A 30.29% range and 88% realised vol force a quoting desk to reprice its own inventory
Should the market-making desk keep quoting both sides at its current width and size, or widen spreads and cut quote sizes until hourly volatility falls back?
The market facts are real; the decision maker is illustrative.
The week and the position
In the week of 21 December 2020, BTC rose 12.05% to close 26,281.66, 5.54% below its 52-week high, inside a 30.29% range from 21,815.00 (Monday 12:00 UTC) to 28,422.00 (Sunday 11:00 UTC). Realised annualised volatility was 88.0%; dollar volume was 16.9bn, 2.12 times the prior 12-week average. Monday was the worst day (-3.14%) and carried the largest one-hour fall (-4.33% at 10:00 UTC); Saturday the best day (+7.21%). ETH lagged BTC by 4.68 points. The desk quotes both sides and must hold small inventory, at widths set in quieter weeks.
Judgment
A volume 2.12 times the average at 88% realised volatility means a quote at the old width is stale between fills, as the Monday 10:00 UTC hour shows. Keep quoting both sides, but widen spreads, cut per-quote size, and cap inventory at what the desk can flatten inside one hour, since a single hour already moved -4.33%. Funding of 0.0205% per eight hours, even at its 0.0712% maximum, is too thin a cushion against a 30.29% range. Hedges with ETH are weak, since the correlation is loose at 0.68 and ETH lagged 4.68 points. Postpone raising size.
What would change it
If later weeks show largest one-hour moves well inside -4.33% and realised volatility well below 88.0%, with fills that do not turn against the desk before it reprices, the old width and size can be restored.
What sets this case apart
The desk's risk is stale quotes and inventory speed, not direction: the week rose 12.05%, yet the loss channel is a Monday hour inside an up week.