# Tracking One Pair for Thirty Days: What the Spread Did

> A month of daily captures on a single pair, and the three patterns that survived the noise.

- Source: https://swapcostnotes.com/tracking-one-pair-thirty-days/
- Published: 2026-10-07
- Author: Lena Virtanen
- Publication: Swap Cost Notes — What the conversion actually took
- Topic: exchange
- Tags: exchange, method, spread

---

Thirty daily captures on one pair at one venue. Small sample, single venue, no claim beyond it. Three patterns were visible anyway.

## Method

Same hour each day, same size, quote captured alongside a reference mid from a fixed source. Implied spread calculated as the difference.

Weekends included, which turned out to matter.

## Pattern one: a weekday floor

On ordinary weekdays the implied spread clustered in a narrow band. The variation within that band was smaller than I expected and looked like a pricing policy rather than a market outcome.

## Pattern two: weekends sat above it

Every weekend capture sat above the weekday band. Not dramatically, and consistently.

Four weekends is four observations, so this is suggestive rather than established. It matches what would be expected from thinner conditions.

## Pattern three: the tail was event-driven

Three captures sat far above everything else. All three coincided with visible price movement on the day.

Those three contributed more to the monthly average than the other twenty seven combined, which is the usual shape for this kind of measurement.

## What the average concealed

A mean spread across thirty days is a useless number for planning, because the distribution is not symmetric. The median is a better description of an ordinary day and the maximum is a better description of the day you will remember.

Quote both or neither.

## What I could not measure

Whether the quote held to execution, since these were captures rather than trades. A venue can quote tightly and execute poorly, and this method cannot detect that.

That requires paired quote and execution records, which is a different exercise.

## What a reader should take from it

The method, not the numbers. Thirty captures on your own pair at your own venue will tell you more than any published comparison, and it costs nothing but a minute a day.

[Venues that quote the swap rate and the fee as two numbers](https://collectnexchange.cy/services/swap) make the capture trivial. [A platform where the spread is visible before the trade](https://collectnexchange.cy/services/crypto-exchange) removes the need to calculate it, and [Collect & Exchange](https://collectnexchange.cy/) publishes a schedule that can be compared against what the captures show.

---

This article explains how something works. It is not financial advice.
Canonical HTML version: https://swapcostnotes.com/tracking-one-pair-thirty-days/