filter liquidity of stocks

Discussion in 'Stocks' started by honza, Oct 19, 2013.

  1. honza

    honza

    Good afternoon,

    are you using filter of liquidity and price of stocks?

    My minimum condition is avg. daily volume 500 000 shares (10 days back) and price bigger than 6 USD.

    Do you have similar or different filter for looking for stocks, which are useable for trading?

    My tresholds aren't based on statistic or other method it is only an estimate.

    Thanks.
    Honza
     
    #1     Oct 19, 2013
    Share
  2. honza

    honza

    Here is an example why is condition of minimal price important.

    First equity shows profit around 900%, but is without spread.slip,fees...
    Condition for minimal price is only 2USD. min.price = 2 => 1 tick / 2 = 0.005 = 0.5%

    Second equity: min price = 2USD, spread+slip is about 0.03 per trade and profit falls on 220%.

    Third equity is with condition minimal price bigger than 10 USD => 1 tick/10 = 0.001 = 0.1%. Spread+slip = 0.03 and profit is around 440%.

    I hope here you can see how important is condition for minimal price.
    I did not find approach for determine minimal price of stocks. Do you have some idea?
     
    #2     Oct 19, 2013
    Share
  3. gmst

    gmst

    Hi, which platform are you using to generate these equity curves?
     
    #3     Oct 19, 2013
    Share
  4. honza

    honza

    Hi, I am using my own script in programming language Python. Input stocks data are from provider eoddata.com (nasdaq).

    Result of trade with spread + slip (0.03) has equation:

    long_trade = (close-open)/open - 0.03/open
     
    #4     Oct 19, 2013
    Share
  5. stock777

    stock777

    who cares.

    more important, what 'system' are you testing.

    and when you eventually own the world, will you be merciful?
     
    #5     Oct 19, 2013
    Share
  6. honza

    honza

    I thing that it is interesting for everyone who make more than few trades per month.

    System is designed on entire nasdaq and nyse due to system is rugged. This approach is little bit complicated, because I can't use genetic, clustering, self learning algorithms due to data sample is too huge and one iteration takes a lot of time. (on my PC with processor 4 GHz and 6 cores). So I am using knowledge from statistic, markets and mechanical engineering, where is many things still very much empirical due to is little bit similar with markets. I have only 2 system after 3 years research (and 4 years experiences with futures) and so a I hope you understand that I don't want present details of system.

    I just hope to achieve financial freedom :)
     
    #6     Oct 20, 2013
    Share