Trend Following Research

Discussion in 'Technical Analysis' started by Trend Following, Aug 28, 2010.

  1. kut2k2

    kut2k2

    Now I'm curious. Why should counter-trend trading be any more predictable than trend trading? If price moves randomly, both methods would be unprofitable, right?
     
    #701     Apr 19, 2011
  2. True. However, since markets aren't random in that sense, who cares?

    Most markets I've analyzed have a counter-trend bias at short time frames, and a trend bias on longer time frames. The doesn't mean you can't trade the opposite style, but the sophistication required of your entry decision making/filter is much higher if you're trading opposite the nature of the market.
     
    #702     Apr 19, 2011
  3. kut2k2

    kut2k2

    Thanks for the info. Is it safe to assume that you're talking about the 40-bar EMA of closing price?
     
    #703     Apr 19, 2011
  4. This is more or less the Mebane Faber "model" that's been tossed around in the press for the past two years.

    On another note, the trend-tracker blog mentioned earlier is now updated thru Mar, http://bit.ly/hBxW4j
     
    #704     Apr 19, 2011
  5. Sure, although in practice it makes no difference which side of the weekend gap you use. For that matter you could use the Wednesday S&P pit close as your price for the week if you really wanted. Actually, you could use any price from any specific time in the week. The effect is way too strong to care about minor details like that.

    Incidentally, this is far from the best system for trading the S&P on weekly bars. It's too tweaky and reverses its position way too often. But it is perfectly good for statistical purposes to prove trends are real - since there are roughly 1500 data points in the modern (post-Volcker) era and and the method uses readily available data, it allows one to disprove the efficient market hypothesis with a confidence nearing absolute certainty.
     
    #705     Apr 19, 2011
  6. It's WAY older than that. I believe it's mentioned as a working system in Market Wizards (or some similarly sage old tome - I don't have it handy). And the intellectual root of this line of thinking, exponential moving average crossover systems, date back to the 70's at least. SMA crossovers are older than that I think.

    The interesting thing is, anyone who's anyone has known this works for about 40 years now. And it STILL works. Funny, that :D
     
    #706     Apr 19, 2011
  7. 100% agreed with you, The Big D.
     
    #707     Apr 19, 2011
  8. Whoa! March really wasn't kind to trend followers. Are all these guys using the exact same model? Wonder if they will get back to even?
     
    #708     Apr 20, 2011
  9. Another good poke at the internal inconsistency of marketsurfer's argument, and one to which he has predictably not responded.
     
    #709     Apr 20, 2011
  10. Quote from kut2k2:

    <b>Now I'm curious. Why should counter-trend trading be any more predictable than trend trading? If price moves randomly, both methods would be unprofitable, right?

    Another good poke at the internal inconsistency of marketsurfer's argument, and one to which he has predictably not responded. </b>
    Thanks for catching this, gabfly, I did not read it earlier.

    Without another edge than the past trend, both methods are guesses-- profits would be subject to guessing then managing the trades correctly -- being lucky and disciplined is a huge part of trading without an edge.

    Both or one or the other could be profitable, as well as not profitable as the case or trader may be. In fact, they could both be profitable at the same entry depending on time frame.
     
    #710     Apr 20, 2011