Forecast Russell 2000 instead of SP500

24Mar10

Other researchers point to the fact that they couldn’t find reliable technical indicators having
‘significant’ prediction power for the SP500 and the Dow, but they could find for less ‘optimal’ markets like the Nasdaq and the Russell 2000.
So, I tried my NN testing with adjusted IWM instead of SPY.

After 37 tests: (from 2000-2010)
winLoseRatios Arithmetic Mean: 51.07%, stdev: 2.28%
So, this is an edge of 1%. Hardly worth playing.

projectedCAGRs Arithmetic Mean (biased to the upside): 9.30%, stdev: 28.79%
the stdev of the CAGR is not convincing.

The parameters:

nNeurons = 20;
nDaysInTest = 260;
nInputSignals = 5; %+3; % close and volume and open, high, low
nDaysInInputVector = 5;

That is again no a good news. I expected more.
Note: I adjusted the price, but not the volume and the IWM has a 2:1 split. So, in theory, the volume should be adjusted as well.
That is a bug (to be honest), but I don’t think its correctness would modify the 51% prediction power of the network.

Advertisements


No Responses Yet to “Forecast Russell 2000 instead of SP500”

  1. Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: