macrocephalopod

Only problem is that the price series is one I generated using 100% random noise, it is completely unpredictable by any signal. Congratulations, you now know how to overfit a backtest. Welcome to quant research.

This is Shannon's Demon - you can get meaningfully positive ER by volatility pumping. Sure, it has issues (its basically short gamma) but what you found is "legit" and not an overfit.

No it really is an overfit.

The fact you chose 7 days is an overfit. However, the fact you see positive Sharpe at any horizon is because of Shannon's demon. @breakingthemark

This is further proof to me that the “shannon’s demon” crowd are actually insane.

Shannon’s demon needs +ve arithmetic return but flat or negative geometric return. I generated this data myself so I know it has zero arithmetic return! The fact that you would argue with me about data *that I made up* is simply a confirmation that you are an idiot.

Please don’t become a Shannon’s demon person. The only real-world use is that it tells you you should rebalance your portfolio from time to time, and maybe tilt a bit toward risk parity (completely uncontroversial). Otherwise, completely uninteresting for practical purposes.