There are 3 contexts of laws required in trading . The appropriate LAWS OF THINKING for trading, the appropriate LAWS OF FEELINGS for trading , and the appropriate LAWS OF ACTIONS for trading.
The Successful trading is based according to these three laws on
1) POWER OF COLLECTIVE SCIENTIFIC THINKING: A GREAT AND SIMPLE SCIENTIFIC PERCEPTION OF THE FUNCTION OF THE ECONOMY THROUGH SOME GLOBAL STATISTICAL LAW. E.g. The law of Universal attraction in economy: that big money attracts more big money in the capital markets, and this by the balance of demand and supply makes securities indexes of the companies , that are indeed the big money, to have mainly stable ascending trend, whenever one can observe such one. Valid statistical deductions can be obtained with simple statistical hypotheses tests about the existence or not of a trend, with sample size half the period of a dominating cycle). (STABLE GREAT SCIENTIFIC THOUGHT-FORM OR BELIEF FACTOR IN TRADING. )
2) POWER OF COLLECTIVE PSYCHOLOGY: A LINK WITH THE POSITIVE COLLECTIVE PSYCHOLOGY.(E.g. that the growth of security indexes also represent the optimism of the growth and success of real business of the involved companies. And we bet or trade only on the ascension of the index, whenever an ascending trend is observable). (STABLE GREAT POSITIVE COLLECTIVE EMOTIONAL OR PSYCHOLOGICAL FACTOR IN TRADING. )
3) POWER OF INDIVIDUALS SIMPLE , CONSISTENT AND EASY TO CONDUCT PRACTICE. (e.g. a trading system with about 80% success rate that utilizes essentially only one indicator in 3 time frames, simple risk management rules of stop loss, take profit, trailing and escalation, and time spent not more than 20 minutes per day. In this way there are not many opportunities of human errors in the conduction of the trading practice. Failed trades are attributed to the randomness and are not to blame the trader). (STABLE SIMPLE AND EASY PRACTICAL FACTOR IN TRADING)
We may make the metaphor that successful trading is the ability to have successful resonance with the activities of top minority of those who determine the markets.
In trading there are 3 components in the feelings that must be dealt with. 1) The feeling of MONEY itself, 2) The feeling of the UTILITY of the money 3) The feeling of the RISK of the money each time. What is called usually money management in trading is essentially RISK MANAGEMENT.
VALID STATISTICS AND PREDICTABILITY
We must make here some remarks about the robust application of statistical predictions in the capital markets.
1) The theory that the efficient markets and in particular that they follow a pure random walk is easy to refute with better statistical experiments and hypotheses tests. The random walk would fit to a market where the sizes of the economic organizations are uniformly random. But the reality is that they follow a Pareto or power distribution, therefore this is inherited in the distribution of the volumes of transactions and also in the emerging trends or drifts.
2) The statistical models of time series are more robust , when they apply to the entity MARKET as a whole and are better as non-parametric , and not when they apply to single stocks and are linear or parametric. The reasons is that a time series as a stochastic process , requires data of a sample of paths, and for a single stock is available only a single path. While for all the market the path of each stock or security is considered one path from the sample of all paths of all the stocks. Linear time series models or derived like ARMA, ARIMA, SARIMA etc are destined to fail for particular patterns like those described in the post 32, because the true equations are non-linear and in addition with random time varying coefficients that derive the random emergence of the 4 basic observable patterns (see post 32 ). In addition the standard application of the time series by the researchers, focuses on stationary time series after they extract a stable exponential trend, while in the reality the main concern should be the random path of the average value of the prices that shapes the patterns and is neither constant exponential trend neither zero ! The "statistical momentum conservation" (See post 10) might then be nothing else than an hypothesis that the random and time varying 1st order in time steps , partial correlation of the prices , is always positive. This can be easily tested statistically. E.g. in the cross exchange rate EURUSD but also in the indexes, the partial correlation of the current to the previous time step bar is measured indeed positive, in almost all time frames, except at the daily time frame, where the cyclic behavior prevails. In the daily time frame the partial correlation is negative , which means if one day is up the next day it is more probable that it is down. In addition, the cyclic behavior is even stronger in pairs of two days with negative partial correlation (two days up two days down etc).In searching for random cycles or periodicity, of say a single index or even instrument , the valid statistical practice requires the creation of a sample of paths over a time interval of a whole period, by collecting the pieces of the path at different periods as the market move as far as the searched periodicity is concerned may be considered as moving independently at independent periods.
The most essential tool for successful and profitable above the average, trading from the three that the title of the book suggests (Law of growth, law of cycles, law of inequalities) is the law of cycles and the awareness to discover cycles in the charts of prices, that are not directly apparent. Especially when the cycles are 1) Daily cycles to be traded with hourly or 4-hours bars and 2) Weekly cycles to be traded with hourly or 4-hours or daily bars 3) Monthly cycles to be traded with daily bars 4) Seasonal 3-months cycles to be traded with daily bars.
3) The less ambitious the statistical application the more valid the result. E.g. applying a statistical hypothesis test, or analysis of variance to test if there is an up or a down trend (drift) or none, is a more valid statistical deduction , than applying a linear model of a time series and requiring prediction of the next step price.
4) Multivariate statistics, like factor analysis, discriminant analysis , logistic regression, cluster analysis , goal programming etc are possible to utilize for a more detailed theory of predictability and of portfolio analysis, and sector analysis of the market and not only H. Markowitz theory.
5) In applying of the above applications of statistics, the researcher must have at first a very good "feeling" of the data, and should verify rather with statistics the result rather than discover it.
6) The "Pareto rule of complexity-results" also holds here. In other words with less than 20% of the complexity of the calculations is derived more than 80% of the deduction. The rest of the 20% requires more than 80% more complexity in the calculations.
The less hypothesis we use in applying statistical hypotheses, the better. That is why non-parametric statistics is better. An exception is our knowledge of the application of the Pareto distribution in various aspects of the market which we is parametric.
That is why we avoid applying very complicated with many hypotheses and time consuming to estimate models to forecast the markets, but we prefer to respond to the market, by measuring only in a valid statistical way, the average position of the price, and the channel around it, the velocity (trend, 1st derivative) and acceleration-deceleration (2nd derivative) of the prices.
The statistical quantities from the front-office in trading need to me measured are
1) the price position in the channel around the average, 2) the velocity (1st derivative) and
3) the acceleration-deceleration (2nd derivative), which is done as statistical quantities by a hypothesis test or confidence interval.
4) The support-resistance levels can be measured also by action-volume histograms. The measurements are done with convenient indicators, and can also define in a statistically valid way, not only , the channels , the trend, reversal, and
5) (Eliot) waves but also
6) the spikes.
7) It is required also an in advance in the past measurement and discovery of the basic stable cycles in the markets (see post 5)
8) An in advanced in the past measurement and discovery that trends duration and length, and volumes follow the Pareto distribution (see post 11,25 etc).
In addition for the back-office of trading we need to measure the
9) probability of success of trade based on the past history of trades, to apply the Kelly criterion, and also
10) the average rate of increase of the trading funds and
11) its variance again from the past history of trading.
The back-office statistical quantities in 9), 10), 11) are related , by simulation as e.g. in the simulator in post 43.
I have also created a simulator to experiment with different rules of withdrawals.
