Portfolio of 8 pre-registered falsification studies on retail/ICT crypto & forex signals — summary of what survived (almost nothing) and the one cross-study pattern that keeps repeating
Also: XAG was the worst performer across every single profile and bundle — zero positive cells. Same asset showed a significant opposite-direction effect in an earlier study (forward-test repo).
If this were a real NY-liquidity mechanism, silver should've shown the same sign, just maybe weaker. It didn't. That's a real warning sign, documented before the actual forward test even starts.
Tested whether delta-neutral funding rate carry survives costs on BTC/ETH/SOL perps vs spot (2019-2026, full history, pre-registered before touching data).
Falsified — but not for the boring reason.
Rebalancing mechanism I pre-registered to fix basis drift? Triggered on 0-2.7% of trades. The 1% threshold was too generous — basis risk is real but passive, it just eats you slowly at exit rather than tripping any alarm.
5th falsification repo.