Skip to content

Predicting the Market from the Inside Out: New Breadth Measures

April 23, 2010

I always wondered if it was possible to predict the market better from the inside out by looking at the constituent stocks of the index to generate signals. After all, it seemed that everyone used breath indicators in some form or fashion. I did quite a bit of  research using classic breadth indicators and came up with only a handful of of indicators that actually worked better than simply using timing on the market itself.  Of course, (as some of you know) I have a very persistent and almost stubborn desire to solve research problems. I won’t give up until I feel I have exhausted all the logical and simple hypotheses that can be tested. Unlike the typical scientific skeptic who writes everything off before testing (ok its an oxymoron but true!), I like to keep an open mind and have the opinion that if enough people claim that something is useful it is worth examining in a serious way.

 In this case, after spending time formulating the confounding variables, I had the breakthrough moment. As it turns out–not all stocks are equal, and classic breadth measures assume that each stock gets an equal vote. In reality, when we try to measure breadth, we are attempting to gauge a broad flow of money in order to have a clearer picture of the market. We are looking at buying and selling power on the premise that it is difficult to hide what is happening across a wide variety of stocks. One of the factors in the Livermore “Active Issues” rankings relates to volume activity. I decided that I would apply this proprietary variable as a filter to identify the stocks worth focusing on. Once I used this approach, suddenly breadth appeared a lot more powerful. So I decided to take it one step further—why not apply an adaptive indicator to the constituent stocks to improve timing signals?  Well as it turned out the sum of the parts beat the whole—using the indicator on the selected stocks to generate signals to trade the market outperformed using the the same indicator to trade the market on its own!

now that is good!—-to be continued soon

5 Comments leave one →
  1. bootstrap permalink
    April 23, 2010 2:15 am

    you sound exited, but i can’t parse your punch-line sentence,
    “Well as it turned out the sum of the parts beat the whole—using the breadth-based indicator on the market outperformed using the the same indicator on the market in a substantial way!”

    can you rephrase that?

  2. April 23, 2010 8:19 am

    Sounds good indeed. Will you be giving a bit more details or leave us with a teaser of a proprietary indicator? 😉

    A side question I was wondering regarding your (and my) research: what is your method to ensure the strategy is really statistically significant:

    ie doing millions tests, you are bound to find some strategies that exhibit statistical significance. For example, If there is a 1% probability that strategies that exhibit statistical significance result from chance, how do you ensure that your statistically significant strategy is also statistically significant (to the second degree)

    @bootstrap: I think David just meant that a breadth indicator looking at individual stocks components of the S&P 500 would give better results than an indicator looking straight at the S&P 500 (or other index)

  3. Dave Svilar permalink
    April 24, 2010 6:47 pm

    When you did your series on “trendy” vs “mean-reverting” stocks you mentioned that it showed promise in terms of breadth indicators. I assume this is a follow up on that? Cool!
    Personally, I’ve found the bullish percent indexes to be the most useful for my own trading. However, I use it more for a gauge of risk/reward rather than an actual timing indicator. Can’t wait to see what you’ve found.

  4. jgpietsch permalink
    May 4, 2010 8:31 am



  1. High Frequency Trading and the Online Retail Investor « CSS Analytics

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: