WHITE SHADOW TIGERIN MIDNIGHT BLUEPublic screen
Market case · ETH, week of 29 July 2024

A sixth-percentile -17.77% week asks whether the ETH carry signal was a signal

Should the research team suspend live capital on its long ETH signal for four weeks of paper scoring, or keep it running through a week this extreme?

The market facts are real; the decision maker is illustrative.

The week and the position

In the week of 29 July 2024, ETH fell 17.77%, from 3,270.16 to 2,688.92, a return in the 6th percentile of 337 weeks of history. The low of 2,630.00 came late on Sunday 4 August, paired with the week's largest 1-hour move, -3.94%, and the worst day, -7.40%. The close sits in the lower third of the range. Eight-hour funding stayed positive, averaging 0.0077%. Dollar volume of 7.13 billion was about equal to its prior 12-week average. The four prior weeks were -14.75%, +10.72%, +8.96% and -7.52%. An illustrative research team must decide whether its signal survived.

Judgment

Judge the signal against its own premise, not against one week. Funding stayed positive all week, averaging 0.0077% per eight hours, so the signal's entry condition, cheap long-side carry, held while price ran against it; longs were not forced out by funding turning against them. That is one clean observation of the signal failing on price, not on its own logic. Suspend the signal's live capital and keep scoring it on paper for four more weeks. Volume near the 12-week average and ETH trailing BTC by 2.99 points make this an orderly directional loss, so a single sixth-percentile week does not reject a signal with a longer sample.

What would change it

Restore live capital if the following week closes back above 3,270.16, or holds above the 2,630.00 low while eight-hour funding falls to zero or negative and dollar volume exceeds the prior 12-week average. Abandon the signal if two further weeks close in the lower third of their ranges with positive funding.

What sets this case apart

The signal's premise, cheap positive carry, survived an extreme down week intact; the decision is suspension of capital for further scoring, not a forecast of price.

Weekly brief · week of 29 July 2024Week of 29 July 2024: Bitcoin falls 14.78% and Ether 17.77%, both closing near the week lowWhen both assets close the week near their lows after a mixed run of prior weeks, how much of the fall came from single days, and what does that say about the trend?Market case · BTC, week of 29 July 2024A 14.78% week in the worst 4% of 337 weeks tests a short-volatility book's delta hedgeShould the desk buy back part of its short-volatility positions or delta-hedge them after a 58.7% realised-volatility week, or keep collecting premium at the current size?Concept · example: ETH, week of 29 July 2024Gambler's fallacyWhy does expecting a reversal simply because a losing or winning streak has run long often lead traders astray?Concept · example: BTC, week of 29 July 2024Hindsight biasWhy does an outcome that looks obvious in retrospect feel like something that should have been predicted?

Research notes on real market weeks, not investment advice. Sources: Binance spot BTCUSDT and ETHUSDT 1-hour candles and USD-M perpetual funding rates; weekly figures computed by WSTiger.

RISCON Labsrisconlabs.com