Search
× Search
Thursday, February 6, 2025

Archived Discussions

Recent member discussions

The Algorithmic Traders' Association prides itself on providing a forum for the publication and dissemination of its members' white papers, research, reflections, works in progress, and other contributions. Please Note that archive searches and some of our members' publications are reserved for members only, so please log in or sign up to gain the most from our members' contributions.

Can We Use Mixture Models to Predict Market Bottoms?

photo

 Brian Christopher, CFA, Quantitative Analyst, Quantitative Developer-Python

 Monday, April 3, 2017

http://blkarb.se/2ormHBH


Print

5 comments on article "Can We Use Mixture Models to Predict Market Bottoms?"

photo

 Valerii Salov, Director, Quant Risk Management at CME Group

 Thursday, April 6, 2017



Johnson SU distribution is continuous. With a continuous distribution, a probability to get a particular price or price increment is zero. And a probability to get a price or price increment within an interval is positive. This contradicts to real pricing observed, for instance, for futures, where a probability to get a particular value is positive and within many intervals is exactly zero. The picture is upside down. Without dismissing the described results, I would like to attract attention to futures prices and two free links "The Wandering of Corn" https://arxiv.org/pdf/1704.01179.pdf April 3, 2017 and "Optimal Trading Strategies as Measures of Market Disequilibrium " https://arxiv.org/pdf/1312.2004.pdf December 6, 2013. The latter also presents a method pp. 43 - 49, 54 - 58 of converting a continuous distribution to a discrete or lattice one under quite general conditions. I would be happy, if these results will be useful within the discussion. Best Regards, Valerii Salov


photo

 Dr. Debashis Dutta, Senior Manager, Financial Service Risk Management, Advisory at EY

 Saturday, April 8, 2017



very interesting Brian. But I am a little confused as I an trying to find why results are too high or too low. Thanks!


photo

 private private,

 Saturday, April 8, 2017



No.. Just to analize /real analysis need more than that


photo

 YUNUS EMRE Ozcan, PHD.c

 Sunday, April 9, 2017



interesting topic , currently I am working on the dirichlet process to model volatility dynamics , its gives a great flexibility for modelling via you can construct distribution over distribution


photo

 Henry Tindal, ETRM Technical Consultant

 Monday, April 10, 2017



You get smelly fingers when you try to pick bottoms

Please login or register to post comments.

TRADING FUTURES AND OPTIONS INVOLVES SUBSTANTIAL RISK OF LOSS AND IS NOT SUITABLE FOR ALL INVESTORS
Terms Of UsePrivacy StatementCopyright 2018 Algorithmic Traders Association