A 16.86% week at the 95th percentile turns two-sided quotes into one-way inventory
Should the desk keep tight two-sided quotes at full size through a week trending this hard, or cut size and widen spreads until intraday direction stops running one way?
The market facts are real; the decision maker is illustrative.
The week and the position
In the week of 4 November 2024, BTC rose 16.86%, from 68,775.99 to 80,370.01, the 95th percentile return against 351 weeks of history, with volume of 24.43bn, 2.27 times its prior 12-week average. The best day was Wednesday 6 November at 8.94%, and the largest one-hour move was 2.74% starting 02:00 UTC, early in the day. The low preceded the high, and the close sat in the upper third of the range. Funding averaged 0.0109% per eight hours. An illustrative desk quotes both sides and holds small inventory. ETH gained 12.66 percentage points more than BTC, and hourly correlation was loose at 0.68.
Judgment
A week this directional punishes the side of the book that keeps buying as price runs. Cut quote size, widen spreads, and skew both quotes against the prevailing direction rather than trusting round-trip fills. Hedge residual inventory with ETH only in part: correlation of 0.68 is loose and ETH moved 12.66 points more, so the hedge leaves basis risk. Accept lower capture in a high-volume week to avoid carrying inventory overnight against the trend, which is the loss the desk cannot unwind quickly.
What would change it
If the largest hourly moves settle back well below the 2.74% of Wednesday and the best and worst days alternate direction across two consecutive weeks, the desk can restore size and tighten spreads. If hourly correlation rises above 0.7 and the ETH-BTC gap narrows, the hedge can carry more of the residual inventory.
What sets this case apart
This case is about adverse selection inside a trending week: the desk's loss is inventory it cannot sell, not the direction call itself, and the hedge only partly covers it.