A 31.20% weekly range and a 13.02% day set the inventory limit for a two-sided desk
Should the desk cut its BTC inventory cap and widen its quotes after Friday's 6.78% rebound, or keep tight two-sided quoting at last month's sizing?
The market facts are real; the decision maker is illustrative.
The week and the position
In the week of 18 January 2021, BTC fell 9.95%, from 35,824.99 to 32,259.90, inside a 31.20% range: a 13.02% drop on Thursday, a 4.28% fall in the hour starting 2021-01-22 00:00 UTC, then a 6.78% best day on Friday. Realised volatility ran at 109.0% annualised and dollar volume was 1.43 times the prior 12-week average. Perpetual funding averaged 0.0446% per eight hours and never turned negative. ETH outperformed by 22.93 percentage points. An illustrative desk quotes both sides and keeps inventory small.
Judgment
Size the inventory cap from the 4.28% hour, not the weekly loss: the desk must survive a single hour, not a week. Widen spreads to match 109.0% annualised volatility and the above-average volume, accepting a lower fill rate as the price of not being run over on Thursday's scale. Funding at 0.0446% per eight hours, always positive, pays the short lean, but it is income on inventory the desk first has to carry. Keep quoting both sides at the smaller size.
What would change it
If hourly moves and realised volatility stay near this week's levels for several more weeks, the wider spreads become the permanent cost of doing business. If volatility falls back and funding turns negative, the cap and the quote bias can both be reset.
What sets this case apart
The decision is how a quoting desk sizes inventory against one bad hour inside a volatile week, not on the weekly direction.