Good to hear from you as well, I check your blog quite frequently. The post on MINE is bookmarked for later reading. No, it goes back to norse mythology and history classes in school, but since Hugin Expert is a Danish company I guess they choose their name for the same reason.
Interesting, but quite technical, paper. Haven't read all of it and I'm not sure I fully understand the things I've read, but doesn't he just prove that provided that you know the conditional distributions of returns given a certain event, and these distributions are IID, you can find an optimal trading algorithm? It seems straightforward that this is possible (IF I understand the paper correctly) but doesn't it just change the problem into specifying/finding the conditional distributions?
I don't have the math background yet to fully understand it without sitting down for 10 hours to dissect it - I won't bother commenting since I don't think I'm qualified too (both mathematically and about the financial markets), but I will say that this stuff is pretty out there, as the suboptimal strategy pointed out, there are obvious problems when certain parameters get too small/large
As with any search and like others have pointed out, finding good input values (recipes) is a huge part of it. With initial assumptions you might get bad results. Then you realize you need a lot of computing power to figure out which input values are not noise, or a very fast search application, or both. This is needed to find statistically significant input values, and then it's up to you how statistically significant you want that to be. It might take a lot of computing before you get there so be warned. Creating code to search for stuff isn't that hard, what's hard is having it search long enough for the initial data set which will then be used for various other markets/instruments. Without access to an expensive high-end data centre, your best bet is very fast code.
I don't remember quoting it, I remember reading it and seeing it is one of the popular articles on ssrn, that's why I suggested it. I don't understand all of it as I said before, but enough to get a better than normal gist of it. Also, because I don't actually trade anything right now as I'm learning more about platforms, strategies etc, I wouldn't know enough about the real markets to comment. Could you explain the paper please though
Here are some results from a strategy with a CART filter. This is the original equity curve from the simple lowpass filter strategy: And here's the same strategy with trades filtered by a CART that gets some volatility indicators as input: In this chart, the CART is too aggressive and also filters out some good trades, so the inputs must be more carefully selected. But it's a start.