Direct answer
Spread by session refers to the idea that the effective spread you experience in forex can vary across different trading “sessions” (for example, based on time of day and typical liquidity patterns). The main risks associated with this concept are interpretation risk (misreading what the session effect implies), market risk (spreads widening when liquidity drops or volatility rises), operational risk (handling and measuring session-based costs incorrectly), and counterparty/execution risk (differences between how spreads are quoted and how orders are filled).
Mechanism or definition
A spread is the difference between the buy (ask) and sell (bid) prices for a currency pair. In practice, what matters to trading cost is the effective cost of execution, which depends on both the displayed spread and how your order is filled.
Spread by session is a way to describe that effective spreads can differ between time windows, because trading activity and liquidity can change by region and hours. When liquidity is higher, spreads often tend to be tighter; when liquidity thins out, spreads can tend to widen. This “session” effect is not a guarantee. It is a pattern that can exist under some conditions and disappear under others.
Operationally, the risk starts when you treat session-based spread differences as a stable rule. If you measure spreads using different cutoffs, time zones, or data sources, your results can shift even if the underlying market is unchanged.
Evidence or example
Consider a simple assumption: you want to estimate the typical execution cost for a pair during two time windows, Window A and Window B. Assume you use the same time zone, the same pair, and the same measurement method for both windows.
Material limitations can still appear:
- Assumption mismatch: If Window A includes transitions (for example, just before or after a major liquidity change), observed spreads may be more volatile than “steady” periods.
- Execution mismatch: Your fills might occur during momentary quotes that differ from the spread snapshot you recorded.
- Provider variation: Two venues (or two quoting methods) can handle order execution differently. Even if the market spread is similar, your realized spread can differ.
A failure mode is using historical session behavior to expect similar costs next month. Historical relationships do not establish future results, especially if volatility regimes change or if liquidity conditions differ from the past.
Limitations and risks
1) Market risk (liquidity and volatility)
Spreads can widen when liquidity drops or volatility rises. That can happen suddenly around economic news, rollovers, or when major market participants are less active. If spread widening occurs, the realized cost can be higher than what session labels alone would suggest.
2) Counterparty and execution risk
Even with the same “session” label, execution can differ due to order type, routing, quote stability, and fill timing. A session-based spread figure can reflect quoted conditions rather than your realized fill.
3) Operational risk (measurement, timing, and assumptions)
Common operational issues include:
- Using inconsistent session time boundaries or time zones.
- Mixing data sources with different timestamp conventions.
- Averaging spreads without accounting for outliers during low-liquidity moments.
4) Interpretation risk (overconfidence in a pattern)
The biggest risk is treating spread by session as a predictable cost schedule. Session effects are conditional on market structure at that time. When conditions change, the pattern may weaken or invert.
Verification or next question
To independently verify what “spread by session” means in your context, focus on definitions and measurement rather than on prediction:
- What time zone and session boundaries are used?
- Are you comparing quoted spreads or realized execution costs?
- How do results change across multiple weeks or different volatility conditions?
- Do your findings hold when you separate steady periods from transitions?
A useful next question is: “Which part of the session effect are you actually observing—liquidity-driven spread changes, execution timing effects, or measurement differences?”