Building Your Own Trading System – Part 2
Loess smoothing calculates a number of regressions. It isn’t a mathematical technique you can just do in Excel out of the box, but if you have a tool like Octave, R or Matlab, Loess smoothing is an available function.
At present, I have a custom written Loess smoothing algorithm that was developed to my requirements by a PhD, for some custom written stock selection software that runs under Unix. It can smooth data essentially immediately on my dual core MacBook Pro. Although it is low on my priority list, I would like to see this as an Excel add-in.
If you are interested in reading further about this exciting approach, check out this link (Black Art of Smoothing).
In conclusion, this article is about identifying the problem that most traders never define – the need to smooth data in their trading model, and presents a solution that is now feasible for use.
I’ll continue with building your own trading system in future articles. As always, I appreciate your comments and feedback.
Pages: 1 2
Tags: Building systems, loess smoothing, moving averages, price smoothing, trading algorithms, Trading Systems
March 16th, 2009 at 5:25 am
Hi- I ran across your post while searching for information on whether the loess filter can be applied to stock prices. Do you use publicly available software such as Caterpillar SSA? I would be interested to find out how I could use that data filter in the markets.
Thanks,
Larry Perlman
Parsippany NJ USA
larryperlman@optonline.net
April 6th, 2009 at 4:41 am
Hello,
The software is custom written in Python. However, Loess smoothing is a relatively common addition to packages such as Matlab.