Direct answer
Spread by pair is typically most active (showing the largest and fastest changes) during trading-session overlap windows, when more participants are active and liquidity is concentrated in several time zones at the same time. In many markets this overlaps the handoff periods between major centers (for example, when one region is nearing its close while another is opening). However, there is no single universally correct “most active” session for every currency pair, because the pair’s underlying liquidity comes from where that currency is most actively traded.
Mechanism or definition
“Spread by pair” refers to how the quoted bid–ask difference varies for a specific currency pair. The bid–ask spread is affected by the cost of providing liquidity (or hedging risk) and by how readily counterparties are willing to trade at particular moments. Liquidity usually rises when multiple markets are open, because more orders are available and more intermediaries can match trades. When liquidity is higher, quoted spreads often become tighter; when liquidity thins, spreads often widen.
Session overlap matters because it changes the balance between order flow and available liquidity. In simplified terms: more active trading centers generally mean more competing quotes and faster order matching. Still, spread behavior is not only a “time of day” effect. It also depends on stable mechanics like the pair’s usual trading interest and on variable conditions like market volatility, momentary order-book depth, and how a provider prices execution.
A practical assumption for any example is that “activity” means frequent widening/narrowing and larger spread variability, not that spreads always widen during a particular session.
Evidence or example (non-real-time)
You can reason about activity without live data by using overlap logic:
- Choose a major overlap window in your local timezone (for example, when two large trading regions are both open).
- Assume higher liquidity and more frequent quoting compared with periods when only one region is open.
- Expect spread by pair to show more movement during that overlap, especially for pairs that draw significant trading interest from both regions.
As a concrete illustration of how to think, consider two pairs with different dependency on global participation: one pair may be influenced strongly by trading centers in North America and Europe, while another may be influenced differently by Asia-centered flows. Even if both pairs are traded all day, their “most active” windows can shift because their liquidity is sourced from different participants and venues.
This also connects to execution: even with the same underlying market conditions, providers can apply different cost components (for example, fixed commissions plus variable spreads, or internal matching practices). Therefore, a chart of “spread by pair” from one venue may not match another venue’s pattern during the same hours.
Limitations and risks
A material limitation is that “most active” is not a stable property across time. Failure modes include:
- Provider or execution effects: A provider’s quoting and execution model can create spread changes that look like session effects but are actually cost or operational behavior.
- Volatility regimes: During high volatility, spreads can widen even in normally liquid overlap periods, making time-of-day expectations unreliable.
- Thin-liquidity moments: Overnight or holiday periods can produce abrupt changes because available counterparties are fewer, even if session overlap would suggest otherwise.
Because outcomes vary with market conditions, costs, and execution quality, historical overlap patterns do not guarantee future results. Also, without real-time observations, any ranking of sessions remains an assumption-based explanation rather than a verified measurement.
Verification or next question
To independently verify which sessions are most active for a chosen currency pair, use consistent assumptions: record spreads and timestamps from the same data source, document the cost structure you include (spread only versus spread plus commissions), and compare overlap windows versus non-overlap windows. If you see differences, check whether they align with liquidity expectations or with changes in provider pricing and execution.
If you want, name the currency pair you care about and the data source/timezone you use, and I can help you set up a verification approach that distinguishes session-overlap effects from provider-specific pricing behavior.