Tag: QQQ

Can We Use Mixture Models to Predict Market Bottoms? (Part 3)

Post Outline * Recap * Webinar Hypothesis * Anaylsis/Conclusions * Jupyter (IPython) Notebook * Github Links and Resources Recap Thus far in the series we've explored the idea of using Gaussian mixture models (GMM) to predict outlier returns. Specifically, we were measuring two things: 1. The accuracy of the strategy implementation in predicting return distributions. 2. The return pattern after an outlier event. During the exploratory phase of this project there were some interestin

READ MORE →

Can We Use Mixture Models to Predict Market Bottoms? (Part 2)

Post Outline * Recap * Model Update * Model Testing * Model Results * Conclusions * Code Recap In the previous post I gave a basic "proof" of concept, where we designed a trading strategy using Sklearn's implementation of Gaussian mixture models. The strategy attempts to predict an asset's return distribution such that returns that fall outside the predicted distribution are considered outliers and likely to mean revert. It showed some promise but had many areas in need of improvement.

READ MORE →

What I'm Watching ( Week of 12.29.14 )

This week it's simple. I'm watching/waiting for volume to reenter the market and confirm or reject last week's holiday action. Last week ( SPY, DIA, QQQ ) all traded to new 52 week highs with my short target ( EWC ) bouncing significantly in sympathy. ( OIL ) price action appears to have moderated a little in terms of downside volatility; buyers and sellers that remained haven't shown conviction one way or the other yet. There is good reason to be skeptical of last week's broad market moves.

READ MORE →