You are viewing a single comment's thread from:
RE: More Playing with BTC/USD Data - Quantitatively
I don't completely follow, but, what you're doing is interesting. It also tracks what I would have liked to have done if I were about 20 years younger.
Thanks, it;s not that complicated once you get used to it. In this article I just described that I have updated my benchmarking system. Many statisticians compare different forecasts together which is like comparing apples to oranges, in my my method I compare a model to it's random self, which is self -contained and it's provably unbiased. If the model is better then it's random self then, its obviously good, hinting towards an edge, possibly a profitable forecasting model.