A good strategy must outperform in many uncorrelated assets?

Quantitative Model

Hello, I've been backtesting many different ideas and i have been talking with a friend about this topic and we conclude that there's no a general outperformer.

But the question is whether a strategy must outperform different assets to be considered a good one? 

If a strategy performs poorly in one asset but very well in others, it would be worthwhile to investigate more. But you should also evaluate if it was due to pure luck or a genuine edge in your strategy.

There's no one-size-fits-all answer to whether a strategy must outperform different assets to be considered good. It depends on your own situation, investment goals, and risk tolerance. A well-planned strategy should align with your goals and be flexible enough to adjust to changing market conditions so you can make informed decisions.
Henry Tsai
No, a "good" strategy is one that generates a positive risk-adjusted return, exceeding the benchmark of your choice, for the same period of time and instruments that you choose to run on.

Keep in mind that 
  • There is no signal rule to determine "good" or "bad".
  • There is no God strategy that generates profit all the time on all asset classes.

If you look at many professional systematic traders, they'll all have a broad basket of strategies that work for different market regime. If you develop a strategy and it looks like it used to work, keep it on file. Maybe it'll work again at a future time.