Search
× Search
Monday, December 23, 2024

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.

A Challenge to Improve Your Trading System Using a Meta-Strategy

photo

 Jonathan Kinlay, Quantitative Research and Trading | Leading Expert in Quantitative Algorithmic Trading Strategies

 Wednesday, October 7, 2015

What is a Meta-Strategy? In my previous post on identifying drivers of strategy performance I mentioned the possibility of developing a meta-strategy.


Print

6 comments on article "A Challenge to Improve Your Trading System Using a Meta-Strategy"

photo

 Ludo L., Ing info chez NA

 Wednesday, October 7, 2015



Jonathan, your US Bond strategy is based on? signal processing ? volatility forecast ? u can give some clue ? :)


photo

 Jonathan Kinlay, Quantitative Research and Trading | Leading Expert in Quantitative Algorithmic Trading Strategies

 Wednesday, October 7, 2015



Well, the underlying strategy is not really the point of the post. But, yes, it make use of signal processing and machine learning concepts.


photo

 Edmond Yip, Project Management

 Thursday, October 8, 2015



Logi regression ?


photo

 Jonathan Kinlay, Quantitative Research and Trading | Leading Expert in Quantitative Algorithmic Trading Strategies

 Saturday, October 10, 2015



Not using logistic regression in this model, although it is a generally useful technique.


photo

 Edmond Yip, Project Management

 Saturday, October 10, 2015



Co integration approach ?


photo

 Franck Bardol, Data Scientist - data to predictions - machine learning

 Sunday, October 18, 2015



congrats, promising results !


but .... just few reading about statistical learning help anybody to have a bird eye view about trading system development.


One of the most important concept to keep in mind is BIAS-VARIANCE Trade Off. https://en.wikipedia.org/wiki/Bias%E2%80%93variance_tradeoff


Roughly speaking for those non familiar with it, expected errors made by a forecasting system can be express in 3 terms :


bias + variance + IRREDUCIBLE noise


In short, you can not reach 100% success due to :


* if you decrease bias (precision related term) you increase variance (capacity of your system to generalize results to non-seen datas). vice versa


* noise means that the underlying mechanism that you attempt to predict/hack is noisy


All of this - demonstrated - theory seem to not apply to your result ....


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