How only the closing price of the Nasdaq100 can give so good forecasts?

[jmbonni]

Hello,

I am the developer of a solution (used in marketskipper) for predicting the price at the close of the next session of the Nasdaq 100 and Russell 2000 index. This solution uses only the previous closing prices of the index and an in-house machine learning.

The results are very good and I am wondering how this is possible given the daily volatility and economic news effects ?

Thank you in advance for your ideas and remarks.

Jean Michel

 

[Incster]

There are at least two possibilities:

(1) You have discovered a great new technique to forecast one of the most well-known and profitable market indexes using a tiny fraction of the amount of data that everyone else has needed to use.

(2) You have made one of the hundreds of mistakes that the millions of people before you have made to give a false-positive.

If you do not share more about what you're doing, I don't know how we can help.

 

[jmbonni]

Thank you for your reply. I understand I do not share à lot but anyway the simple fact that knowing only a monovariable serie is used is suffisant to très to draw hypothesis. One of mine is that there are hidden constraints which pull down or up the price. But what should be these constraints?. And I think this is not modelisable mathematically but finely grain probability is possible with machine learning.

 

[jmbonni]

And this means that these indices prices are much more influenced by past and constrains than we think and economics news play a minor role.
Another idea is that the constraints comme from past movements which reflect the anticipations of economics news moves...

 

[jmbonni]

If someone is interested, I just released an app on Google Play to get the the forecast. Search for "marketskipper".
(It is free because I want first to understand)