There are many who complain that the indicators have lag, and prefer not to use indicators at all, but only the prices. This is rather stupid! The indicators when are measuring a statistical quantity MUST have lag, because we are not interested for the price at the now only, but in a short-term past horizon too, which defines the statistical momentum which is statically conserved. It is not a race of speed, it is a challenge of successful perception. The science of statistics is the best for the moment one can have , from the collective scientific thinking, and collective consensus, and we must be honest and humble to admit is restricted abilities, but also trust , respect it and be confident for the success when applying it. When we are applying the statistical mode of thinking for the markets, we never run serious dangers of being "burned"and "busted" in our deductions, as statistics claims everything only up to some probability, or probability inequality and intervals.
There are some also that claim to have "intuitive guessing" about how the markets will move beyond the observable state of the markets. This of course cannot be included easily in the standard statistical inference. But it seems to me that sometimes, this is in certain human and social environmental conditions is too much to ask from yourself, and it may fire-back to systematic opposite to the actual markets moves guessing! I believe that fortunately pure statistical inference from the observable states of the market only, may be adequate for very profitable trading.
THE TOP 6 FACTORS OF ATTENTION IN MANUAL TRADING ARE
That is why we avoid applying very complicated with many hypotheses and time consuming to estimate models to forecast the markets, but we prefer to respond to the market, by measuring only in a valid statistical way, the average position of the price, and the channel around it, the velocity (trend, 1st derivative) and acceleration-deceleration (2nd derivative) of the prices.
The statistical quantities from the front-office in trading need to me measured are
1) the price position in the channel around the average, 2) the velocity (1st derivative) and
3) the acceleration-deceleration (2nd derivative), which is done as statistical quantities by a hypothesis test or confidence interval.
4) The support-resistance levels can be measured also by action-volume histograms. The measurements are done with convenient indicators, and can also define in a statistically valid way, not only , the channels , the trend, reversal, and
5) (Eliot) waves but also
6) the spikes.
7) It is required also an in advance in the past measurement and discovery of the basic stable cycles in the markets (see post 5)
8) An in advanced in the past measurement and discovery that trends duration and length, and volumes follow the Pareto distribution (see post 11,25 etc).
In addition for the back-office of trading we need to measure the
9) probability of success of trade based on the past history of trades, to apply the Kelly criterion, and also
10) the average rate of increase of the trading funds and
11) its variance again from the past history of trading.
The back-office statistical quantities in 9), 10), 11) are related , by simulation as e.g. in the simulator in post 43.
I have also created a simulator to experiment with different rules of withdrawals.
There are many who complain that the indicators have lag, and prefer not to use indicators at all, but only the prices. This is rather stupid! The indicators when are measuring a statistical quantity MUST have lag, because we are not interested for the price at the now only, but in a short-term past horizon too, which defines the statistical momentum which is statically conserved. It is not a race of speed, it is a challenge of successful perception. The science of statistics is the best for the moment one can have , from the collective scientific thinking, and collective consensus, and we must be honest and humble to admit is restricted abilities, but also trust , respect it and be confident for the success when applying it. When we are applying the statistical mode of thinking for the markets, we never run serious dangers of being "burned"and "busted" in our deductions, as statistics claims everything only up to some probability, or probability inequality and intervals.
There are some also that claim to have "intuitive guessing" about how the markets will move beyond the observable state of the markets. This of course cannot be included easily in the standard statistical inference. But it seems to me that sometimes, this is in certain human and social environmental conditions is too much to ask from yourself, and it may fire-back to systematic opposite to the actual markets moves guessing! I believe that fortunately pure statistical inference from the observable states of the market only, may be adequate for very profitable trading.
THE TOP 6 FACTORS OF ATTENTION IN MANUAL TRADING ARE
1) NEVER USE ALL YOUR FUNDS FOR TRADING. DIVIDE THEM TO TRADING AND NON-TRADING FUNDS BY THE RATIO f=R/a^2 RULE (see below for this ratio or in posts 3,13,33). THE DIVISION OF FUNDS AT EACH PERIOD IS ADJUSTED TO CONFORM WITH THIS PERCENTAGE RATIO. NEVER WITHDRAW PER PERIOD FROM THE NON-TRADING FUNDS MORE THAN HALF OF THE AVERAGE PROFITS OF THE TRADING FUNDS PER PERIOD. This division and adjustment of the funds has been applied for many years in buy and hold investments by professor Michael LeBoeuf.
2) THE ONLY CERTAINTY, WHILE TRADING IS ALSO OUR FIRST PRIORITY: WE MAY DETERMINE THAT OUR LOSSES AT EACH POSITION WILL NOT BE LARGER THAN A SPECIFIED PERCENTAGE DEFINED BY THE KELLY CRITERION (see posts 3, 13, 33)
3) FOCUS ON MACROSCOPIC INSTRUMENTS LIKE STOCK INDEXES WITH PERMANENT STRONG LONG TERM TREND, even if you want to trade at short time scales. (e.g. of the American Economy which is young and strong and indexes like Dow Jones, SnP500, Nasdaq etc).The statistical quantities from the front-office in trading need to me measured are the price position in the channel around the average, the velocity (1st derivative) and the acceleration-deceleration (2nd derivative), which is done as statistical quantities by a hypothesis test or confidence interval. The support-resistance levels can be measured also by action-volume histograms. The measurements are done with convenient indicators, and can also define in a statistically valid way, not only , the channels , the trend, reversal, and Eliot-waves but also the spikes. In addition for the back-office of trading we need to measure the probability of success of trade based on the past history of trades, to apply the Kelly criterion, and also the average rate of increase of the trading funds and its variance again from the past history of trading.
4) FOR VERY LOW RISK AT OPENING POSITIONS ON THE PREVIOUS INDEXES WITH PERMANENT STRONG TREND, OPEN BETTING UPWARDS, AT TERMINAL SPIKES AGAINST THE TREND. This is the Bill Williams technique.
5) READ THE NEWS AND FINANCIAL STATEMENTS BUT THE ASSESSMENT OF THE PATTERNS OF THE MARKET REQUIRES THAT IT IS DONE IN MANY SUCCESSIVE TIME FRAMES CHARTS. This is a basic recommendation by Alexander Elder, which, by now, it is a common knowledge to traders
6) BE FLEXIBLE IN RESPONDING TO THE MARKET AND DO NOT HESITATE TO FOLLOW PROMPTLY ANY UNEXPECTED CHANGES OF THE TREND OF THE MARKET. (This is the Bill Williams psychological "Holy Grail" of trading)
Probably the best instantaneous rewarding "why?", of manual trading is the joy and satisfaction in playing, among the situations of higher or lower uncertainty of what will happen in the global economy and markets, so as to plan and conduct a strategy that lets you know on occasions what will happen with acceptable low uncertainty.
We mentioned in the post 22 of the 3 basic modes of Demand-Supply and the resulting 4 basic price patterns
1) Spikes. 2) Trends, 3) Flat Waves (or curles, or the technical analysis triangles and rhombus) 4) Stationarities (ranging markets).
The 3 modes of the Demand-Supply (Domination, competition, cooperation) are consequences of the Law of polarity (in the 12 laws of the financial markets). Most of the spikes are produced by the coupling mode of competition in the Demand-Supply.
The price waves and patterns created by the demand-supply follow of course the laws of periodicity and universal attraction or economic inequality as far as period and amplitude is concerned. In other words all of the above 4 patterns may be derived as more complex combinations of the 3 basic laws of
1) Momentum conservation or trend (the first partial correlations is positive in an appropriate time scale)
2) The market moves in predetermined periodicity or cycles known in advance. (Cycles from 60-80 years, 22.2 years, 1 year, seasonal (3 months), monthly weekly, daily, and intra-day cycles).
