Is martin so bad? Or do you have to know how to cook it?

Aleksey Mavrin 2020.01.09 08:03 #621

Edgar Akhmadeev:
That's what I'm saying - as the quality of EA trading improves the need for MM like Martin decreases. Probably doesn't get to zero though. I haven't found out yet. Work in progress.

Let's be clear about what we mean. Whether or not a strategy involves raising the stakes to diversify risk and increase IR is considered a Martin. I think not, because Martin is a strategy that essentially involves nothing but raising the bet.

Anti-Martingale vs Martingale : Avalanche Expert's idea of a

Dmitiry Ananiev 2020.01.09 12:29 #622

And you disable one of the inputs and look at the result.

Why is martin bad: During optimization parameters are selected so as to pass all the "bad" parts of the history and not to fail. And even a test on a long history does not guarantee that there will not be a series of losses.

Safe martingale. Damned Martin Machine learning in trading:

Edgar Akhmadeev 2020.01.09 15:19 #623

Aleksey Mavrin:

Martin is a strategy that essentially involves nothing more than raising the stakes.

The theoretical publications write that any improvement in the strategy's 50/50 ratio(random entry) greatly improves the Martin.

But whether Martin improves the strategy is the question I'm exploring in my lab assignment.

Is martin so bad? Or do you have to know how to cook it? - page 63