Many thanks to all for your thoughts and contributions. As I have said repeatedly, I have no doubt whatsoever that there are trends and that such trends are caused by supply and demand in the economy and human behaviour. I do not believe in the EMH. I do not believe that the markets are a markov chain. I believe that there is autocorrelation in markets for periods of time as participants adjust to ongoing economic reality and herd instinct. What I do believe however is that these trends have become increasingly difficult to follow and profit from using mechanical systems in the recent past. Analysis of autocorrelation, of the profitability of trend following systems, of the âefficiencyâ of trends all point in the same direction. Such analysis is confirmed in the real world by the effective collapse of the methods of various long term trend followers in the mid 2000s: Dunn and JWH were lead to near ruin, Abraham and a slew of others were forced back to the drawing board and made changes to their portfolios and their systems. 2011 and 2012 wrought similar havoc among trend following CTAs. And a similar rush of head scratching. So the purpose of this thread was not to deny that trends exist nor to claim that it is impossible to benefit from them financially through the use of systematic trading. It was to provoke thought on how systems might survive into the future and continue to benefit from trends, while avoiding being destroyed by noise and sideways choppy markets. Some believe that trend following systems can be turned on during trending periods and counter trend systems can be turned on during choppy, sideways markets and vice versa. Some believe that running a multitude of different systems at the same time will enable the program as a whole to profit from any market condition. I have made much the same request for discussion in a number of different forums with widely varying results. An anonymous well wisher on Nuclear Phynance made the following helpful suggestion: âIMO this thread belongs to trash.â Some kind souls have, as usual, heaped derision on my efforts and responded with scorn and unpleasantness. Others seem to wish to take anything I say as a personal insult and feel the necessity to respond in a like manner. Happily, the majority take what I say in the spirit in which it is intended. Riccardo Ronco for instance instead pointed to useful and interesting research papers. To all of those, thank you for your contributions.
You say: "So the purpose of this thread was not to deny that trends exist nor to claim that it is impossible to benefit from them financially through the use of systematic trading. It was to provoke thought on how systems might survive into the future and continue to benefit from trends, while avoiding being destroyed by noise and sideways choppy markets. Some believe that trend following systems can be turned on during trending periods and counter trend systems can be turned on during choppy, sideways markets and vice versa. Some believe that running a multitude of different systems at the same time will enable the program as a whole to profit from any market condition." Thus your solution for trending and then chop is two systems One called a trending system and the other called a counter trending system. Trending and counter trending are not opposites as you point out. Handling trending and opposite trending would just use the one systemmic approach and the full offer would be taken for both. As I mentioned, I use one system for both trending and chop. I do not use an entry/exit; pehaps you do since you mention a host of those who still do and have troubles. I think of price as often being in the shadow of a brief higher volatility moment. The information group of this moment can take on one of three shapes depending up on up to four possible useable values of the shape. (OHLC). My solution while I only have to use one system is to examine the context of the chop in this shadow. Chkitout recently modified Brooks' systemmic approach in a non important non data context period to make a mistaken point. See his red marks elsewhere. My position going forward from the event of the beginning of the shadow is to hold and assess the context of the chop. I do this by going to the fastest observeable fractal where there is always trending and never chop. This is a contextually siginificant change. None of your referenced operators do this. Thus I am observing sentiment on a trading fractal level and on a faster fractal sub fractal level of the trading fractal. For me all fractals are nested in a fixed unchanging ratio. (3:1) For deeply and precisely resolved technical considerations, I observe (during this non money making shadow interval), the sentiment changes on the subfractal pompared to the unchanging trading fractal. I resolve the condition, situation and context at a specific point after the shadow commenced. In effect this puts me on the correct side of the market from that point onward and I am trading on the sub-fractal continually taking small profits (See the level of trading of Lucrum). Thus at the time the market resumes trending (a BO of this lateral) I am in synch always with the trading fractal and I take the opportunity to continue to move forward profitably. Thus, keeping one system that totally relates to the systemmic operation of the market is always possible. As a person takes the trip of dealing with CW's "noise", "flaws" and "anomalies", he eliminates these through deductive reasoning. In effect, the usual process of collecting "edges" reaches a peak number and, from then on, the edges are merged into systems where two groups form and they are named: "trend systems" and "counter trend systems". Bridging these two systems depends open higher principles. some of the shifts that facilitate this are: 1. shifting from entry/exit trading to hold reversal trading. 2. shifting form difining exits and entries as separate entities to understanding exit and entry are an identity with just one contextual variant (sentiment). 3. shifting from probabilities o certainties by having a moment in an information parcel where no more change can occur (prbailities of alternatives goe to zero from that point onward. 4. developing the "functional toolks for all of the above by using the correct Algebra for the functions in non continuous functions. You suggested you were opening a conversation. Notice what has prevented you so far. Ras suggested to the magician he is in a difficult place. The magician will not change his limited view for his reasons. Ras doesn't care because he will not waste time coaching such a person as the magician. Ras does make contextual comments because he knows ohters can "get" the value of his coaching. You do not coach. You as here to gather information to solve your problems in a better way. One conclusion that I am presenting to you is subtle. Notice that I do not predict but I do let each information parcel resolve itself as it "locks in" to its "certain" identity before the end of the information parcel. People are making crass and harsh comments to you. They do this from a vantage point. I do not feel I can help you but you are an example for others to consider. You do need to gather more intelligence to be able to helpyourself.
Thank you Jack for your pleasant response. I do not need or seek "help"; I merely enjoy sparking rational and civilized discussion, to which you have contributed. The majority of people on internet forums, particularly anonymous ones like this, enjoy the surrogate violence of being unpleasant in words. Perhaps they also enjoy physical violence, who knows? The search engines will not have caught up with it yet, but if you are interested you can no doubt find the attached thread which I have written elsewhere, in a place where, happily, I can repel those with a vitriolic intent: "David Attenborough Explains Internet Brutishness"
All directional trading depends on prediction. Period. and other trading depends on predicting volatility. No amount of rhetoric can change this fact. surf
Where I live in Canada (Ottawa), blackjack nowadays is played using 8 decks, a sealed shuffling machine and cards are reshuffled when ~1/2 of the cards have been dealt. I'd like to know how you can see "exact probabilities" in that setting. BTW, do you know of a single casino still playing blackjack w/ only 2 decks?
Cruise ships based casinos are single deck but the deck is continuously being shuffled--- with the used cards. My experience only on one ship.
No problem seeing the exact probabilities, but if your real point is the difficulty of finding an exploitable edge in that situation, I fully agree. Las Vegas, maybe. Can't say for sure, I haven't been there in over a decade.
Yes but as I painstakingly pointed out, there are different levels of prediction. The unadorned word "prediction" too often conjures up images of mystics forecasting unlikely events ala supermarket tabloids with their annual top-20 "psychic forecasts" for the new year or whatever. But we humans anticipate things every day, and these anticipations are nothing but predictions of likely events, and most of them come true. Stop acting like anticipating certain price actions by using logic and experience is some mystical event. It's not, and it shouldn't ever be treated as such.
There are a small percentage of traders that make money jumping on moves for the ride. There is another small percentage tht fade moves and make money. Most occupations that have a potential pot of gold at the end of the rainbow -- Rock-n-Roll, Acting, Proffesional Sports -- have very few pots for very few people. A few people make tons of money trading, somewhat more make a good living and after that it goes downhill very quickly. It is more about the trader than it is about following or fading.