3) In the very long run of many decades (at least larger than 11.1 years) the securities indexes have a constant growth or ascending trend.
The 3 modes of the Demand-Supply (Domination, competition, cooperation) are consequences of the Law of polarity (in the 12 laws of the financial markets). Most of the spikes are produced by the coupling mode of competition in the Demand-Supply.
The price waves and patterns created by the demand-supply follow of course the laws of periodicity and universal attraction or economic inequality as far as period and amplitude is concerned. In other words all of the above 4 patterns may be derived as more complex combinations of the 3 basic laws of
1) Momentum conservation or trend (the first partial correlations is positive in an appropriate time scale)
2) The market moves in predetermined periodicity or cycles known in advance. (Cycles from 60-80 years, 22.2 years, 1 year, seasonal (3 months), monthly weekly, daily, and intra-day cycles).
3) In the very long run of many decades (at least larger than 11.1 years) the securities indexes have a constant growth or ascending trend.
For example a currency pair (e.g. EURUSD) is already a coupling of two economies (US and EU) , and this coupling is mainly or is rotating among the above three modes: Domination, Competition, Cooperation. Different economies are inclined to dwell more on one mode than another, depending on the global fundamentals of the economies, their trade, and interaction.
The most usual pattern coming out from Domination is the waves. The most usual pattern coming out from competitions is spikes, and the most usual pattern coming out from cooperation is a trend. In the average (rounding numbers, and about at the hourly bars time-frame) , about 10% of the time the market produces spikes, about 20% of the time the market produces trends, about 30% of the time the market produces waves and about 40% of the time the market produces stationarities (ranging or no pattern at all). Of course these numbers get different if we shift to shorter time scales (more stationarities/flat-waves) and different if we shift at larger time scales (more trends).
Also the transition probabilities that the 4 patterns succeed each other are not equal. After a stationarity or a wave pattern it is more probable that a spike or trend will follow. Trends will start and end usually with spikes (initial-terminal spikes). If a trend will end with a terminal spike, it is highly probable that a reaction initial spike will occur for a new trend.
We may classify these patterns in two categories a) The horizontals or volatility short
that include the waves and the stationarities b) The verticals or volatility long that include the trends and spikes. The transition probabilities from the horizontals to the verticals is higher than the average. The horizontal patterns are utilized so as to take the vertical patterns from their early start. In particular the skew waves (skew triangles) indicated directly which direction will be the coming spike or trend, and are thus as valuable as the information from acceleration or divergence which is prior to a trend-start. The horizontal patterns are also related with clear and strong support-resistance levels. One of the best trading systems is to utilize the skew-waves ( triangles) as information for a coming breakout of a support-resistance level, and combine it with a periodicity, e.g. horizontals that shape daily when all sessions are closed in forex.
The full qualitative mathematical definitions of domination, competition, and cooperation as well as the proofs that only the above 4 price patterns are the possible solutions is outside the scope of a rather practical Blog like this.
All the 4 trading patterns are tradable.
1) The spikes are traded as (reversal or continuation) break-outs of previous trends, waves, or stationarities. We anticipate the reversal break-out from the deceleration of the momentum of the previous pattern. Thus such break-outs are essentially counter-trend trading and prior to the spike. But there is also a post-trading after the spike which is safer. A way to anticipate if a spike will reverse to an anti-spike or it will continue is to analyze if the spike is an ending-spike or exhaustion spike of a previous trend (in which case it will reverse to an anti-spike) or if it is initial-spike that initiates a new trend. Similarly spikes inside a flat channel are called inner-spikes and more often they reverse to counter-spikes, or external to the flat channel in which are called outer-spikes and more often the initiate a trend and continue. In both cases we may apply the 3rd-wave trading rule, therefore avoid trading the spike itself but essentially trade what comes after the spike. (see below). Because the 1st wave here is the spike, (and if it is an initial spike that it will continue), the 2nd wave or retrace is almost flat (and in the average it lasts double the duration of the spike, and its retrace can be predicted with the Babson median line) , therefore we trade the 3rd wave (continuation of the spike) with a stop-entry just very little away from the extreme of the spike. In this way we prefer the spike only because it is easier to detect it among other patterns, and because of its high momentum, which results also in to higher potential profits. During the trading we may pyramid. We exit by trailing or deceleration of the spike momentum. The sub-rule buy at a support level, sell at a resistance level still holds here, if we define appropriately support-resistance levels.
Approximate equation for this patterns is the p(t)=exp(a(t)t)+e(t) . Where a(t) is the intensity of the growth and e(t) is the random disturbance term and of course p(t) is the price while t is the time.
2) The trends when they have sub-waves (called here expeditions), are traded by their internal (Elliot) sub-waves, where we trade only the sub-waves (vectors) that go along the trend. They are the longer. Those that go against the trend are the shorter. We do not trade the shorter. The subwaves according to their order of appearence either along or against the trend are named here as 1) Lighning (usually a spike) , 2) Thunder, 3) Blow, 4) Wave4, 5) Wave5 etc. Actually we pyramid at the start of the subwave Blow, then sub-wave4, then subwave6 etc or at the break-out of these subwaves from the forward extremes of the previous longer subwave. If there are no sub-waves at all (called here excursions, or excursions are called the subwaves of a trend while the trend itself in such a case is called expedition) , then they are traded simply by open-and-hold (and below we will improve the trading with pyramiding and adjusting of position size inside the trade and between support-resistance lines. Of course there is also the coarser pyramiding, only at the support-resistance levels, that is only at the start of the subwaves) . We trail by the shorter sub-waves height and exit also on the appearence of deceleration (usually divergence) of the momentum of the trend. If we accept the concept of non-horizontal skew supports-resistance lines then: The sub-rule buy at a support line, sell at a resistance line still holds here.
Approximate equations for this patterns is the
1) p(t)=exp(a(t)t)+e(t) . Where a(t) is the intensity of the growth and e(t) is the random disturbance term.
2) p(t)=A(t)*Cos(b(t)t+c(t))+exp(a(t)t)+f(t)*t+d(t)+e(t) . Where A(t) is the amplitude of the wave , b(t) defines the frequency of the wave, the c(t) the phase and the a(t) and f(t) are exponential and linear intensities of the growth, d(t) the level , and e(t) is the random disturbance term.
3) The flat-waves (curles or the technical analysis triangles and rhombus) are traded as 1st preference, externally by (continuation or reversal) break-outs on their boundaries. As 2nd preference they are traded internally by their sub-waves in only one direction or two directions and when they are not of expanding or diminishing amplitude (triangles) only the middle 1/3 as the trading of the flat of constant amplitude waves. The deceleration of their momentum is again the anticipation of the break-out. The break-out will lead usually to a spike (or trend) which traded as above. The diminishing triangles are preferred to be traded externally rather than internally.
When a flat channel reduces in width, thus suggesting a diminishing triangle, it is more likely that a new trend will start, compared to a flat channel that keeps its width constant. While the expanding triangles may be traded internally too. Also if the flat wave pattern is after a spike, we anticipate that it will have sufficient long duration and we may trade it internally too. But if it is at a reversal we anticipate that it will not last for long, so we prefer to trade it externally only. In other words if we anticipate that it will last for long we trade it internally but if we anticipate that it will not last for long we trade it externally. It is important to notice that because flat channels change easier than trending channels, the internal trading of a flat channel is the last priority among modes of trading of the patterns and somehow worse than the internal trading o a trending channel. But the external trading of a flat channel is a better priority than the internal trading of a trending channel, because it is a pro-active setup relative to its anticipated break-out. . The sub-rule buy at a support line, sell at a resistance line still holds here.
