TickerLayer Data Quality

Market data you can verify.

We do not ask you to trust our market data. We benchmark it: price fidelity, freshness and consistency measured against selected market references, per asset class, under a published methodology.

  • Measured from outside, as a customer
  • Published, versioned methodology
  • 124k matched samples across 6 asset classes
New York84 msLos Angeles140 msUS edge137 msSão Paulo170 msLondon21 msMadrid34 msIstanbul41 msMumbai246 msAsia relay154 msHong Kong184 msTokyo256 msJakarta170 msOriginRound trips from the origin · measured Sep 10, 2026

Four questions, four measurements

Fidelity. Freshness. Reproducibility. Consistency.

Every number comes from an identified run. Open one to see how it was measured.

Price fidelity, Forex

0.13bp

Median deviation from the selected multi-source consensus, 20 symbols.

How is this measured?
Definition
Median across symbols of each symbol's median |TickerLayer mid − reference mid| / reference mid × 10,000.
Reference
selected multi-source consensus
Aggregation
Per symbol first, then the median of per-symbol medians so active names cannot dominate.
Window
25 minutes, sampled every 1 s
Universe
fx-core-20 v1, 20 of 20 symbols
Samples
30k matched samples
Method
fx-consensus-v1

Freshness, delivery p95

280ms

Publish timestamp to client receive on the public stream, slowest asset class.

How is this measured?
Definition
Observer receive time minus the ts carried by each WebSocket frame, per frame.
Scope
Delivery path only. Market-event freshness is reported per asset where the reference supports it.

Settled-bar consistency

93.4%

One-minute bars fetched live and again from history later, unchanged, across 502 bars.

How is this measured?
Definition
Latest settled 1m bar via the prev endpoint, then the same interval from the historical range endpoint.
Why
The bar your application saw live must stay explainable when you retrieve that interval later.

Consistency, REST vs WebSocket

97.9%

Same quote on REST and WebSocket across 676 checks in 5 asset classes.

How is this measured?
Definition
One REST quote every ten seconds compared with the stream state around the request.
Tolerance
Exact equality is required for the headline; midpoint difference is published for the rest.

Reference overlay

The difference is measurable.

Two lines that read as one. The strip beneath is the gap, sample by sample.

Stocks: US:GE against the NBBO

Against the selected market references, one sample per second.

94.2%US:GE within 1 bp of the NBBO midpoint
325.00325.50326.00326.50327.00327.50+13.93 bp0 bp-13.93 bp14:2414:3014:3714:4314:49

Full Stocks benchmark

Reference models

Different markets need different truth.

Each asset class uses a reference model suited to its market structure.

Cross-interface consistency

Built to reconcile.

The quote you read over REST is the quote we stream. The bar you saw live is the bar you get back later.

REST vs WebSocket

Share of REST quotes matching a WebSocket frame received within 1.5 s of the request.

  • Stocks99.2%128 checks, median gap 0.00 bp
  • Forex92.6%149 checks, median gap 0.00 bp
  • Crypto100.0%149 checks, median gap 0.00 bp
  • Indices99.2%131 checks, median gap 0.00 bp
  • ETFs99.2%119 checks, median gap 0.00 bp

Live vs historical

Settled one-minute bars fetched live, then again from history at least three minutes later.

  • Stocks97.8%92 bars re-fetched
  • Forex100.0%91 bars re-fetched
  • Crypto100.0%92 bars re-fetched
  • Commodities65.6%90 bars re-fetched
  • Indices100.0%69 bars re-fetched
  • ETFs100.0%68 bars re-fetched

Methodology

The methodology is part of the product.

Universe, reference model, matching rule, aggregation and limitations are written down per asset class and versioned.

Observer
Runs outside the production path with an ordinary API key: one public WebSocket connection plus REST calls. It sees what you see.
Matching
Time-based state matching once per second. Never event N against event N.
Aggregation
Per symbol first, then symbol balanced. Event-weighted figures are published alongside.
Universes
Fixed, versioned lists per asset class. Recognisable names appear because people know them, not because they score well.
Missing data
A figure that cannot be measured is published as not available. Never zero, never a stale number dressed as live.

Benchmark runs

Every number has a run behind it.

We keep measuring, publish each run as it lands, and improve the feed from what the runs show.

Last runSep 10, 2026

Asset classRunMethodWindowSamples
StocksTLQ-STOCKS-20260910-1424global-stocks-quality-v2Sep 10, 2026, 14:24 to 14:49 UTC41k
ForexTLQ-FX-20260910-0845fx-consensus-v1Sep 10, 2026, 08:45 to 09:10 UTC30k
CryptoTLQ-CRYPTO-20260910-0845crypto-consensus-v1Sep 10, 2026, 08:45 to 09:10 UTC22k
CommoditiesTLQ-COMMODITIES-20260910-1018commodities-reference-v1Sep 10, 2026, 10:18 to 10:43 UTC7,420
IndicesTLQ-INDICES-20260910-1018indices-reference-v1Sep 10, 2026, 10:18 to 10:43 UTC8,904
ETFsTLQ-ETFS-20260910-1345etfs-quality-v1Sep 10, 2026, 13:45 to 14:06 UTC14k

Benchmark results describe observed behaviour under the stated methodology during the published window. They are not a guarantee of future performance and not a service level agreement unless separately agreed in writing. These comparisons evaluate TickerLayer quotes against selected market references during the stated observation window. They do not imply exchange affiliation, endorsement, or certification.

See the feed on your own terms.