SymbologyData
Regulatory & Financial Data Platform
← Back to Execution Quality

Execution Quality Methodology Specification

exq-2

Every parameter and formula powering the FINRA 5310 regular and rigorous review reports is immutable, versioned data served directly by the platform API.

Core Execution Quality Parameters

Methodology Version
exq-2
Current production standard
Min Orders per Cell
100
Cells below are suppressed
First Benchmark Month
202608
Amended Rule 605 start date
NPI Max Amount / Mid
1.0×
Excludes corrupt vendor rows

The 144 Benchmark Tiers

Amended Rule 605 groups orders into cells defined by the 4-tuple key: (universe, order_class, notional_bucket, lot_class). This yields 2 universes × 3 order classes × 8 notional buckets × 3 lot classes = 144 cells per reporter-month.

8 Notional Buckets ($):

$0$250$1000$5000$10000$20000$50000$200000

3 Lot Size Classes:

FRACTODDROUND

2 Market Universes:

sp500_proxyother

606 to 605 Order Class Mapping:

606 Order Class Benchmarked Amended 605 Order Types
market MXXNN
marketable_limit LYNNN, LYNYN
nonmarketable_limit LNYNN, LNYYN, LNNNN, LNNYN

Excluded order types: Stop orders (MXXNY, LYNNY, LNNNY), other 606 order types, and the 606 options section.

Tracked Execution Quality Metrics

Metric Code Direction Aggregation Weight W Description
fill_rate Higher is better Executed Shares Here E Fill rate at market center: Executed Here E / Covered Shares C
pi_rate Higher is better Quote-Relative Shares Q Price-improved share rate over quote-relative shares: P / (P + A + O)
npi_bps Higher is better Executed Notional (Σ Q × mid) Net price improvement in basis points of executed notional: 10⁴ × (P·p̄ − O·ō) / Σ Q·mid
efq_pct Lower is better Executed Shares Here E Effective-over-quoted spread percentage: avg_efq_pct
eff_sprd Lower is better Executed Shares Here E Dollar average effective spread per share: avg_effective_spread
eff_sprd_pct Lower is better Executed Shares Here E Effective spread as percentage of receipt quote
real_sprd_15s Lower is better Executed Shares Here E Average realized spread at 15 seconds
real_sprd_15s_pct Lower is better Executed Shares Here E Realized spread at 15s as percentage
real_sprd_1m Lower is better Executed Shares Here E Average realized spread at 1 minute
real_sprd_1m_pct Lower is better Executed Shares Here E Realized spread at 1min as percentage
speed_1s Higher is better Executed Shares Here E Share of executed volume completed within 1 second

Formulas & Peer Normalization

1. Mix-Adjusted Peer Benchmark

To eliminate symbol-mix bias (where venues trading only mega-caps appear superior), every venue is compared against the volume-weighted performance of competing venues on the identical symbols and size tiers:

Benchmark(V,T)=∑sW(V,s,T) PeerMean−V(s,T)∑sW(V,s,T)\mathrm{Benchmark}(V,T)=\frac{\sum_{s} W(V,s,T)\,\mathrm{PeerMean}_{-V}(s,T)}{\sum_{s} W(V,s,T)}

2. Net Price Improvement (bps of Notional)

Normalized across share prices using the row-level quote midpoint, where P and O are price-improved and outside-quote shares, p̄ and ō their average amounts per share, and Q the quote-relative shares:

NPIbps=104⋅P pˉ−O oˉ∑Q⋅mid\mathrm{NPI}_{\mathrm{bps}}=10^{4}\cdot\frac{P\,\bar{p}-O\,\bar{o}}{\sum Q\cdot\mathit{mid}}

3. PFOF vs Execution Quality Delta

Converted to cents per hundred shares at the venue’s average executed price:

Δcph=(NPIbps−Benchmarkbps)×VenueNotionalVenueShares\Delta_{\mathrm{cph}}=\bigl(\mathrm{NPI}_{\mathrm{bps}}-\mathrm{Benchmark}_{\mathrm{bps}}\bigr)\times\frac{\mathrm{VenueNotional}}{\mathrm{VenueShares}}

Mandatory Disclosures & Caveats

  • A venue's Rule 605 statistics describe all flow it handled, not this firm's orders.
  • 606 route shares are per order, not per share, and list only the top 10 venues plus any venue with at least 5% of non-directed orders.
  • FINRA Rule 5310.09 also requires the firm's own order-level review of its executions.
  • This report is not legal or compliance advice.
  • Net price improvement is in basis points of executed notional (quote midpoint × shares). The PFOF comparison converts the difference to ¢/100 sh at the venue's average executed price; it is an estimate.