Approximate equations for these patterns are the
1) p(t)=A(t)*Cos(b(t)t+c(t))+d(t)+e(t) . Where A(t) is the amplitude of the wave , b(t) defines the frequency of the wave, the c(t) the phase, d(t) defines the level of he prices and e(t) is the random disturbance term.
2) p(t)=exp(a(t)t)*Cos(b(t)t+c(t))+d(t)+e(t) . Where exp(a(t)t) is the expanding or dumping amplitude of the wave , b(t) defines the frequency of the wave, the c(t) the phase , d(t) defines the level of he prices and e(t) is the random disturbance term.
4) The stationarities (ranging market) are traded almost as the waves in two modes internally and externally. Internally: We open 1/3 away from the middle line and we close almost at the middle line. We also trade them as the constant applitude waves, externally, anticipating break-out usually to a spike (or trend).
Stationarities that are after a spike may be traded internally , because we anticipate that they will have sufficient duration. But if they are at reversals we anticipate no sufficient long duration and we prefer to trade them externally only (break outs). In other words if we anticipate that it will last for long we trade it internally but if we anticipate that it will not last for long we trade it externally. Here the optimal adjustment algorithm on the position size applies, but not pyramiding. The sub-rule buy at a support line, sell at a resistance line still holds here.
Approximate equation for this patterns is the
1) p(t)=d(t)+e(t) . Where d(t) are linear level and e(t) is the random disturbance term.
The most predictable effect modulated by cycles (see post 5) is the reaction to an super-exponential moves (a blow-up at the end of trend in the form of super-exponential move or terminal spike). (See e.g. https://www.ted.com/talks/didier_sornette_how_we_can_predict_the_next_financial_crisis and http://www.er.ethz.ch/ Such super-exponential terminal patterns of trend may occur usually as result of overgrowth of the one of the two populations in a demand-supply coupling rather that of domination and not so much of competition or cooperation. See also post 22. .The frequency of emergence and the size of such super-exponential blow-ups follows the law of inequalities in other words the Pareto or Log-normal distribution and is thus by fat more often than pure randomness would predict!).
THE HIDDEN SINGLE PATTERN TO ALL THE ABOVE 4 PRICE PATTERNS IS THE SUPER-BUBBLE!
In other words a single move (which is contained to all the above patterns) but very strong and with accelerating momentum. THEN THE PREDICTED NEXT MOVE IS ITS REACTION
Here are some pictures and examples of super-bubbles as detected by the alligator indicator, from the book by Bill Williams "Trading Chaos" 2nd edition, chapter 9. The superexponetial-bubble is the simplifying "eye" that turns the chaos of price movements in to sequence of randomly occurring simple pattern (that of super-bubble) which thus allows for randomly intermittent predictability. The average waiting time for a super-bubble to occur is usually larger than the duration of the sum-total of the super-bubble and its reaction. But of course if searched among 10-15 instruments such super-bubbles occur one after the other.
The super-bubble is of higher predictable reaction, if the previous longer time interval , the market had already a slower trend in the same direction as the super-bubble (so that the spike-super-bubble is an terminal or exhaustion move!)
https://www.scribd.com/doc/177005696/Bill-Williams-Trading-Chaos-Second-Edition
Finally here is a ted.com video by Didier Sornette that explains why super-bubbles are unique, universal but predictable behavior of the markets.
https://www.ted.com/talks/didier_sornette_how_we_can_predict_the_next_financial_crisis
The Log-Periodic Power Law Singularity (LPPLS) applies as the bubble reaches it point of crashing. In other words faster and faster waves appear with smaller amplitudes. (
see http://www2.math.su.se/matstat/reports/serieb/2009/rep7/report.pdf
https://warwick.ac.uk/fac/sci/maths/research/events/2013-2014/statmech/ght/programme/sornette_2.pdf
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0165819
https://www.sciencedirect.com/science/article/pii/S187538921000249X
https://en.wikipedia.org/wiki/Didier_Sornette
)
In order to conduct successfully an intra-day system of transactions , that is successful in the long run and easy to keep on applying it the next points must be met.
1) It must be relatively utterly simple! Only the "eye of simplicity"can put order and tame the chaos of intra-day price patterns! It must be manual and not automated!
2) Therefore it has to be one only pattern among the 4 price patterns (see post 32)
3) To deal with this one only pattern, we may apply simplifiers like , velocity or rate of change of prices, acceleration, support-resistance.
4) Celestial periodicity will give the long-run stability, but it need not be one only frequency or period but a few neighboring frequencies or periods in the spectrum of celestial frequencies or cycles.
5) But most of all the strongest simplifier is that , when measuring the velocity or rate of change , by a stratified sampling hypothesis test, then it has to be an extreme value , which will indicate a reaction or closing of the cycle. This in particular means that we entirely avoid the parametric predictive models of econometry that assume predictability at every time step, as for such to be succsesful they would need to be pod stochastoc coeficients and there practically no such econometric models, and we resort to the more robust and with less assumptions non-parametric statistics and in particular of a single non-parametric measurement of the velocity of the prices, with stratified sampling.
6) It must be a phenomenon tested scientifically with valid quantitative procedures , with sufficient good (intermittent) predictability , for many years.
7) The financial result should be adequate (e.g. >= 1MDS).
8) The financial result, in my case, is to be used not only for economic freedom, but also for a worthy goal e.g. so as to finance my innovative research in the new millennium digital mathematics.
9) The solution to all of the above 8 points leads to one only system: The (solar spikes) super-bubbles system:
(See end of post 44)
<2 change="" font="" is="" not="" of="" prices="" rate="" the="" zero.="">
<2 change="" font="" is="" not="" of="" prices="" rate="" the="" zero.="">THE SUPER-BUBBLES DO NOT HAVE ONLY THE EASIEST PATTERN RECOGNITION (E.G. WITH PROGRAMMABLE STATISTICAL HYPOTHESIS TEST) BUT ALSO THE HIGHEST PREDICTABILITY OF NEXT EVOLUTION, AMONG THE OTHER MOE COMPLICATED PATTERNS.
THE OVER ALL STATISTICAL BEHAVIOR OF PRICE PATTERNS
We notice that although the patterns are essentially 4 categories in details there are 6 distinct statistical patterns of the random or statistical posItion, velocity and acceleration of the prices.
A class of stochastic processes can be defined as the behavior of the markets based on these 6 basic patterns P1, P2, P3, P4, P5, P5, P7. The Pi i=1,2,3,4,5,6,7 are essentially random statistical patterns with random and variable parameters of size duration, and relative analogies that define them. E.g. we may have a Pareto distribution of the duration and price height of the patterns because of the inequalities in the markets. The 7th pattern P7 of stationary behavior we may call intermittency pattern. We may then assume a class of Markov processes with random and variable transition probabilities, where each random type of pattern occurs and then a next one occurs. Bu the transition probabilities are not arbitrary! E.g. spikes occur usually at the begging (initial spikes) and the end (terminal spikes) of trends, up and down trends with stationary channels in between them, shape cycles, that in the average of stable period, and some times of fixed beginning and end.
The probability that a type of pattern occurs changes also according to the time scale. In 2, 5 or more years annual bars time frames, the non-waving trend pattern is dominating, while say in 5-minutes bars time-frame flat patterns are dominating.
