NAMO data is weird

Discussion in 'Technical Analysis' started by wmwmw, Sep 18, 2021.

  1. wmwmw

    wmwmw

  2. jharmon

    jharmon

    I think you don't understand how the McLellan Oscillator is calculated, and the difference between a market-cap-weighted index and the how advancing and declining issues are (mostly) equal weight.

    Don't guess on what it means - backtest to see if your guess has any edge.
     
  3. wmwmw

    wmwmw

    What to backtest? LOL.
    I have 30 years trading experience and I check NAMO everyday.It is always on a big up day NAMO is up and big down day NAMO is down.
    I said I have never seen it before and you want me to backtest?
    How do you backtest something that never happens before?
     
    Last edited: Sep 18, 2021
  4. jharmon

    jharmon

    So Namo goes up and index goes down - that's easy enough to backtest and determine how many times it has happened before.

    It won't be "never happened before"

    Do you even backtest bro?
     
    jtrader33 likes this.
  5. wmwmw

    wmwmw

    You can't understand plain English?

    If I wrote down 13, you will say it is a 3.
     
  6. Overnight

    Overnight

    I believe what wmwmw is stating is that he has been looking at the MAMO for 30 years, and has never seen anything like what has happened on Friday, with the MAMO up and the index down.

    If that is totally true, then a backtest would show the same result. His 30 years of watching the MAMO is the same as a backtest. There's no need to do it, because his memory IS the backtest.
     
  7. jharmon

    jharmon

    Just because someone says he's "never seen it before" and "has 30 years trading experience" is a basic "false authority/appeal to authority/appeal to accomplishment" logic fallacy.

    Oh, but there is a need to do it.

    This has happened 7 times in the last 2 years alone which I've determined through about five minutes of coding. Obviously there may be some differences between data sources and I freely admit that my code may contain errors. I post it here for others to test/replicate/critique rather than making sweeping statements about it.

    Code:
    import norgatedata
    import pandas as pd
    
    indexticker = '$COMP'
    advancingissuesticker = '#NASADV'
    decliningissuesticker = '#NASDEC'
    
    timeseriesformat = 'pandas-dataframe'
    start_date = pd.Timestamp('2020-01-01')
    indexdata = norgatedata.price_timeseries(indexticker,start_date=start_date,timeseriesformat=timeseriesformat)
    
    # Search for a situation
    # "Nasdaq was down 137 points yet NAMO was up."
    
    indexdata['chg'] = indexdata.Close - indexdata.Close.shift(1)
    indexdata.drop(columns=['Open','High','Low','Volume'],inplace=True)
    
    # Add in A/D data
    adv = norgatedata.price_timeseries(advancingissuesticker,start_date=start_date,timeseriesformat=timeseriesformat)
    dec = norgatedata.price_timeseries(decliningissuesticker,start_date=start_date,timeseriesformat=timeseriesformat)
    
    # Cleanup colummns
    adv.drop(columns=['Open','High','Low','Volume'],inplace=True)
    dec.drop(columns=['Open','High','Low','Volume'],inplace=True)
    adv.rename(columns={'Close':'adv'},inplace=True)
    dec.rename(columns={'Close':'dec'},inplace=True)
    
    # Merge A/D data into index data
    indexdata = pd.merge(indexdata, adv, how="left", left_index=True, right_index=True)
    indexdata = pd.merge(indexdata, dec, how="left", left_index=True, right_index=True)
    
    # Calc McOsc for Nasdaq
    indexdata['mcosccalc'] = 1000 * (indexdata.adv - indexdata.dec) / (indexdata.adv + indexdata.dec)
    indexdata['mcosc'] = indexdata.mcosccalc.ewm(span=19,adjust=False,min_periods=19).mean() - indexdata.mcosccalc.ewm(span=39,adjust=False,min_periods=39).mean()
    indexdata['mcoscchg'] = indexdata.mcosc - indexdata.mcosc.shift(1)
    
    # Find "Nasdaq was down 137 points yet NAMO (Nasdaq McClellan Oscillator) was up."
    # To prove/disprove "I have never seen this before." by wmvmw
    
    indexdata['wmvmw'] = ((indexdata.chg <= -137) & (indexdata.mcoscchg > 0))
    
    # Show those rows that satisfy the condition
    print (indexdata[indexdata['wmvmw']])
    
    Output:

    Code:
                      Close         chg     adv     dec   mcosccalc      mcosc   mcoscchg  wmvmw
    Date
    2020-03-20   6879.520020 -271.060059  1307.0  2017.0 -213.598068 -49.251003   2.724048   True
    2020-09-16  11050.469727 -139.850586  2046.0  1358.0  202.115158  10.166898  10.642873   True
    2020-11-09  11713.780273 -181.450195  2373.0  1202.0  327.552460  31.498248  13.775218   True
    2020-11-10  11553.860352 -159.919922  2133.0  1393.0  209.869537  37.234422   5.736174   True
    2021-03-08  12609.160156 -310.990234  2007.0  2080.0  -17.861511 -57.224157   6.734503   True
    2021-04-14  13857.839844 -138.259766  2292.0  1885.0   97.438354  -3.683592   6.163399   True
    2021-09-17  15043.969727 -137.950195  2238.0  2209.0    6.521250  -2.064219   2.165337   True
    
     
  8. Overnight

    Overnight

    Then code an algo to take advantage of his flaws, make millions of dollars and retire. Holy shit dude! Where's your journal on ET?!?
     
  9. wmwmw

    wmwmw

    No, no,you understood wrong again.
    There were days that market were down and NAMO is up.
    But I have never seen a BIG down day like Friday and NAMO was up.
     
  10. Overnight

    Overnight

    AGAIN? I've never interacted with your shit before. And since that is your attitude...

    Cry havoc, and let slip the dogs of war!
     
    #10     Sep 19, 2021