Emini Divergence revisited

Discussion in 'Index Futures' started by lescor, Jan 18, 2004.

  1. lescor

    lescor

    I'm not much into reading other people's journals, but from what I've seen, the "Emini Divergence Journal" by no_pm_please was one of the best.

    http://www.elitetrader.com/vb/showthread.php?s=&threadid=20761

    He basically taught a very solid system and did it with great clarity. I have been trading this system since September exactly as it was taught in the journal and am very satisfied with the results. The market's volatility and volume dropped off significantly in the last quarter of '03 so the trade set-ups were less frequent, but hopefully things will return to normal (abut 4-6 trades/week).

    Here are the results of all the trades I've done:
    45 trades, 60% win rate, avg win 3.6, avg loss -1.9, 62 points total, expectancy 1.4pts/trade. My numbers are in line with what no_pm was generating over a two year period, and also my own backtested results.

    The reason for this post is to see if anyone else has been trading this system and to get your feedback on it or any revisions you've made. The author of the journal also gave some insight to a second, related strategy, but illness caused him to abruptly stop posting and he never gave the last few details of this system. Has anyone figured it out? I just think there was too much value in that thread for it to have ended the way it did, so I'd like to bring it back to life.

    And lastly, to Mr. no_pm_please: If you read this post, please know that your journal was fantastic and that there are a lot of people who hope things turned out well for you and who would love to see you back on these boards.
     
  2. Ditch

    Ditch

    I've just read the thread, but probably overlooked the hints to the second strategy. I would appreciate it very much, if you could post a link to that. Thx.
     
  3. Lescor,

    That thread was what got me to register here instead of being a lurker. That, and one from VisionTrader got me interested in divergences in general. Since then, there was also a thread about ADX/DI that got me into looking at the DI (DMI for some people) and not the intersect stuff people talk about but the divergences the two lines show. And then there's dbphoenix and his instruction about S/R levels and looking for divergences at pivot points. It's all starting to add up.

    As far as me using No PM's system on it's own, I don't do that. Perhaps I should based on your results. But I have the Chaiken and the ADX and the Keltner up as part of my indicators and I can see, almost every day, where his instructions of a setup come up.

    My biggest problem with divergence trading is the trouble with trend days. I lose money anytime that happens. Probably you would say that the times No PM's setups don't work are on those days, or would you not? If I could just find a reliable way to avoid my divergence setups on strong trend days, I'd be very happy. I've been studying how to determine trends and just don't see it yet I guess, but this is another subject for another thread.

    It could be that your findings and your success, even with the losing trades, is the lesson I need to learn. Losing is OK as long as the winners are larger overall. Maybe I don't need to avoid the losers with No PM's setups....just accept them as part of doing business, your "cost of goods sold" so to speak.

    I appreciate you posting your results. At the time of NO PM's thread there were a few LOUD voices here that said it would not work over time. I am glad to see you confirm his setup works. It also makes me glad because I got the feeling NO PM was a pretty decent guy and I'm happy to see that his contribution was real and useful.

    jd
     
  4. Boomer

    Boomer

    i too have become interested in divergences, and i am about to go back and read his thread and review his strategies. i look forward to this thread also...
     
  5. lescor

    lescor

    No_pm used three methods to avoid trend days, and I do too. I don't try to anticipate trend days and skip setups as a result, I just take every signal that appears and let the historical stats take care of the bottom line. I've got enough confidence in this system now that a month ago I funded a seperate account to trade it, risking 2% of account equity per trade.

    Here are the three rules to avoid big losses on trend days:

    "The easiest way to avoid a big losing day is to stop taking trades if the first trade is a loser and you suspect it may be a trend day.

    A second method is to wait for the ADX to drop below 30 and cross above it again before starting the setup process.

    A third method is to wait for the trend to be broken by touching the other side of the Keltner Channel before looking for another trade."

    I skimmed most of the thread again looking for the few nuggets of info on the second method, but I'll have to try again when I have more time. It involved pivot points.
     
  6. Thanks Lescor.

    Question: you use those three things to avoid trend days or, as you say in the first paragraph, you take all setups anyway? Or you take all setups AFTER using the three things above?

    thanks,
    jd
     
  7. MarkB

    MarkB

    Here is a link to the thread which no_pm started on his other method:

    http://www.elitetrader.com/vb/showthread.php?s=&threadid=21347

    I really don't think he was referring to pivot points per se. Not to take away from his excellent and lucid posts and writing skill, but after reading some of his other posts on the board, I think he had a tendency to use the term "pivot" when he meant swing highs and lows. Of course, should a pivot s/r level land within the keltner channel during a trade, there is a high probability that it will serve as a swing point as well, thereby warranting a shift in stop placement as according to his method.

    As Lescor knows, I've been using no_pm's method as well. Indeed it has proven profitable, despite a few week long periods with no signals generated. Lescor has been even more successful that I have been with this technique, as his patience is considerable :)

    Fortunately Lescor has kept more careful notes than I have. I'm looking forward to a correlation of winning trade percentage vs. time of day, as I believe that I've spotted a volume consideration which may at times generate false signals. Our best signals seem to have come right near the open, followed by those in the later (after 2 PM ET) afternoon. I think that signals generated right before or during the lunchtime lull account for the majority of losing trades. Perhaps others who've been routinely applying this method have noticed the same thing? As I've discussed with Lescor, should this theory prove to be correct and not merely anecdotal, there is a logical explanation as the system is volume based.

    And indeed, I couldn't have phrased it better: no_pm, I too sincerely hope that all is well with you. Please visit when you can; not to answer questions, just to let us know that you are okay.

    MarkB
     
  8. MarkB,

    It makes sense what you said about the setups not being as reliable around the start of the lunch hour vs other times. When I see a DV (divergence) on my indicators like Chaikin or OBV, and it's near 9am (noon for eastern time people), I am usually skeptical. The volume between 7am and 9am drops off maybe not because of a direct relationship between price action but as a result of less traders at their computers. This would not effect your trading if you used just candles or just price action, but if you used volume as related to volume the last hour like a volume based indicator like Chaikin does, it might screw things up. Does this make sense to you?
     
  9. MarkB

    MarkB

    Yes precisely JackDaniel.

    What I've seen is a new high (for example) right before lunchtime. Then a higher high with lower volume creating the divergence. And as you said, fewer people trading then. If you enter, the trade then meanders about, and often a new high is made when everyone comes back from lunch, resulting in a losing trade.

    Mark
     
  10. Lancer

    Lancer

    Are the profits generated by those particular/few divergence signals so great that they justify missing out on trend days? IMO, divergence signals are valid on trend days (so trend days need not be avoided), but for those counter-trend pullbacks following the divergence signal, consider what signals get you back in the direction of the trend; (avoiding a stop-out on the divergence position).

    Maybe a counter-trend reversal signal can be found on the next lower time-frame chart? For example, for a trend continuation signal (counter-trend close and reverse), maybe consider divergence on a lower time-frame chart, or fast/slow oscillator crossover or sling, or MACD crossover, or price oscillator xover zero, or ? all back in the direction of the trend.

    Alternatively, maybe there needs to be a trend filter so that system rules are adaptable to market conditions?

    Why get whacked trade after trade due to "failed" divergence signals on trend days, or miss out on trend days altogether. What rules will improve the method so that trend days can be traded as successfully as consolidation days?
     
    #10     Jan 18, 2004