Consistent edge

I would personally most likely own portfolio consisting of several uncorelated strategies / asset classes each with low Sharpe ratio. In this case I want to be 110% sure that each strategy has some edge and I am not trading noise. It means those strategies have to be tested on long data history and have little degrees of freedom and be nearly unoptimalized. There must be statistical hypothesis testing done on them and I need to have strong convintion there is really premium trading this strategy or investing in it. I do not care so much about performance at the first step. Strategy must have positive expectation of course but equity curve could be bumpy. We could get to better and smoother performance by packing these strategies / asset classes together and applying money management.

Equity asset class is nice example. I can have high confidence that equities have premium over very long term. They are simple beta investment, you can track them very easily, there is no optimalization when you simply track them via index index fund or ETF. They have small Sharpe ratio, but they do have nice performance over long run. And this is type of asset class / strategy I am looking for.

There is lot of strategies like this as I mentioned in my previous post about looking for edge. I personally like to look for them in academic journals and sites like SSRN. The main advantage of this is that articles must meet some standard before they are posted and academics usually pay big attention to statistical testing which is exactly what I am looking for. There is posible to use also sites like http://www.cxoadvisory.com/. This webpage is especially good because each day they are writing short note on one strategy (so I do not need to read whole scientific paper) and they sometimes do double testing (which is handfull too).