Direct answer
Spread by session can change during volatile markets mainly because the “spread” you see is produced by how quotes are formed and how orders are executed at that moment. Volatility often reduces available liquidity and increases sudden shifts in prices, which can widen the effective spread for some sessions, especially near session start/transition times.
Mechanics and definition
“Spread by session” means the spread (the difference between bid and ask) is not a single fixed number across all market hours, but can vary by predefined time windows (sessions). Even when a provider labels spreads by session, the final realized cost for an order depends on several layers:
- Market liquidity at that time. When fewer counterparties are willing to trade, the bid/ask gap typically grows.
- Latency and quote freshness. Quotes can lag behind fast moves. If the market moves between when a quote is generated and when an order is matched, the effective spread can differ.
- Order handling rules. Providers may route, queue, or partially fill orders differently across sessions. If matching happens in chunks, each chunk can face a different immediate bid/ask.
- Price gaps and jump risk. During volatility, the market may move in steps rather than smoothly, creating discontinuities where spreads can jump.
Evidence or example (with stated assumptions)
Assume a session schedule defines a “base” spread for a time window. In calm conditions, liquidity is steady, quotes refresh quickly, and orders are typically matched close to the displayed bid/ask. During volatile conditions, suppose liquidity thins and price updates become sporadic:
- A user places an order at the end of one session and expects the next session’s spread to apply immediately.
- Due to timing, the order may be matched using quotes produced just before the boundary, or using the closest available liquidity.
- If the market has moved sharply during that interval, the bid/ask gap at matching time can be larger than the baseline implied by the session label.
This can happen even without any “mistake” by the user. The key point is that session-based spread labeling is an input, while execution is a time-dependent process.
Limitations and risks (material failure modes)
- Quoted vs realized spread mismatch. The spread displayed at one instant may not equal the spread actually paid when the order is filled.
- Gaps at session transitions. Session boundaries can coincide with lower liquidity or delayed quote availability, increasing jump and widening risk.
- Liquidity withdrawal. If counterparties step back during volatility, order matching may rely on less favorable immediate prices.
- Uncertain timing across systems. Network delay, server processing, and platform behavior can change what part of the quote stream your order actually interacts with.
Verification and what you can check independently
To verify how spread by session affects realized costs, use a method that does not assume future behavior:
- Compare session windows over multiple volatility regimes (quiet vs volatile periods) using your own historical execution records.
- Track timestamps, order lifecycle, and fill details (when orders were sent and when fills occurred) to see whether spreads widen around boundaries.
- Separate displayed quotes from fills. If your platform provides both, measure the difference between quoted spread at submission time and the effective cost implied by fills.
If you want, share your exact definition of “Spread By Session” from your platform or contract terms (text only), and I can help you map the timing and order-handling concepts to that wording—without making promises about outcomes.