SMA(30)/SMA(45) combination, 2 dimensional case beats the 1 dimensional cases


Recently, I was advised that instead of using Moving Averages that are far apart (like 20/180), a combination where the moving averages are close to each other may prove more useful.
For that I test the combination of SMA(30)/SMA(45) pairs.
You can convince yourself, that the crossing of the SMA(30)/SMA(45) happens faster than the crossing of SMA(20)/SMA(180),
therefore this new version can react to changing market quicker.


  • only MA(30): winLoseRatios Arithmetic Mean: 51.45%, stdev: 5.28%, avgBarGainPercent mean: 0.08%, stdev: 0.28%, projCAGR: 4.21%***p_test: 2502
  • only MA(45): winLoseRatios Arithmetic Mean: 52.07%, stdev: 4.91%, avgBarGainPercent mean: 0.07%, stdev: 0.28%, projCAGR: 3.74%***p_test: 2947
  • MA(30)/MA(45): winLoseRatios Arithmetic Mean: 53.55%, stdev: 4.24%, avgBarGainPercent mean: 0.17%, stdev: 0.29%, projCAGR: 9.41%***p_test: 2500

This is an example in which the 1 dimensional cases were not successful, but combining 2 1 dimensional NN into a 2 dimensional NN improved the result.
In a previous post, I said that combining 2 dimensions didn’t improve the result.
Here is the example when it improved; so ignore please that previous blog post.

So far the SMA(30)/SMA(45) combination is one of our best results.
It beats the SMA(20)/SMA(180) version by about 1%.
The SMA(20)/SMA(180): winLoseRatios Arithmetic Mean: 53.66%, projCAGR: 8.56%
SMA(30)/SMA(45): winLoseRatios Arithmetic Mean: 53.55%, projCAGR: 9.41%


No Responses Yet to “SMA(30)/SMA(45) combination, 2 dimensional case beats the 1 dimensional cases”

  1. Leave a Comment

Leave a Reply

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

You are commenting using your 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: