How to "sit" on a position during a big move

Discussion in 'Trading' started by Ghost of Cutten, Aug 25, 2010.

  1. Let's have a thread on "sitting" i.e. holding on to a position that is working well during a long-term trend. What techniques do you use to stay in for a big move, how do you avoid getting faked out by typical corrections and pullbacks, how do you know when the trend has finally ended and it's time to bail, how far away do you trail a stop, if you use stops (if not, what's your risk control method).

    There's a clear tradeoff between competing objectives - staying in the move i.e. avoiding being stopped out prematurely; versus not staying in once the move has ended. When is a 10% counter-trend move just a test or a correction, and when is it the start of that 50% bear market that is going to kill all the longs who don't have an uncle point.

    There seem to be three main approaches to trading big moves. The first one, which we could call the "hedge fund" approach, is to formulate a view, place the position, have no stop, and just sit and pray that it works. This guarantees big profits when you are right, and career-ending drawdowns and client redemptions when it doesn't. For those who can't rape dumb money by exploiting the trader's option inherent in non-clawback bonuses equal to 20% of annual volatil^H^H^, I mean profits, this is probably not a sustainable approach. Some of the more pragmatic hedgies have modified this by scaling back their exposure if losses exceed a certain point, but it is rare to see a hedge fund player who will cut and run just because their favourite market has corrected 10-15%. Also, if the typical correction is just that, then it actually makes more sense to add, not scale back, when one occurs.

    The second approach is to avoid the career-ender by using a stop. A common variant is to have a more mechanical method which basically follows momentum, with a trailing stop - like the Turtles. So, let's call this the technical trend-following (TTF) method. The vulnerability here is having close stops and thus frequently being stopped out of a move too early, or having wide stops and suffering large drawdowns when the move finally ends (e.g. the turtle method). A further disadvantage is that by being less selective about the moves you enter (since the technical approach will buy false breakouts as well as genuine ones), you will suffer lots of losses from false breakouts. This might be good for reasonable returns if you can stomach the volatility, but it seems like there is scope for improvement - perhaps by reducing the incidence of false breakouts, and by having some kind of method for reducing exposure by scaling out once price and sentiment get to the kind of extreme that commonly accompanies a major reversal.

    Finally, we have what could be called the pure speculator approach, which is to try and exploit the manic depressive tendency of markets to go from one extreme to the other. In theory, this is done by accumulating positions before a trend has become widely recognised, or at least early on in the move, then holding on as the trend builds in momentum and recognition, and finally exiting into extreme strength or weakness once the move has turned into the latest flavour of the month. The momentum extremes that accompany hot/dumb money piling into the late stages of a move (with accompanying Bloomberg headlines parroting it so frequently that even the dentists have piled in) are often reasonably easy to recognise by the rapid price acceleration and speculative/media frenzy that accompany them. But recognizing trends before they occur, or even early in their development, is more difficult. Recognizing when a scary-looking correction is just a temporary pullback rather than the end of the move, is no child's play either. Can this approach be implemented by mere mortals, or is it too much to ask?

    Perhaps the most promising improvement is to combine the better elements of all three. Let us imagine that we have access to a good global macro hedge fund manager (or stockpicker), a good technical trader, and a good speculator. We could then have good trend selection skills (from the macro trader or stockpicker), the ability to hold on through lots of noise (from the technical trader), and finally the ability to liquidate into price/sentiment extremes near the end of the move (from the speculator). In theory, this would be about as close to the holy grail as one could reasonably expect to get.

    Your thoughts please...
  2. Cheese


    'Sitting' is usually known as holding, namely to hold a position. For daytrading the beginning and ending of holding is a function of reversing. You reverse (buy) on the end of downswing and you reverse (sell) on the end of an upswing.

    The daily session (eg CL) is made up of gyrations, being the sequential swings up and down (or vice versa), flowing from the open to the close. You methodology needs to exploit those swings.

    To identify how you should set up your system, look at the metrics of your market. Look at a number of trading days. In each one, add up the maximium points in each swing. For example in CL you could treat 35 points as the minimum for each swing. The combined maximums of the days swings is the days total maximum. Now you begin to see what is available to you. From there you set about the task of adopting or devising a system to capture as much as you can from each swing.
  3. Hmmm, this is what I do... Except I don't pray and there is a stop. It happens when I conclude that the premise behind the original trade is flawed.
  4. Tough proposition to "sit" and even harder to put down on paper.

    One thought.

    Once I have made my selection with whatever method of madness, I'm done. I'm aware of the pitfalls of confirmation bias.
    It's hard to think clearly when and if the good news appears like a slap on the back saying "atta boy".

    I prefer to read the negatives, the bad news. Does the bad news offend my position?
  5. I think the only real way to sit on a position is to be a pure technical trader, and to know what the metrics of your market are.

    For example, if I see a group of price bars all closing up, but my momo indicators are sitting around the mean of that market, and my trend indicators indicate a definite tread, I am inclined to think that this is a pull back and I will take no action unless that pullback exceeds a certain volatility parameter.

    I do take money off the table at each volatility benchmark. And there is a period of course when no identifiable tread exists, and since it is unwise to guess, then there is no position.

    I guess the real trick is to tune your indicators to respond rapidly to a change in sentiment, which will of course be reflected by the price.
  6. 1. trade the exact pattern u wanted to trade,
    sell 75% at the end of it, leave 25% for gravy

    everything after that is gravy
  7. LeeD


    Ghost of Cutten, you need to add to the list scaling in and out of the position. In a long trend it' feasible to scale out and scale back in multiple times.
  8. You beat me to it. While it would be über-cool to be prescient, no one knows how far a market will retrace or whether it is a reversal in the making. There is only action and reaction, the latter preferably based on predetermined criteria. I regard anything else as mostly wishful thinking.
  9. I just wait till options expiration and either I get stock called away or assigned stock on a put.
  10. The $$$ profit from a winning trade is considered "house money" by the TF guys I know. They see trade profits as almost separate from initial trade capital. I don't see it that way myself, though.
    #10     Aug 25, 2010