Week of 21 August 2023: a 6% swing that ends flat on volume 33% below normal
After a week that fell 10.63%, a week that swings 6% but closes almost unchanged can mean several things — what does the price data support, and what does it leave open?
BTC moved 6.01% from low to high yet closed -0.34% at 26,101.77; ETH ranged 7.73% and closed -1.60% at 1,658.33. Both assets bottomed at 21:00 UTC on Tuesday 22 August and peaked Wednesday 19:00 UTC. Dollar volume ran at 0.67 times the prior twelve-week average for BTC.
What the data shows. Both assets followed the same path. BTC fell to 25,300 in the hour starting 21:00 UTC on Tuesday 22 August, then rose to 26,819.27 by Wednesday 19:00 UTC and closed the week at 26,101.77. ETH set its low of 1,580 and its high of 1,702.11 at the same two hours, closing at 1,658.33. The two moved closely hour by hour (correlation 0.90), but ETH fell further, so the ETH/BTC ratio dropped 1.26%. The close sat near the middle of the weekly range for BTC (52.8%) and somewhat higher for ETH (64.1%). Volume was thin: 4.92 billion dollars for BTC, 0.67 times its prior twelve-week average, and 2.75 billion for ETH at 0.78 times. Both weekly returns ranked mid-pack against 288 weeks of history (44th percentile for BTC, 39th for ETH). Realised volatility, the size of the swings that actually happened scaled to a yearly figure, was 26.2% for BTC and 33.2% for ETH.
What it does not show. A flat close after a wide swing does not tell us whether sellers dried up or buyers simply stayed away. Volume running a third below its recent average says only that fewer dollars changed hands, not why. Nothing here identifies who traded at Tuesday's low or whether Wednesday's high marked any change in direction. The prior four weeks included a -10.63% BTC week, so this week sits inside a recent decline, but the data cannot say whether that decline continues.
What a careful desk would check. The largest hourly move of the week differed by asset: ETH fell 1.66% in the hour starting 21:00 UTC on Tuesday 22 August, the same hour its low was set, while BTC's largest hourly move, -1.03%, came later, in the hour starting 21:00 UTC on Wednesday 23 August. A desk would compare its stop distances with those hours rather than with the yearly volatility figure, and would ask whether thin volume (0.67 times average) makes this week's range a reliable guide to liquidity at all. A range set on light volume can narrow quickly once volume returns.
Funding. BTC perpetual funding averaged +0.0014% per eight hours, with a high of 0.0053% and a low of -0.0028%; ETH averaged -0.0009%, with a low of -0.0066%. These are close to zero, so neither longs nor shorts paid a meaningful standing cost on average across the week's eight-hour funding periods.
What would change this reading
- If a later week trades below Tuesday's low of 25,300 for BTC, this week's range reads as a pause inside the prior -10.63% decline rather than a base.
- If volume rises back towards the prior twelve-week average while price holds above 25,300, the flat close gains weight as evidence of absorption.
- If funding turns clearly negative while price holds, short positions are paying longs and the balance of pressure may be shifting.
Terms
- close location
- where the weekly close sits between the week's low and high, as a percentage; a higher figure means the close finished nearer the top of the range
- realised volatility
- how much the price actually swung over a period, expressed as a yearly percentage