GemloomResearch Log No. 5Topic: On-chain flows
On-chain capital is only a clue: the boundary between today's whale transfers and review evidence
Today's scan covered 516 Binance USDT spot trading pairs, of which 268 had market caps within the research scope, and only 1 passed all hard filters; in the past 24 hours, the review council sent 3 to review and approved 2. Among on-chain large transfers, there were 12 whale-to-whale transactions totaling USD 1.54 million across 6 tokens. We treat on-chain anomalies only as bonus points and review evidence, not as buy signals. There is only 1 closed signal, and the sample is too small to draw conclusions.
Key points- The sample of on-chain large transfers is limited and cannot be used alone as a directional judgment; it serves only as one piece of review evidence.
- The coin selection funnel was extremely strict today: only 1 of 516 trading pairs passed the hard filters, and the review council approved 2 signals.
- The closed trade track record has only 1 sample, so the win rate has limited reference value and needs further accumulation.
Today's funnel: from 516 to 1 passing hard filters
The latest full-market scan is a snapshot covering 516 Binance USDT spot trading pairs, of which 268 had market caps within the research scope, and only 1 passed all hard filters. Reasons for failing the hard filters included: 209 with market caps outside the research scope, 168 with risk tags, 38 not yet in a bottom zone, 80 with insufficient or unstable liquidity, and 2 with too short a listing period. Figure 1 shows this funnel.
Strong sectors were bStocks, Payments, Storage, and HODLer. In the potential board top 10, POLYX was listed with a potential score of 55, a 75% decline below its peak, and sectors of Public Chain/L2 and Infrastructure; among the sub-scores, cross-exchange anomalies scored 2.0. The potential board is scored by a rules engine, not selected by AI.
On-chain large transfers: 12 transactions, USD 1.54 million, what can they tell us
In the past 24 hours, on-chain large transfer statistics only had the 'whale-to-whale' category, with a total of 12 transactions, an amount of USD 1.54 million, involving 6 tokens. This sample is not large, and there are no exchange net inflow or net outflow data, so we cannot directly judge whether chips are being accumulated or distributed.
The theme note already pointed out: exchange net outflows are often seen as accumulation signs, while transfers into exchanges may mean selling pressure, but they could also just be internal exchange transfers. For this reason, we treat on-chain anomalies only as bonus points and review evidence, not as buy signals. Today's 12 transactions are not large in amount and are scattered across tokens, so no directional conclusion can be drawn for now.
Review council: the stances of 4 reviewers and approved signals
In the past 24 hours, the review council sent 3 to review and approved 2, with 0 red-flag vetoes and 1 with insufficient approval votes. The two approved signals were GALA and YGG, both with 4 buy votes, and all 4 reviewers gave buy votes.
Figure 2 shows the stance distribution of the four reviewers: Fundamentals analyst 3 buy votes; Skeptic 2 buy votes and 1 abstain; Technicals 3 buy votes; Risk officer 2 buy votes and 1 wait-and-see. The two approved signals received buy votes from all reviewers, while the case that did not pass was due to insufficient approval votes. This shows that even when the candidate pool is very small, AI review still maintains multi-role checks and balances.
The track record sample is still too small: what it means for users
In the public track record, there is 1 closed signal, a win rate of 100.0%, an average of 9.89% per trade, a median of 9.89%, a maximum drawdown of 0.0%, an average holding period of 4.4 days, 13 signals currently held, and 0 closed and 0 profitable in the last 7 days. The exit reason was 'gain N% take-profit N/N' 1 time.
There is only 1 closed signal, and the sample is too small to conclude that the strategy is stable; it can only be used as an early reference. The on-chain capital evidence currently consists only of whale-to-whale transfers, with no net outflow or net inflow direction, so it cannot be used as a reason to buy. We insist on putting on-chain anomalies into the review evidence chain, rather than using them as a switch to trigger trades.
