One daily conversion meets a 3.11% hour and a Sunday low set late in the UTC day
Should the payments company split each day's receipts across several fixed conversion windows, or keep converting once a day at its single fixed hour?
The market facts are real; the decision maker is illustrative.
The week and the position
In the week of 10 February 2020, BTC fell 2.31% to close at 9,917.27, in the lower third of an 8.94% weekly range. It gained 3.77% on Tuesday and lost 4.25% on Saturday; the largest hour was 3.11%, mid-day, and the weekly low of 9,638.12 arrived late Sunday, after Thursday's high of 10,500.00. Volume was 1.51 times the prior 12-week average, realised volatility was 50.2% annualised, and perpetual funding averaged 0.0874% per eight hours. An illustrative payments company converts each day's customer receipts into BTC once, at one fixed hour.
Judgment
One fixed-hour batch takes a single price on days whose extremes sit hours apart. Split each day's receipts across two, at most three, fixed windows at the same total daily amount, so no one hour decides the whole day's conversion price. The extra spreads and fees are the price of this, and they are worth paying against a week that contained a 3.77% day, a 4.25% day and a 3.11% hour. Keep the daily total unchanged; only the timing is divided.
What would change it
If realised volatility falls and the largest hourly move drops below the added spread and fee cost of splitting for several weeks, one daily window is enough again. If funding turns negative, the perpetual no longer trades above spot and the basis case for splitting weakens.
What sets this case apart
The choice is operational, not directional: how many times a day to transact, priced against a single hour that moved 3.11% while the week netted only -2.31%.