A relative-value signal gave back 13.57% against BTC right after a 29.52% rally
Should the team keep its long-ETH-short-BTC signal live at reduced size after one 13.57% losing week, or retire it and wait for a new signal?
The market facts are real; the decision maker is illustrative.
The week and the position
In the week of 11 November 2024, ETH fell 3.37% to close at 3,076.00 while the ETH/BTC ratio dropped 13.57%, so BTC rose and moved more. The week before, ETH had gained 29.52%. Dollar volume hit 15.66bn, 2.87 times the prior 12-week average, and ETH closed in the lower third of a 14.20% range, 24.38% below its 52-week high. Funding stayed positive, averaging 0.0187% per eight hours. An illustrative research team runs a signal long ETH and short BTC and must judge it now.
Judgment
A 13.57% ratio loss on volume 2.87 times the prior 12-week average is real evidence against the signal, but not a clean falsification: the prior week added 29.52%, the week's range was 14.20%, and ETH finished in the lower third, so one week after a spike is a thin sample from 352 weeks of history. Keep the signal, cut its size to what a fresh, untested idea would carry, and record this week as one observation rather than a verdict. The team accepts missing upside if the signal resumes.
What would change it
If ETH closes a following week below the 3,014.50 low of this week, or the ratio loses further ground while BTC keeps rising, retire the signal. If ETH closes back above the 3,442.50 high, restore the original size.
What sets this case apart
The verdict turns on the 13.57% ETH/BTC ratio move, not on ETH's own modest -3.37% week, which hides the relative underperformance.