Every claim we tested, and what killed it

Buy bitcoin from 21:00 to 23:00 UTC, while stock exchanges are shut, every day

Our strongest backtest ever (deflated Sharpe 0.980). Traded forward for real: 40 trades averaging −$7.21, against +$61.62 expected. The backtest still reproduces to the cent, so it was not a data error. A great backtest is not proof.

"CCI mean reversion on gold, London session: $10,000 → $46,519"

Before costs it earns nothing: within 0.08 points of zero in all 15 parameter settings. After costs, −0.42 points a trade at 19 trades a day.

"Dead simple NQ trend strategy: 71% winners, profit factor 3.24"

A sliver of edge before costs (+0.57bp a trade) becomes −0.57bp after them, at 816 trades a year.

"NQ highest-close breakout with an ADX filter keeps performing year after year"

It makes +2.9bp a trade, but less than just holding: Sharpe 0.48 against 1.03. Being long at the same minute on random days, filter on, does as well.

"64.5% of NQ sessions break both sides of the 8–9am range, so fade to the other end"

True, and meaningless. With the bars shuffled into random order, both sides still get broken 69.2% of the time, against 69.4% in real data.

"Automated smart-money strategy: +$34,630, then +$15,370 on a holdout"

+0.004R a trade before costs across 2,797 trades, and −0.039R after them. Positive in 2 of 12 years.

"Lunch-hour range: trade the sweep, enter on the gap, target the midpoint. 289 trades"

Win rate 37.4%. Coin flips with the same stop and target score 38.2%.

"On Wednesdays, when the first-hour low forms first, the high breaks first 78.57% of the time"

By then price has already rallied toward the high. A random walk from the same spot reaches it first 79.7% of the time.

"Opening-range break with an EMA filter and ATR target: 83% win rate over 66 trades"

The reading closest to the post wins 76.7% of trades and still loses before costs, Sharpe −1.72. A coin flip with the same target and stop wins 91.8%.

"Two average stocks, rebalanced daily, turned $100,000 into $7 million"

Across 16,290 stock pairs, daily rebalancing added about 0.8bp a year over simply holding the pair. The Sharpe of that extra: 0.00.