A 42.79% week with a 5.14% hour forces a market maker to shrink its quotes
Should the desk cut quote size and widen spreads through this kind of volatility week, or keep quoting near normal size and absorb the inventory swings that follow?
The market facts are real; the decision maker is illustrative.
The week and the position
In the week of 28 December 2020, ETH rose 42.79%, from 685.10 to 978.28, closing in the upper third of its 48.46% range and 1.34% below its 52-week high. The largest hourly move was 5.14% on Sunday, and Sunday itself gained 26.32%, while the worst day fell only 2.09%. Realised annualised volatility ran at 103.1%, dollar volume hit 7.41bn (2.21 times the prior 12-week average), and perpetual funding averaged 0.0697% per eight hours, peaking at 0.1715%. ETH beat BTC by 17.23 percentage points with loose 0.48 hourly correlation.
Judgment
A desk quoting both sides earns spread but pays for every unit of inventory it cannot offload, and a 5.14% hour can move price past both quotes before the hedging leg is placed. Cut quote size to what can be unwound within one hour at this volatility, widen spreads to cover the 103.1% realised volatility, and skew quotes towards selling since funding of 0.0697% per eight hours is charged to long holders. Accept lower volume capture in exchange for smaller inventory losses.
What would change it
If hourly moves shrink well below 5.14% and funding falls back towards its 0.0256% minimum for several weeks, quote size and tighter spreads can be restored. If funding turns negative, the skew should flip towards holding inventory.
What sets this case apart
The decision is not directional: a desk that quotes both sides still loses when one hour moves 5.14% and funding charges it for the long side of its inventory.