Day-Night System

Discussion in 'Automated Trading' started by syswizard, Jul 15, 2019.

  1. I posted these articles on the ES journal forum and they received very little comment.
    The first article is an update of the second. The latter being an aged, extremely scholastic study with a lot of detailed statistics. It badly needs to be updated as the data used is over 12 years old....and a lot has changed with respect to the EMini market since then.
    I would appreciate any and all comments on the two articles.
    Note: please download to your hard-drives to see my highlights and comments.
    Article 1: https://tinyurl.com/syswizard-daynight-1
    Article 2:https://tinyurl.com/syswizard-daynight-2
     
  2. ph1l

    ph1l

  3. Big AAPL

    Big AAPL

    Sys...you've been here a while. Pull up a daily of the ES. A down bar in an uptrend (and vice versa) banging up against a major MA almost always reverses. Forget the blanket statements about "the market only works at night". It is true only to an extent. Still, it requires some analysis. Keep your stops realistic and reenter when you are wrong. JFYI Down trends tend to be more aggressive.
     
    EsKiller likes this.
  4. tommcginnis

    tommcginnis

    Three thoughts:

    1) It's the living embodiment of the semi-strong Efficient Market:
    • the market went up.
    the market went up on news.
    • most of the market news was earnings.
    • most earnings news come out after hours.
    It would be foolish to think that market-moving effects are going to be digested while the market hangs out idle til we're done.
    (Your p.3 pretty much spells this out, FWIW.)

    2) Most of the *down* news over the past 30+ years has been macro, and while a lot of that comes before the open, *most* of it comes a mere hour or so. Not much time to digest. So what would make sense here?
    • Markets fall more (often and to a great degree) during the day.
    • Markets fall when the herd spooks: excepting Brexit, China de-val, and the overnight swoon of The Orange Instability election, when do we get massive volume after hours?

    3) Survivorship Bias: looking at the rise in the S&P colors the events to the bifurcation of the S&P's relevant daytime/nighttime events. Repeat this study on the Nikkei! WOW!

    SO!
    Up: business (risk) event: earnings: overnight.
    Down: market (risk) event: macro: daytime.
    Who?: The S&P rose == what of the Nikkei?

    Them's my thoughts. THANKS for the post!
    TMc
     
    birdman, nooby_mcnoob and Big AAPL like this.
  5. Dude - the statistical evidence is overwhelming....the academic study (article 2) proved that. What they failed to do is create a Sharpe Ratio for the final equity curve result.
    Article 1 also fails in that regard.
    Still, it's hard to ignore this concept. It just needs to be enhanced with filters and other buy/sell logic.
     
  6. Tom - thanks for the response but you gotta back-up your assertions with statistics as was done in article #2.
    That article likely took over 5 man-years of effort.
    Thoughts are one thing.....research is entirely different.
    I'll go with the research every time.
     
    Last edited: Jul 15, 2019
  7. tommcginnis

    tommcginnis

    Errrrr, no I absolutely do not have to do any such thing. Noooooo no no no no. {Time for Consuela from Family Guy...}

     
  8. i think you are on the right track but only have revealed about half of what you need to succeed.

    think of data as continuous stream where only movement matters and time is omitted. when you focus on the movement and not when the movement happens, plans become more likely to happen.
     
  9. lindq

    lindq

    Plans without time. Far out, man.
     
    tommcginnis likes this.
  10. how would someone harness the ocean using time? so i think time is just as useless trying to harness market data.
     
    #10     Jul 16, 2019