data_quote_start / data_quote_end
data_quote_start and data_quote_end are symbol metadata timestamps that indicate the earliest and latest times for which quote (bid/ask) updates are available for a symbol. They define the outer bounds of quote data coverage. If you compute spreads, mid-prices, or quote-based liquidity measures, these fields are the correct coverage references.
Why It Matters
Quotes can have different availability than trades. Some venues publish trades reliably but have limited or later-starting quote capture, and order book depth may start even later. If you only use overall data_start/data_end, you can accidentally include symbols that lack quote coverage during the period you analyze, leading to missing spreads or biased liquidity metrics.
Are quote coverage bounds enough to model execution?
They are necessary but not sufficient. Execution realism depends on quote stability, latency assumptions, and order book depth. Use quote bounds to ensure data exists, then apply additional filters for tradability (spread, update frequency, depth).
How should you treat data_quote_end for currently active symbols?
For active symbols, data_quote_end will typically move forward over time as new quotes arrive. Store the metadata snapshot time if you need reproducible results. For historical studies, treat data_quote_end as the last covered quote timestamp, not as a delisting date.
Which is better for spreads: trades or quotes?
Spreads are defined on quotes (best bid/ask). Trades can validate execution and compute realized spreads, but they do not replace quote coverage. If your metric is spread-based, start with quote coverage fields.
Real-world example
You measure bid-ask spreads in 2020 but quote data for many venues only starts in 2021. If you ignore data_quote_start, you’ll have missing quotes and may incorrectly treat those markets as having no spread or as being inactive, which biases cross-venue comparisons.