In the Capital Market, and in much of the Financial Market, most indicators that somehow analyze entropy seek to do this focused on volatility.
I do not think this approach is wrong at all, since volatility can really help to identify chaotic markets, but the problem is that we need a lot of more information to be more certain if the market will follow a trend, or react in a chaotic way until it returns to the natural entropy values of the financial instrument, for the period of time we are evaluating.
In fact, this seems to me a complex and appropriate problem to seek solutions based on machine intelligence and with other algorithms.
Another relevant point is that probably it will not be possible to identify the cause of the high or low entropy, and this is not the point we seek here, because to reach this level of understanding would require the robot to have a deep understanding of the book, depth, order flow and new facts of the market in real time.
In fact it would be very relevant to be able to build a robot with the planned strategy of chance with all that potential – and I currently work on several fronts in this direction in consulting in the corporate market – but the complexity of simulating this arsenal of data is still very high.
____
Rogerio Figurelli – @ 2018-11-08
