Week of 8 May 2023: Bitcoin falls 5.32% on a quarter of its usual dollar volume
When both assets fall but trading volume runs far below normal, what can the price path say about the decline, and what can thin trading not tell us?
BTC fell from 28,430.09 to 26,917.62 (-5.32%) and ETH from 1,870.41 to 1,799.21 (-3.81%). BTC dollar volume was 0.27 times its prior 12-week average. BTC funding averaged 0.0034% per eight hours, so longs paid shorts only slightly.
What the data shows. BTC printed its weekly high of 28,631.01 in the first hour of Monday 8 May, then drifted down to a low of 25,811.46 at 19:00 UTC on Friday 12 May, a range of 10.92%. The close sat at 39.2% of that range, in the middle third. The worst day was Monday (-2.68%); the best was Sunday 14 May (+0.53%). The largest hourly move was -2.43%, starting 17:00 UTC on Wednesday 10 May, in the same hour ETH posted its own largest move (-2.27%). The two assets moved closely together (hourly correlation 0.92). Both finished about 16% below their 52-week highs. BTC's weekly return ranked at the 21st percentile of 273 weeks; ETH's at the 30th.
What it does not show. Volume was the notable signal: BTC traded 9.25 billion dollars, 0.27 times its prior twelve-week average, ETH 5.63 billion at 0.83 times. Thin trading can make a move look larger or smaller than it would in a crowded market, and price candles say nothing about who traded or why. Funding near zero (average 0.0034%, high 0.0069%) tells us neither side paid much to hold; it does not tell us how large the positions behind it were.
What a careful desk would check. Realised volatility, the yearly-scaled size of the swings that actually happened, was 41.6% for both assets, and the largest hourly move was -2.43%. A desk would compare its stops and margin buffers against that hour rather than the average week, and would treat executions on 0.27-times volume as less reliable than the price suggests. The prior four weeks alternated for BTC (6.99%, -8.96%, 5.95%, -2.75%), so this decline extends an uneven pattern rather than a clean trend.
Ether against Bitcoin. ETH fell less than BTC, so the ETH/BTC ratio rose 1.60%. The gap of 1.51 percentage points is modest, and with correlation at 0.92 this reads as one broad move, not one asset separating from the other.
What would change this reading
- If BTC volume returns towards its prior twelve-week average in later weeks, this week's 5.32% fall on 0.27-times volume reads as a thin-market move rather than a change in trend.
- If a later week trades below Friday's low of 25,811.46, the middle-third close at 39.2% was a pause inside a decline.
- If the ETH/BTC ratio keeps rising while both assets fall, the modest 1.60% divergence is becoming a real rotation, which changes which hedge matters.
- If funding turns clearly negative from its near-zero 0.0034% average, short positions would begin paying long ones, a different posture from this week's balance.
Terms
- realised volatility
- how much the price actually swung over a period, expressed as a yearly percentage
- hourly correlation
- how closely two prices moved together hour by hour, from -1 to 1
- ETH/BTC ratio
- the price of one Ether expressed in Bitcoin, showing which of the two outperformed