Rate History Explorer

Why percentile beats prediction

Almost every article about timing a currency exchange eventually admits it cannot tell you what rates will do next. That admission is correct, and it is also where most such articles stop being useful. Knowing that nobody can forecast the next month does not help you decide whether to move two thousand pounds today.

What you can establish without predicting anything is position. Every pair has a range it has actually traded in, and today’s rate sits somewhere inside it. If you are near the top of the decade’s range on the side that favours you, you are doing better than most days would have given you. If you are near the bottom, waiting is defensible provided something else is not costing you more while you wait. That is a decision you can make from data rather than from a forecast.

The trap is treating position as a signal about direction. A rate at the top of its range is not about to fall, and one at the bottom is not due a recovery. Ranges persist for years and then break for reasons nobody published in advance. Use the percentile to judge whether you are getting a fair shake today, not to guess tomorrow.

The part that actually costs you money

Volatility numbers put the timing question in proportion. A major pair with annualised volatility around 7 percent will typically move less than one percent in a week, while the gap between a good provider and a bad one on the same transfer routinely exceeds five percent. That means provider choice usually dominates timing by a wide margin, and the hidden cost calculator measures the side worth attacking first.

Timing does dominate in one narrow case: very large, genuinely deferrable transfers, where even a one percent move dwarfs any fee difference. If that describes your situation, use our guide to timing alongside the percentile above, and consider splitting the transfer across several dates rather than betting it all on one.

How this chart is built

Every point is a published daily reference rate, not an interpolated approximation, which means you can verify any single day against the original source. Days on which markets were closed simply do not appear rather than being filled in with invented values, so weekend gaps are normal. See our editorial standards for the full sourcing policy.