We notice that, by utilizing only pattern P3, of non-waving trend, and intermittency P7 we may derive all, other patterns with appropriate patterns of transition probabilities of the P3! I have coded a simulator of such a class of stochastic processes, superimposed on many time frames called Multi-time scales Rainbow Walks stochastic processes.
This is the overall behavior of patterns of the markets, and there are some invariant properties like
1) Cyclic behavior as alternation with up and down trend patterns with flat channels at the bottoms and tops
2) Statistical momentum conservation (see post 10) where the 1st time-step partial correlation of a price is almost always positive,
3) A Pareto distribution of the duration and height of the patterns, due to the inequalities of the enterprises in the economic system (see post 10, 25, 57,63 )
We notice that a common sub-rule for all the 4 price patterns is the
Buy at a support line, sell at a resistance line.
The support-resistance levels are the poles on which the 3 modes of a) Domination b) Competition c) Cooperation (see post 22) , start and end so as to form the 4 price patterns. Therefore any system based on the support-resistance levels is very basic, fundamentally simple and statistically successful. One of the best ways to detect the support-resistance levels (poles or attractors) is through the mid-levels of the reversals or the continuation or stationary patterns(waving, diminishing, expanding or constant amplitude. The best tool here is the statistical, e.g. price-activity histograms).
SUPPORT-RESISTANCE TRADING METHODS In particular a very simple and high probability of success trading strategy with support-resistance is the Single Support-Resistance System. This system is meaningful for cross-rates like EURUSD or cross-rates of currencies closely correlated e.g. in the same continent etc where the balance and stationarity of the market is more important and often rather than any continuing trend. Because in this system we bet for the stationarity-stability of the market rather than any kind of continuing growth. According to this method, we buy and sell simultaneously, AT a single (statistical or maxima-minima, or decimal xx00, xx50, or better mid-level of a continuation or reversal pattern or boundary of a flat channel, etc) support-resistance level. Then the statistical behavior of the market is that with high probability it will hit the support-resistance it will deviate a little and will return to hit it again. When it deviates, we close the position-direction with profit (putting also a reasonable stop-loss, e.g. equal to the targeted profit, to the other part of the position) , and when it will return, (even not completely to hit the support-resistance) we close the losing position-direction. The total is a small profit, which nevertheless it happens with high probability and very often. A variation of it, is the double or ending support-resistance system, in which we close the winning position only if it reaches a next support-resistance, so that we have time for the losing position to deviate or retrace. And in this variation we may open the positions at a support-resistance which is convenient as the market stays there for long, or anywhere (e.g. after a spike even if it seems ending ) , as long as we end the positions on a support-resistance. Because it it independent from where we open the pair o positions, but it does matter how we close around a support-resistance. As a support-resistance is also a statistical one, in other words if we calculate online-real time the histogram of price activity, of how often the price is found in particular price-bin (pip or tick) , the support resistance will be a (local) maximum of the histogram, we have also a mathematical proof why such a system would be winning when repeating it. (Here the conservation in adequate time duration of zero-momentum of the market at a support-resistance is required for the proof).
The CORRECTIVE ESCALATION from a support-resistance level, trading method
According to this method it is not so important the pattern recognition method and the pattern recognition in other time frames, or the focus on the 3rd wave in the current time frame, as it is the corrective adjustment and escalation of the position. Less than 20% of the success is in the initial high probability forecasting based on charts, technical analysis, indicators, and pattern recognition, and more than 80% of the success is on the corrective escalation technique after the Kelly criterion. We may start at a support-resistance level, (which may be a continuation pattern at its channel boundary or mid level) and we open position at the most probably direction of little size only, if we do not predict well and the prices go to the opposite direction we increase and open in the opposite direction , and we repeat it so as to have always a net position size towards where the prices are going. We do not use stop-loss as the role of the stop loss is the level where we open the opposite position. When the market breaks out finally to a direction we escalate (pyramid) and increase further the position. Typical sequences of position sizes are (1,-2,3, 5,4,3,2) or (-1,2,-3,5,4,3,2 ) We close by trailing or to the next support resistance level. Or we close at the predicted (by Babson median) end of a wave-vector between the 1st and 3rd (in Elliot counting).The starting leverage is usually 1 and increases to 2, 3 etc. While the percentages f of risked funds, start with 1% and in the average is is safe if it is at the 5%-6%. In general if p1 is the probability to win a trade, and p2 the probability to lose it (p1+p2=1), and the reward-to-risk ratio is equal to b, then the percentage f of funds risked in the trade should be about f=(b*p1-p2)/b. This is the Kelly theorem (see http://en.wikipedia.org/wiki/Kelly_criterion ). Nevertheless we prefer to risk in the average only 5%-10%, of what the Kelly formula suggests, that is about 2%, and the reason is that after simulations (see post 43), the paths with 2% exposure, are much more smooth and psychologically bearable in their volatility and draw-downs, compared to the exposure suggested by the Kelly formula! Let us say that with pattern recognition we have a success rate of 65% and the reward-to-risk ratio is 70% , then by the Kelly formula we must not exceed risking (0.7*0.65-0.35)/0.7=15% If it was a pattern recognition with 70% success rate (something very difficult) then the Kelly exposure would be 39% ! With the CORRECTIVE ESCALATION let us say that we start say with risking 1%, and then we correct in the opposite direction with 2% and then we re-correct in the initial direction with 5% (in total 1%-2%+5%=4% in the correct direction or -1%+2%-3%+5%=3%),then we continue pyramiding with 4%, then 3%, then 2% and no more. Thus in total 4%+4%+3%+2%=13% thus less than the Kelly exposure.Such sequences of trades are called excursions, and permit the definition of hierarchical sampling, where while at the level of trades the success rate may be say 55% at the sample layer of excursions it may be 75% or higher. With the CORRECTIVE ESCALATION we have the chance, even if the probability of an up or down move to be 50% to increase the success rate of the trades, to 70% - 75% with the above sequence (given of course the average momentum conservation).
The above show why trading based on invariants like support-resistance lines is probably the most powerful method.
We may summarize the trading of the 3 patterns (flat channel, trending channel, spike) as follows
1) There is external trading of channels and there is internal trading of channels. External trading of flat channels is better (of lower risk and higher performance) than internal trading, only at the end of the flat diminishing (contracting triangle) channel , when it is ready for a breakout to a spike. Otherwise the 3rd wave internal trading of flat channels is better.
2) The internal trading is of flat channels and of trending channels. The internal trading of trending channels is considered better, than of flat channels.
3) The external trading of flat channels, is of flat channels after spikes or not. External trading of flat diminishing channels before spikes is considered the best opportunity, for spikes as it is pro-active.
Also expanding flat channels are the worse for external trading, the best (among flat channels) for internal trading. Contracting flat channels, are the worse for internal trading, but good for external trading.
A tree of questions to make a pattern recognition and make trading decisions is the next.
1) Is it a) a flat-channel? (constant width or diminishing, expanding width)
b) a trending waving channel?
c) a spike?
2) If it is a spike, is initial or terminal? Internal or external? (if terminal / internal we prepare for re-action trading, if initial/external we prepare for 3rd wave continuation)
3) If it is a trending waving-channel, are we on the start of its 3rd wave? Or at the start of its (2n+1)-wave? (for internal one-sided trading)
4) If it is a flat-channel, are we on the start of its 3rd wave? (for internal trading). If not is the channel diminishing or expanding leading to a new trend? (for external trading).
