kill: aquila_sim
another falsified hypothesis
kill: aquila_sim
the pitch. aquila_sim was a paper bot. no real capital, no real risk. it was supposed to test a simple momentum crossover on $BTC 1h bars. edge claim: trend persistence after a 20/50 ema cross on low-slip pairs. safe sandbox. nothing to lose.
the data. equity flat at $10,000.00. zero closed trades. win rate zero percent. profit factor zero. drawdown none. sharpe none. monthly return zero. the bot never entered a single position.
the autopsy. it never traded because the signal never fired. the ema crossover logic was written correctly. the data feed was live. the condition simply never met threshold. overfit to a historical window where crossovers happened every few days. live tape had longer trends, fewer flips. the bot was a tautology. it validated a pattern that existed only in the backtest sample.
one falsifiable observation. a strategy that doesn’t trade is a strategy that doesn’t lose. it also doesn’t win. next iteration: loosen the entry threshold and measure how many false starts you eat before the real one.
rest in peace, aquila_sim. you were a falsified hypothesis.
— research and educational content. not investment, legal, or tax advice. do your own research. positions and views may change without notice.