Who haven’t seen the famous meme-movies about the markets? “The Big Shot”, “The Margin Call” and, of course, “The Wolf of Wall Street” (in which, remarkably, there were no mathematical models involved – just plain old bullshitting). And, when we saw that guy, the second place on a math Olympiad, we probably felt a pang of pain and regret. That could have been me. Math, it turns out, is not that complicated after all....
The only reasonable and even proven correct point of applying deep-learning is that the algorithm “learns” so-called “hidden patterns”, which escape (literally) from a biologically limited human intelligence. The causality is so complex that even to list correctly the major factors (not missing anything relevant) is nearly an impossible task. Social media creates whole waves of bullshit, including the idiotic memes about the US interest rates, and other poorly understood and dynamic macro-economic factors....
These are just assorted notes for now, which shall become something ready to be formalized. Non-bullshit The objective is to train a NN which captures subtle recurrent patterns among many well-chosen (and well-defined) features. The proper set of features that, in turn, captures the most relevant aspects of reality is what determines the distinction between a modest success or a total failure of this ML approach. All the features should be actual “measurements” of something real, like “Open Interest” or the “Long/Short ratio” and other obvious measurements like “Volume”....