5) Check the acceptable background time-frames T1, T2, T3 (etc) to find a tonal pattern there. We are interested even better in particular for a spike or 3rd wave tonal pattern. as this has high predictability and low risk. Then rank the T0 (focal) time frame pattern in relation to the direction/phase/timing suggested by the tonal pattern and intermediate background patterns.
The global multi-time-frame filtering may let hidden probabilities to be better predicted on the patterns of the focal time-frame especially if the tonal pattern is a spike or 3rd wave, like the next
1) If a trending channel will more probably decelerate and stop trending, or even reverse.
2) If a flat channel will more probably breakout up or down.
3) If a spike more probably will react to a counter-spike or continue.
After answering the above 5 questions, we then chose among the next four modes of trading
1) Trade only spikes, posterior with the 3rd-wave rule
2) Trade only with the 3rd-wave rule (this includes trend-channels, continuation flat-channels and reversal patterns).
3) Trade internally with the nth-wave for flat channels, or one-sided with (2n+1) wave for trending channels
4) Trade all the above and also externally, at the end of continuation diminishing flat-channels.
The choice of any of the above modes, which are in order from lower risk to higher risk, is also very much depending on having found among the background time-frames a tonal pattern which is spike or 3rd wave, as this has high predictability and low risk.
A standard way to apply the pattern-trading, is to choose a simple indicator, at the focal time frame which gives many signals, called the beat of the trading, but then execute only those signals that are in congruence with the pattern rules both at the focal time frame but also to the background time frames, and in particular the tonal-background time frame. The beat of the trading usually depicts the inner waves of flat or trending channels at the focal time frame. This will create the necessary intermittency and Pareto 20-80 rule on the opportunities. For example in the daily time frame or the 1-minute bars time frame we may choose the cross of moving average of 2 bars and 5 bars, or the force-index of 2 bars, or the Bollinger bands of 5 bars, and then select signals according to the daily, weekly and monthly time frame.
Classical technical analysis patterns
1) The classical technical analysis of continuation patterns are the flat channel of constant amplitude and the flat channel of diminishing amplitude (triangle) or flat channel expanding amplitude, when before them was a trending pattern. In other words are composite patterns based on the above enumeration of the basic patterns. They are ideal for external trading. They are usually after trends without channel or spikes and most often continue the price movement rather than reversing it. The flat channel of increasing amplitude, is supposed to breakout as reversal rather than as continuation (and in the securities and commodities market usually breaks-out downwards).
2) The reversal patterns of classical technical analysis are very short-term rather flat-like channels that reverse a channeled trend to a new opposite direction channeled trend. But a simple way to detect and predict a reversal is the next: When after a trend, the prices are stationary, and the channel pattern is not one of the constant width flat-channel, or triangle-channel , then it is a reversal pattern, and the prices will reverse trend. They are ideal for internal trading of the 3rd wave towards the new trend (Blow vector) (see 3rd wave trading rule below) .
The internal trading of a flat channel can be considered a trading between two parallel support-resistance levels. One of the best ways to do it , is to open a double neutral position (buy and sell of the same size), somewhere in the mid of the channel [and usually immediately after we detect a reversal or continuation or stationary pattern ], and close the winning position after a reflection on any of the two boundaries (support-resistance), while close the losing, at the opposite boundary, or anywhere the profit is positive and after at least 1/3 of the width of the channel. This type of the initial neutral double position trading, can result in to external trading too, if the prices breakout from one of the boundaries. In that case we close the losing position, and we go one with the winning at least till we break-even. The advantage is that we get a trading no matter which direction the market chooses, and also psychologically is appealing for a start of the trading. We call it the initially neutral parallel support-resistance levels trading.
Spikes Trading:
3rd wave (selective to opportunities) trading rule:
Among the various ways to trade the above patterns, we highlight the the 3rd wave towards the new trend (Blow vector) which a very selective and safe trading method among the patterns. In this mode of trading, we focus on waving patterns only (flat, continuation, or reversal, or waving-trend), and we want to take the waving pattern from its start. The best way to do it, is to detect the 1st wave (called Lightning) of the channel (usually a spike if its a reversal and a new waving channel is forming), then wait for the retrace of the 1st wave , that is the 2nd wave (called thunder), and enter, at the beginning of the 3rd-wave (called Blow). We may or may nor pyramid along the 3rd wave. The expected level that the 3rd-wave will reach is calculated with the median line of Babson. Since the channel (flat or trending) requires at least two waves to be shaped, we realize that trading the 3rd wave is essentially trading the earliest possible wave after the channel has been shaped.
This method, if the 1st wave is a spike (essentially initial and external) is also a spike-trading method. And as we may accept skew support-resistance likes, and the 2nd-wave starts and ends in the support-resistance or boundaries of the channel, it is also a support-resistance-to-support-resistance method. We may notice that this method applies not only to waving trending channels, but also to flat waving channels! And it may also apply to non-waving trends, if we look to the larger time-frame and analyze if, there, the non-waving trend is the 3rd wave. So we realize that the selective under risk-management 3rd wave rule can essentially trade all patterns.
3rd waves are detected mainly at changes of the market patterns. There are mainly 4 types of changes (another term could be mutations)
1) A channel (waving) trend changes in to a flat channel (continuation pattern)
2) A channel (waving) trend changes on to a reverse channel (waving) trend (reversal pattern)
3) A flat channel changes in to a channel (waving) trend.(Break out trend starting)
4) A flat channel changes in to a flat channel , higher or lower from the previous.(Break out flat channel starting)
All changes occur with a 1st wave (very often spike) , and we may wait for the 3rd wave.
Here are some videos with details of how to trade the 6 price patterns
http://www.youtube.com/watch?v=1ByzZ6b4ET8
http://www.youtube.com/watch?v=A49xzKYGyg4
http://www.youtube.com/watch?v=6YZ4ORz-UJ0
http://www.youtube.com/watch?v=hyELHtxTp7I
http://www.youtube.com/watch?v=4bIiQyfqW2U
http://www.youtube.com/watch?v=SqCFhhou3SM
http://www.youtube.com/watch?v=TzqDzQwPJnE
http://www.youtube.com/watch?v=_0-SJNnnHFM
http://www.youtube.com/watch?v=h8BP24tDJhA
http://www.youtube.com/watch?v=cf80pXwlda0
http://www.youtube.com/watch?v=J5fc25zRkWI
http://www.youtube.com/watch?v=V1VdCj57Zas
Even if a market would be a sequence of spikes alternating with flat continuation channels (that is a market without trends) , with 50% probability of the next spike up or down (a random walk of spikes) , still an external trading of the breakouts of the flat intermediate channels would be a very profitable trading.
For both the flat-wave pattern and the stationarity pattern, works also a trading mode based on the Maximum Likelihood Price-Level principle. In other words as the mid-level of the channel is also as a "center-of-gravity" of the prices , it holds that the probability is maximum for the price to be there. Therefore it is optimal to open positions say 1/3 of the channel width around and away from the mid-level and close the position at the mid-level. In this way the success rate of the trades is maximized. (see also post 24)
5) We may include also as pattern the no-pattern at all, that is a case not falling in to any of the above 4 cases. Such no-patterns are not traded al all.
Sometimes trading the Spikes and Trends is called trend-following, while trading the wave-patterns and stationarities is called counter-trend trading (and scalping, if it is in fast time frames like 1 hour , 5 minutes etc).
Other times couter-trend trading is also called the trading of trends where we enter in a reversal a bit earlier when the deceleration and divergence appears, before a clear break-out to the new trend.
