Each study fixes its rules before it runs, compares against random luck and the simplest alternative, and keeps its last years unseen. Most ideas fail. We publish those too.
Hedge half of BTC, ETH, SOL, BNB and XRP when traders are crowded, funding is extreme or leverage surges, and compare with simply holding less.
No. The two deepest falls of six years, BTC -77% in 2021-22 and -53% in 2025-26, started without a crowding signal. Crowding caught only the quick leverage flushes. A 200-day trend filter helped in long bear markets but gave it back in choppy ones. Holding a bit less all the time did as well as any signal.
Large deposits to exchange wallets are often read as holders getting ready to sell.
The flows followed the price instead of leading it. Big deposit days marked moves that had already happened. Tokens with the heaviest deposits lagged slightly over the following week, too little to trade on its own.
104 exit rules (targets, stops, break-even moves, trailing stops, partial exits) on the same 48,626 entries, chosen on 2022-24 and judged on 2024-26.
No. Every fixed exit won about as often as it needed to break even, and random entries with the same exits did the same. Exits change how results look, not what they add up to.
Enter with a limit order where price broke a level, target +10%, stop -30%, on every Binance perpetual including delisted ones.
It won 70% of trades, exactly what +10% / -30% needs to break even. Random hours and random coins with the same exits did as well. One thing held up: waiting for the retest beat buying the breakout at market by about 3 points per trade. It avoids a bad entry; it does not create a good one.
The same desk rules on the ten largest coins, with 30 different target and stop pairs.
All 30 lost to simply holding the same coins. Big caps trended up, and the rules sold them early.
Every test is written down and timestamped before it runs. Changes after a run are listed in the report, never folded in quietly.
Each result is compared with a thousand random versions of itself. If luck does as well, the idea gets no credit.
Holding, or holding less, is always in the comparison. A strategy has to beat doing almost nothing.