Comments on: Build an ETF Trend System in Excel

By: Tom

Tom — Thu, 04 Nov 2010 22:20:30 +0000

@Carlos: dude this posts is like so old.

By: Carlos

Carlos — Thu, 04 Nov 2010 16:21:45 +0000

This is an interesting system.
However, instead of a VBA macro building dinamycally the averages, one could just define a Excel spreadsheet with pre-prepared averages and graphs, replacing the values with data from other ETF.
Graphs ane instantly built.
The key point is to KNOW HOW TO USE THE GRAPH created by this technique.
Where is the BUY signal, where is the SELL signal.
Good job !

By: Tom

Tom — Thu, 14 Jan 2010 10:40:51 +0000

Graciela,

I don't understand your question. The links to all my videos are in the Tutorial section.

By: Graciela

Graciela — Tue, 12 Jan 2010 16:45:33 +0000

buenos dias ¡
disculpe thommas Ott, le escribo ya que me gustaria ver todos los videos de rapid miner 5.0 pero no le entiendo muy bien a la velocida que usted habla ingles para mi es demasiado rapido so me gustaria que me diera un link donde este los videeo tutoriales ya que son muy utiles para mi
esperando respuesta favorable me despido de usted
GRACIAS

By: Tom

Tom — Sun, 12 Apr 2009 20:19:11 +0000

Jim: Thanks so much for commenting. Do you have a background in programing, research, or both? I’m curious to find out how you ended up in this industry and applied your coding to the financial markets.

By: Jim Chappell

Jim Chappell — Sun, 12 Apr 2009 14:11:12 +0000

I always am greatly amused when people who have just discovered “automated arithmetic” methods try to apply them in predicting the markets. I have developed, invented, and written NN codes and models for MUCH more than a decade. I have even imbedded genetic algorithms in the back prop sections of consensus nets. I have used specialized wavelet bases and independent component analysis etc. etc. In short, I have used more advanced techniques than you have even (yet) dreamed of. What is my point? Financial markets and their instruments CAN NOT BE PREDICTED with any regularity that will produce a consistent profit. Eventually you will discover that the ONLY way to consistently make money using computational “stuff” is to sell the “stuff” to the wide-eyed. But, have fun!

By: Tom

Tom — Sun, 16 Nov 2008 22:34:56 +0000

Jason: agreed. I talk about the coefficient of determination (R^2) in one of my blog posts somewhere.

By: Jason

Jason — Thu, 13 Nov 2008 21:44:58 +0000

One thing to note is the fitness of the linear regression, if one or more of the linear regressions are not particularly in terms of the R squared statistic you might be able to infer that the movement is choppy, whereas good R squared will not be proof but it will be an indicator that your signals are telling you the right thing.

Better to have no signal if it is going sideways than a false positive.

By: Tom

Tom — Thu, 21 Feb 2008 11:00:57 +0000

excel chic: Try downloading the example spreadsheet #2 or #3, the macro is already loaded into it.

By: excel chic

excel chic — Tue, 19 Feb 2008 13:42:51 +0000

On an aside, when I copied and pasted the macro from the post text, it gave me an error (I think unidentified sub).