There are trading systems specialized to one only or some only of the trading patterns. E.g. the famous 50 years old, turtle trading system, trades trends (excursions and expeditions) and spikes only. Or e.g. scalpers trade mainly the ranging and waving patterns.
There is a rule of time-scales and the 4 patterns: The larger the time scale the more often and the clearer the trend without noise. So directional (only up or only down) trading is the best. The shorter the time scale, the more often and the clearer the ranging pattern. So bi-directional (both up and down) trading is the best. Spikes and waves appear clearer in special intermediate time scales, like 48-hours, or one month. This is a consequence of the law of attraction , the power law of economic inequalities and the power distribution of the trends and volumes,
These 4-price patterns create an optimal portfolio of 4 trading strategies specialized to trade respectively the 4 price patterns. The largest percentage of the portfolio, is allocated to the strategy that trades the spikes, less percentage is allocated to the strategy that trades usual trends, less percentage is allocated to the strategy that trades waves, and finally the least percentage is allocated to the strategy that trades stationarities. happily because each patterns excludes the occurrence of the other patterns, the same funds are used each time , and the percentages have the meaning of exposure, leverage, and percentage of funds risked each time.Notice that the more rare the pattern, the higher the exposure and capitalization speed and the lower the risk because the longer the time that we are outside the market. It may seem strange to the common sense, but the highest speed of capitalization is concentrating on spikes where also is the longest time that you are outside of trading.Really God's speed in capitalization.
We remind that as a summary the ways that optimal portfolio theory can apply in trading is through at least 3 perspectives
1) Optimal portfolio of instruments ( post 14)
2) Optimal portfolio of strategies on the 4 basic price patterns (post 32)
3) Optimal portfolio of time scales or characteristic frequencies. (post 37)
So we could state a Pareto-rule (20-80 rule) for trading the 4 price patterns:
"More than 80% of the profits are obtained from the trading of less than 20% of the opportunities among the 4-price patterns"
And obviously the best profits are obtained from the rare spikes , and the 3rd wave trading rule, which is a local rule to apply the Pareto 20%-80% rule of filtering opportunities . The global filtering rule is of course the filtering based on many background time-frames and patterns on them. The way to do is to search among the accepted background patterns at what time-frame the pattern is most clear and strong, and considered it as tonal pattern. Obviously half only of the opportunities at the focal time frame are accepted, those that agree in direction timing and phase with the tonal pattern. But we may also rank all acceptable opportunities of the focal time-frame of interest according to if or not the intermediate patterns from the total till the focal are favorably agreeing in direction, phase and timing. In this way much less of half the opportunities of the patterns of the focal time-frame are selected to trade. In this way by filtering locally with the 3rd wave rule and globally with the nested time frames after the tonal time-frame, the success-rate increases from say 52% to 80% or more!
It seems simple to describe or draw the above 4 price patterns and the trading of them. But any 100% automation of them (in the present state of the art of software and platforms 2011) with a robot or EA, would be unsatisfactory. It would be inadequate, even if it was done at the daily focal frequency (hourly time frame) or monthly focal frequency (daily timeframe) where there is low noise, or even if instead of indicators it was utilized readily available pattern recognition algorithms. Where the coding fails is at the stochastic pattern recognition. Most of the current examples (2011) of such automations succeed to have 50% only performance of the corresponding manual execution performance. Human senses inspection and pattern recognitions is always better, and it is adequate and satisfactory. The (manual) Bill Williams trading, is an example (described in 3 of his books) where over daily bars and by trading partly only and few only of the cases of the above 4-patterns, for the last 50 years (with 15 minutes only once per day inspection)he succeeded in having after the leverage an average annual 300% ( or 25% monthly) rate of return. Other example of such trading over daily bars, is the commodex protocol (which contrary to Bill Williams method, is not fully revealed, rather automated, and is sold only as signals, by Philip Gotthelf) which the last 50 years has an average annual rate of return of 130% (or 10.8% monthly).
The pyramiding is theoretically justifiable as optimal in the trading, after the Pareto duration of the above 4 patterns.
The trading systems can be classified also according the various combinations and sequences of the 4 price patterns (e.g. Trend->Trend reversed, Trend->stationarity, Spike->trend, Trend->stationarity->trend etc). In addition if we inspect 2 different time-frames one focal, and one background, then the classification, is of pairs of the above combinations. I have tried it, but it becomes quite complicated. The pairs or triples of the 4 patterns in a single focal time-frame seem to me adequate.
Also the transition probabilities that the 4 patterns succeed each other are not equal. After a stationarity or a wave pattern it is more probable that a spike or trend will follow. Trends will start and end usually with spikes (initial-terminal spikes). If a trend will end with a terminal spike, it is highly probable that a reaction initial spike will occur for a new trend.
We may classify these patterns in two categories a) The horizontals or volatility short
that include the waves and the stationarities b) The verticals or volatility long that include the trends and spikes. The transition probabilities from the horizontals to the verticals is higher than the average. The horizontal patterns are utilized so as to take the vertical patterns from their early start. In particular the skew waves (skew triangles) indicated directly which direction will be the coming spike or trend, and are thus as valuable as the information from acceleration or divergence which is prior to a trend-start. The horizontal patterns are also related with clear and strong support-resistance levels. One of the best trading systems is to utilize the skew-waves ( triangles) as information for a coming breakout of a support-resistance level, and combine it with a periodicity, e.g. horizontals that shape daily when all sessions are closed in forex.
Larger economies , like US, and EU are more likely in competition and cooperation modes of coupling. Therefore spikes, and trends are more often. A large and a small economy, e.g. US dollar and Swedish Korona, or US dollar and Australian dollar, are more likely in domination mode of coupling, thus the waves pattern would be more often than normally. The larger the economy, the smaller the volatility of its economic indexes, and measures, the smaller the economy, the larger the volatility of its economic indexes, and measures. The exact relation follows again a Pareto (or a power) distribution.
The full qualitative mathematical definitions of domination, competition, and cooperation as well as the proofs that only the above 4 price patterns are the possible solutions is outside the scope of a rather practical Blog like this.
All the 4 trading patterns are tradable.
1) The spikes are traded as (reversal or continuation) break-outs of previous trends, waves, or stationarities. We anticipate the reversal break-out from the deceleration of the momentum of the previous pattern. Thus such break-outs are essentially counter-trend trading and prior to the spike. But there is also a post-trading after the spike which is safer. A way to anticipate if a spike will reverse to an anti-spike or it will continue is to analyze if the spike is an ending-spike or exhaustion spike of a previous trend (in which case it will reverse to an anti-spike) or if it is initial-spike that initiates a new trend. Similarly spikes inside a flat channel are called inner-spikes and more often they reverse to counter-spikes, or external to the flat channel in which are called outer-spikes and more often the initiate a trend and continue. In both cases we may apply the 3rd-wave trading rule, therefore avoid trading the spike itself but essentially trade what comes after the spike. (see below). Because the 1st wave here is the spike, (and if it is an initial spike that it will continue), the 2nd wave or retrace is almost flat (and in the average it lasts double the duration of the spike, and its retrace can be predicted with the Babson median line) , therefore we trade the 3rd wave (continuation of the spike) with a stop-entry just very little away from the extreme of the spike. In this way we prefer the spike only because it is easier to detect it among other patterns, and because of its high momentum, which results also in to higher potential profits. During the trading we may pyramid. We exit by trailing or deceleration of the spike momentum. The sub-rule buy at a support level, sell at a resistance level still holds here, if we define appropriately support-resistance levels.
