade over 24h using WebSocket data and score the ratio against a non-monotonic curve peaking in the 50-1,000 range. Requires an active WebSocket feed; venues without WebSocket coverage return insufficient data for this metric.
Mid-price tracks the global market closely and impact dynamics on large trades match the reference pattern. Both metrics in this dimension sit near the best of the peer basket.
Mid-price tracks the global market closely and impact dynamics on large trades match the reference pattern. Both metrics in this dimension sit near the best of the peer basket.
M07Typical
A well-connected exchange should track global prices closely. Persistent deviation suggests stale feeds, isolated price formation, or limited arbitrage activity. We compute mean absolute deviation of the venue's mid-price from a reference basket of peer venues at 60-second intervals. Below 5 bps is typical; above 30 bps is atypical; above 50 bps indicates significant price isolation.
M07Typical
A well-connected exchange should track global prices closely. Persistent deviation suggests stale feeds, isolated price formation, or limited arbitrage activity. We compute mean absolute deviation of the venue's mid-price from a reference basket of peer venues at 60-second intervals. Below 5 bps is typical; above 30 bps is atypical; above 50 bps indicates significant price isolation.
M08Typical
After a large trade, prices should move permanently in the direction of the trade if flow is informed. Wash trades revert fully because no real information was exchanged. We measure the 60-second post-trade reversion ratio for trades above $50k notional; venues where most large trades fully revert score lower. A score above 70 indicates genuine price impact consistent with informed flow. Requires 50 qualifying large trades per day.
M08After a large trade, prices should move permanently in the direction of the trade if flow is informed. Wash trades revert fully because no real information was exchanged. We measure the 60-second post-trade reversion ratio for trades above $50k notional; venues where most large trades fully revert score lower. A score above 70 indicates genuine price impact consistent with informed flow. Requires 50 qualifying large trades per day.
Volume and liquidity shares are broadly in proportion across the peer basket; price-discovery contribution is in line with the venue's size. The dimension sits in the upper half of the peer range.
Cross-venue characteristics sit in the middle of the peer-basket range. The volume-to-liquidity ratio and price-discovery contribution are roughly consistent with the venue's position in the basket.
M10Atypical · Low
An exchange's share of global trading volume should be proportionate to its share of global liquidity. A venue claiming disproportionate volume relative to its depth is flagging a structural inconsistency. We compute the ratio of volume share to depth share against a peer basket using the same ±2% band as M02. Both extremes of this ratio score lower; the healthy range sits between 0.7 and 1.5.
M10Typical
An exchange's share of global trading volume should be proportionate to its share of global liquidity. A venue claiming disproportionate volume relative to its depth is flagging a structural inconsistency. We compute the ratio of volume share to depth share against a peer basket using the same ±2% band as M02. Both extremes of this ratio score lower; the healthy range sits between 0.7 and 1.5.
M11Typical
Exchanges that generate genuine price discovery lead the market; others follow. We measure Pearson correlation of the venue's per-minute returns against a peer basket at seven forward lead times (1, 2, 5, 10, 15, 30, 60 minutes) and score the maximum correlation. A high score means this venue's price moves predict where other venues will be minutes later; a low score indicates the venue follows rather than leads global price formation.
M11Typical
Exchanges that generate genuine price discovery lead the market; others follow. We measure Pearson correlation of the venue's per-minute returns against a peer basket at seven forward lead times (1, 2, 5, 10, 15, 30, 60 minutes) and score the maximum correlation. A high score means this venue's price moves predict where other venues will be minutes later; a low score indicates the venue follows rather than leads global price formation.