Skip to content

Finding Tops: Going Short Above the 200 ma Using MABI

November 4, 2009

As previously mentioned, MABI is a great tool for identifying tops.  This is because it when  the MABI oscillator is low (<.5) the market is churning–ie the number of unchanged issues is high, and the gap between advancers and decliners is getting smaller. To give you a simple example, I will define overbought conditions using the z-score of the 100-day SMA–I will use this versus the 200-day because it is more responsive . The z-score is just the bollinger band location relative to the SMA expressed in standard deviation units—ie a z-score of 2.1 means that you are slightly above the upper band, while “0” defines the SMA. In the test, the additional requirement was that the market had to be above the 200ma. I used the HV osc–which is the historical volatility oscillator that measures 10-day historical volatility on a relative basis. When HV is low, this is bearish and often characteristic of tops that tend to have narrow ranges and strong persistence. The results of combining these variables is quite remarkable–shorting above the 200 ma is typically considered a losing proposition. Going short under these circumstances was very profitable—a strong edge:

Finding Tops: Next Day Short Edges Above the 200MABI <.5, HV 10 osc<.5
  win% avg next day return
100-day z-score    
>0 50% 0.07%  
>1 52% 0.08%  
>1.5 56% 0.13%  
>2 66% 0.26%  
>2.25 80% 0.34%  
Finding Tops: Next Day Short Edges Above the 200MABI <.25, HV 10 osc<.25
  win% avg next day return
100-day z-score    
>0 58% 0.21%  
>1 56% 0.20%  
>1.5 70% 0.35%  
>2 80% 0.28%  
>2.25 100% 0.10%  
3 Comments leave one →
  1. Anton permalink
    November 4, 2009 8:45 pm

    David, what is the number of observations for each of these tests? How robust is the idea – will it provide a similar edge in any other market? consistently? thanks

    • david varadi permalink*
      November 4, 2009 8:50 pm

      hi anton, all very good questions, i did not include the # of trades but for all but the z scores above 2.25, there were more than 20 trades (there were 8 above 2.25 for the first sample and 2 for the second). the data is for the last 12 years, however going back 40 years the edge still existed though not as large. nonetheless it was a persistent and robust effect.


  2. mike permalink
    November 5, 2009 6:12 pm

    David– off topic- but I think you’ll enjoy this

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 )

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s

%d bloggers like this: