I updated my previous post with a link to a thread here discussing it. There may not actually be a "paper"... I misspoke. With regards to timeframe.... If you were able to calculate the "optimal" trading strategy, i.e perfect entry and exit you would find that the maximum profit would occur with the minimum time frame. Sure, the profit per trade would decrease but the number of trades would be higher. The caveat to all of this is that we have to pay commission so there likely will be a "minimum profitable timeframe"....I.e you can't profit of a 1 tick price move if your roundtrip costs are 2 ticks
Key word being "empirical". Redneck is correct. The markets are not fractal because they're not self-similar. If one has instead a very loose definition of "fractal", then it becomes useless. Yes, you have uptrends on a daily chart and an hourly chart and a 15m chart and a 5m chart and a 1m chart and a 1t chart. But not all at the same time. So of what use is the notion of fractal applied to price movement? If one were to apply this notion to the market since early 2008, for example, there would have been no point in going short at any time for any interval.
I agree that the loose definition of "fractal" does not yield any obvious trading strategies. The only thing relevant to trading that I can think of is the subject line: If markets indeed display a fractal nature, shouldn't trading on all time-frames be equally profitable?
Or you could get some healthy teeth pulled at the Dentist. Sorry Jack just messing around, I am not smart enough to know if your method has merit.
I'm not sure how a market could be fractal in the true sense because there is always that pesky minimum tick size. Just this noob's opinion.
I wish things were that cut and dry. However, TFs are an interesting variable to consider. I can't think of a time that I would ever leave them out of the equation. Reminds me of the saying, "The right thing at the wrong time is still a wrong thing.
To the OP: Mandelbrot invented an excellent theory. It does apply to markets. The reason it does is based upon the market's independent and dependent variables having the common characteristic of "granularity". The six basic crystaline forms also have this common characteristic and, therefore, crystals form in one of six ways consistently. It is possible to view structures on any scale as a consequence. As this is considered, then the relationship of different scales becomes apparent. Do not equate TF's with different fractal scales. You are reading many sources of information from different people. It is easy to discount and disregard their contributions and avoid the consequences they give you. That is what I do to stay spot on for creating the system of operation of the markets. Trends exist on all fractals. Trends have three parts: two ends and a middle. Both ends mark a trend's failure. Therefore, trends overlap. A market cycle is composed of two opposite trends. All of the above reveals that a trend must contain a number of moves. To have the interlocking nature of fractals in markets understood is not an easy undertaking. As you see no posters here are knowledgeable nor skilled. The key theories you need to use were created by just a few people: they are: Keynes ( paradigm theory); Carnap (logic theory); Boole (Algebra, base 2); and Mandelbrot (fractal theory). The theory of the market's operation is finite and complete. All parts have complete unique defined definitions and all may be expressed without any errors or omissions. The independent variable has a fixed number of cases and a consistent test to determine the case at hand. The dependent variable determines if and when the independent variable test is significant. Since trends have two ends and a middle, then there is a universal concept that is a consequence. This is that all sets of the system follow and order of events AND that the order is consistent and totally reliable at all times. There is a major limitation when using the system of opreation of the markets. The market has to have sufficient liquidity. Sufficient liquidity is attained when all participants can participate without delay in fulfilling their wants and needs, i, e., there is no sensing of hesitation nor stalling. Fluidity might be a description. Optimizing for effectiveness and efficiency is not being addressed so far in this thread. Observing the system of operation of a market is always possible by any human. Markets are slower than humans are. The full offer of the market is complex to understand simply because one account cannot be used to take the full offer of the market. To take the full offer of the market, all fractals must be used. Usually, knowledgable and skilled people recognize and understand seven interlocking fractals. These extend across the fastest tick level to the slowest Econometric level. Bull and Bear trends are within the Econometric level as a sub fractal. The paradigm of the sytem of operation of the market, has a complete Hypothesis Set (HS) and the corresponding Parametric Measure (PM). The measure is variable velocity. These were implied early in the 20th Century. By the time of the advent of mainframes in computers, the system of the market's operation had been made explicit. As intellectual property, there are few ways to maintain property rights in our present culture. So far the only transference that is possible is by doing a mind building process. The benefit of this constraint is that no one else can have it. It can be represented on less than ten regular sheets of paper using conventional representations. To transfer the system from one computer to another takes about a week (40 hours of programmer time) where a system technian goes through and corrects the remotely transferred original copy to make up for the failure of the currect web and network to handle a planned automated transfer. The comments from this point on are just provided as information that comes up along the way that may seem to be unexpected for one reason or other. 1. There are four subsets of trend types; two are incomplete and beyond the minimum complete trend is the drift trend (each of the last two have an odd number of moves always). Three types of sentiment turns exist. The OOE of the turns is completely defined. Turn types are defined as from>>to in expressions of dominance of sentiment. In the independent variable, the three types of sentiment turns are are paired with all possible volume End Effects. (See Modrian Table). 2, The key failure of those who use the independent variable to correctly determine the depenedent variable is that slugs of data are not usally differentiated into two piles: measurable and not measurable. About 100% of analysts, today, using all data, fail to set aside inelligible data slugs. Hence, the vapid performance of all large organizations (They earn money through commissions and fees as a result). 3. Usually all analysts use the wrong variable for analysis. This keeps the huge market's operation shielded from any effective and efficient considerations. The finance industry, thus, excuses itself from using Science in its operations. Here in ET I am considered to be of no import. This conventional judgement, yields the appropriate consequences. The content of what I post is of high density substantively speaking. A recommended test is to read it every six months. After ten or twelve years your viewpoint may shift a little.
Yes and no. You could have 20+ 15 min chart setups before 1 finishes on a daily chart. But you think that counting the ticks you would earn on daily, approximately evens out the result. If that is true you would have to always enter with the same position size ,meaning stop loss on a 15 min is 10 ticks and on a daily 150 ticks away but you have to enter with the same position. But if you are adjusting your position ,so that you always have the same risk % (on a above example, daily chart will have 15 times smaller position) then things change and 15 min chart is much more profitable.