A +7.37% week that also lost 4.68 points to BTC: did the signal earn its return?
Should the team mark its ETH signal as validated and move it towards production, or keep it in research because this week's gain may be market beta rather than signal skill?
The market facts are real; the decision maker is illustrative.
The week and the position
In the week of 21 December 2020, ETH rose 7.37%, from 638.10 to 685.11, closing 80.7% of the way up its range and 2.95% below its 52-week high. The path was violent: a 30.08% range, a -7.99% day on Wednesday (low 551.28) and a +7.48% day on Sunday (high 717.13), with realised volatility at 104.4% annualised and a 5.62% hour. Volume ran 1.52x the prior 12-week average. ETH still lost 4.18% against BTC, so its return trailed the market it trades in.
Judgment
A profitable week on a 30.08% range and 104.4% realised volatility proves little. The return was small relative to the noise, and ETH lost 4.68 points to BTC, so most of the week's gain came from the broad market rather than from the signal's ETH choice. Keep the signal in research, not production, and judge it only on returns measured against a BTC-matched benchmark across multiple weeks. Accept that this week's evidence is discarded and no size is added.
What would change it
If the signal shows positive ETH-minus-BTC returns across several subsequent weeks, or its daily positions were on the right side of both Wednesday's -7.99% and Sunday's +7.48% rather than merely long, the validation case holds and the move to production can proceed.
What sets this case apart
This case tests attribution, not performance: a positive absolute week in which the asset underperformed its benchmark cannot validate a selection signal.