Direct answer: what “rollover for pair correlation” actually means
Rollover (also called swap) is the interest-related cost or credit applied when a position is held across a roll-over time. “Pair correlation” usually refers to how two currency-pair prices move together statistically. In most explanations, correlation is not part of the rollover math; instead, correlation is relevant because it can change how the combined set of positions behaves, which can affect your net exposure to costs and returns.
So the key idea is separation:
- Rollover mechanics come from interest-rate differentials and instrument conventions.
- Pair correlation comes from observed price co-movement.
- Their link is indirect: correlation influences what positions you hold together and how they move, while rollover is charged according to the currencies and timing of each position.
Mechanism or definition: from interest inputs to a daily rollover figure
A simple, evergreen way to understand rollover for a forex pair is:
- Identify the two currencies in the pair (e.g., base vs quote currency).
- Determine the interest-rate differential between them (in practice, this is often mapped to benchmarks such as central-bank rates or money-market expectations).
- Convert that differential into a daily rate and then into a cash amount for your trade size.
Because the exact implementation differs by provider, you usually need to rely on the provider’s swap/rollover specification. Still, the conceptual calculation is commonly modeled as:
- Swap amount per day ≈ (interest differential) × (position notional) × (instrument scaling factors)
- The sign (cost vs credit) depends on whether you are effectively long the higher-yield currency or short it.
Two practical conventions often matter:
- Daily rollover vs trade hold period: the swap applies per day you carry the position past the rollover cutoff.
- Triple swap on certain days: some platforms charge a larger swap amount on specific days (often around the weekend) because there can be an extra effective day of interest.
Evidence or example: checking the inputs and the “triple-swap” step
Since no live prices or provider tables are provided here, consider a worked template with explicit assumptions.
Assumptions (made for illustration only):
- You hold a position that is subject to a rollover time.
- The provider’s rollover is defined as a swap per standard lot per day, with a larger multiplier on a particular weekday (a “triple swap” convention).
- Your position size is 1 lot (or you use the lot-to-cash conversion your provider states).
Template calculation:
- Start with the provider’s base swap for the pair (the daily swap amount for your trade direction).
- Count the number of rollover days you actually held the position past the cutoff.
- If one of those rollovers falls on the provider’s special day, multiply that day’s swap by the provider’s convention (often a factor such as 3).
- Add them up to get the total rollover cost/credit for the holding period.
Where correlation comes in (indirectly): if you also hold another currency pair with similar dynamics, correlation affects whether both positions remain open and whether you may experience similar price moves at the same time. That can change your overall net outcome, but it does not replace the step where rollover is computed from the two currencies and the swap rules.
Limitations and risks: where this understanding can fail
- Provider-specific swap rules: Even if you understand interest-rate differentials, the exact mapping to a charged/credited swap amount can vary by platform and instrument. Without that document, any numeric calculation is uncertain.
- Timing and execution: Rollover is sensitive to the rollover cutoff and the exact time your position remains open. Small timing differences can change whether a day’s swap is applied.
- Triple-swap exceptions: The “triple swap” convention is not universal in wording or timing; providers may implement different multipliers or schedules. Assuming the wrong convention can materially change total swap.
- Correlation is not a pricing identity: Pair correlation is usually estimated from historical data and can change as market conditions, liquidity, and spreads evolve. It describes co-movement, not the underlying interest-rate differential.