A once-a-day receipt conversion met a 15.64% down day and an 8.07% hour
Should the payments company keep converting each day's receipts in a single batch at one fixed time, or split the batch into tranches through the day to shorten exposure to any single hour?
The market facts are real; the decision maker is illustrative.
The week and the position
In the week of 7 January 2019, ETH fell 25.82%, from 153.89 to 114.16, inside a 38.98% range, closing in the lower third of that range. Thursday 10 January was the worst day at -15.64%, and the largest one-hour move was -8.07% starting 06:00 UTC, early in the UTC day; the low came late in the day on Sunday. Volume was 1.1 times the prior 12-week average. ETH lagged BTC by 13.01 percentage points, and the ETH/BTC ratio fell 14.92% at a 0.9 hourly correlation. An illustrative payments company converts customer receipts into the asset once a day.
Judgment
A single daily batch leaves every receipt converted in one hour's price, and this week that hour moved 8.07% within a day that moved 15.64%. Split the conversion into tranches spread across the day and shorten the maximum wait between receipt and conversion, sized so no tranche carries more than the hourly move this week measured. Accept the higher transaction cost and a slightly worse average price than a well-timed single batch. Do not move the batch to a favoured time of day: the week's high came early and its low late, which gives no repeatable window.
What would change it
If later weeks show hourly moves persistently far below -8.07% and one-hour spread costs under tranching exceed the slippage the batch absorbed this week, the single daily conversion can be restored. A stable ETH/BTC ratio would also weaken the case for shorter intervals.
What sets this case apart
The issue is the length of the conversion window inside a violently trending week, not the direction of the asset or the choice of conversion price.