Approximate equation for this patterns is the p(t)=exp(a(t)t)+e(t) . Where a(t) is the intensity of the growth and e(t) is the random disturbance term and of course p(t) is the price while t is the time.
2) The trends when they have sub-waves (called here expeditions), are traded by their internal (Elliot) sub-waves, where we trade only the sub-waves (vectors) that go along the trend. They are the longer. Those that go against the trend are the shorter. We do not trade the shorter. The subwaves according to their order of appearence either along or against the trend are named here as 1) Lighning (usually a spike) , 2) Thunder, 3) Blow, 4) Wave4, 5) Wave5 etc. Actually we pyramid at the start of the subwave Blow, then sub-wave4, then subwave6 etc or at the break-out of these subwaves from the forward extremes of the previous longer subwave. If there are no sub-waves at all (called here excursions, or excursions are called the subwaves of a trend while the trend itself in such a case is called expedition) , then they are traded simply by open-and-hold (and below we will improve the trading with pyramiding and adjusting of position size inside the trade and between support-resistance lines. Of course there is also the coarser pyramiding, only at the support-resistance levels, that is only at the start of the subwaves) . We trail by the shorter sub-waves height and exit also on the appearence of deceleration (usually divergence) of the momentum of the trend. If we accept the concept of non-horizontal skew supports-resistance lines then: The sub-rule buy at a support line, sell at a resistance line still holds here.
Approximate equations for this patterns is the
1) p(t)=exp(a(t)t)+e(t) . Where a(t) is the intensity of the growth and e(t) is the random disturbance term.
2) p(t)=A(t)*Cos(b(t)t+c(t))+exp(a(t)t)+f(t)*t+d(t)+e(t) . Where A(t) is the amplitude of the wave , b(t) defines the frequency of the wave, the c(t) the phase and the a(t) and f(t) are exponential and linear intensities of the growth, d(t) the level , and e(t) is the random disturbance term.
3) The flat-waves (curles or the technical analysis triangles and rhombus) are traded as 1st preference, externally by (continuation or reversal) break-outs on their boundaries. As 2nd preference they are traded internally by their sub-waves in only one direction or two directions and when they are not of expanding or diminishing amplitude (triangles) only the middle 1/3 as the trading of the flat of constant amplitude waves. The deceleration of their momentum is again the anticipation of the break-out. The break-out will lead usually to a spike (or trend) which traded as above. The diminishing triangles are preferred to be traded externally rather than internally.
When a flat channel reduces in width, thus suggesting a diminishing triangle, it is more likely that a new trend will start, compared to a flat channel that keeps its width constant. While the expanding triangles may be traded internally too. Also if the flat wave pattern is after a spike, we anticipate that it will have sufficient long duration and we may trade it internally too. But if it is at a reversal we anticipate that it will not last for long, so we prefer to trade it externally only. In other words if we anticipate that it will last for long we trade it internally but if we anticipate that it will not last for long we trade it externally. It is important to notice that because flat channels change easier than trending channels, the internal trading of a flat channel is the last priority among modes of trading of the patterns and somehow worse than the internal trading o a trending channel. But the external trading of a flat channel is a better priority than the internal trading of a trending channel, because it is a pro-active setup relative to its anticipated break-out. . The sub-rule buy at a support line, sell at a resistance line still holds here.
Approximate equations for these patterns are the
1) p(t)=A(t)*Cos(b(t)t+c(t))+d(t)+e(t) . Where A(t) is the amplitude of the wave , b(t) defines the frequency of the wave, the c(t) the phase, d(t) defines the level of he prices and e(t) is the random disturbance term.
2) p(t)=exp(a(t)t)*Cos(b(t)t+c(t))+d(t)+e(t) . Where exp(a(t)t) is the expanding or dumping amplitude of the wave , b(t) defines the frequency of the wave, the c(t) the phase , d(t) defines the level of he prices and e(t) is the random disturbance term.
4) The stationarities (ranging market) are traded almost as the waves in two modes internally and externally. Internally: We open 1/3 away from the middle line and we close almost at the middle line. We also trade them as the constant applitude waves, externally, anticipating break-out usually to a spike (or trend).
Stationarities that are after a spike may be traded internally , because we anticipate that they will have sufficient duration. But if they are at reversals we anticipate no sufficient long duration and we prefer to trade them externally only (break outs). In other words if we anticipate that it will last for long we trade it internally but if we anticipate that it will not last for long we trade it externally. Here the optimal adjustment algorithm on the position size applies, but not pyramiding. The sub-rule buy at a support line, sell at a resistance line still holds here.
Approximate equation for this patterns is the
1) p(t)=d(t)+e(t) . Where d(t) are linear level and e(t) is the random disturbance term.
The most predictable effect modulated by cycles (see post 5) is the reaction to an super-exponential moves (a blow-up at the end of trend in the form of super-exponential move or terminal spike). (See e.g. https://www.ted.com/talks/didier_sornette_how_we_can_predict_the_next_financial_crisis and http://www.er.ethz.ch/ Such super-exponential terminal patterns of trend may occur usually as result of overgrowth of the one of the two populations in a demand-supply coupling rather that of domination and not so much of competition or cooperation. See also post 22. .The frequency of emergence and the size of such super-exponential blow-ups follows the law of inequalities in other words the Pareto or Log-normal distribution and is thus by fat more often than pure randomness would predict!).
THE HIDDEN SINGLE PATTERN TO ALL THE ABOVE 4 PRICE PATTERNS IS THE SUPER-BUBBLE!
In other words a single move (which is contained to all the above patterns) but very strong and with accelerating momentum. THEN THE PREDICTED NEXT MOVE IS ITS REACTION
Here are some pictures and examples of super-bubbles as detected by the alligator indicator, from the book by Bill Williams "Trading Chaos" 2nd edition, chapter 9. The superexponetial-bubble is the simplifying "eye" that turns the chaos of price movements in to sequence of randomly occurring simple pattern (that of super-bubble) which thus allows for randomly intermittent predictability. The average waiting time for a super-bubble to occur is usually larger than the duration of the sum-total of the super-bubble and its reaction. But of course if searched among 10-15 instruments such super-bubbles occur one after the other.
The super-bubble is of higher predictable reaction, if the previous longer time interval , the market had already a slower trend in the same direction as the super-bubble (so that the spike-super-bubble is an terminal or exhaustion move!)
https://www.scribd.com/doc/177005696/Bill-Williams-Trading-Chaos-Second-Edition
Finally here is a ted.com video by Didier Sornette that explains why super-bubbles are unique, universal but predictable behavior of the markets.
https://www.ted.com/talks/didier_sornette_how_we_can_predict_the_next_financial_crisis
The Log-Periodic Power Law Singularity (LPPLS) applies as the bubble reaches it point of crashing. In other words faster and faster waves appear with smaller amplitudes. (
see http://www2.math.su.se/matstat/reports/serieb/2009/rep7/report.pdf
https://warwick.ac.uk/fac/sci/maths/research/events/2013-2014/statmech/ght/programme/sornette_2.pdf
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0165819
https://www.sciencedirect.com/science/article/pii/S187538921000249X
https://en.wikipedia.org/wiki/Didier_Sornette
)
In order to conduct successfully an intra-day system of transactions , that is successful in the long run and easy to keep on applying it the next points must be met.
1) It must be relatively utterly simple! Only the "eye of simplicity"can put order and tame the chaos of intra-day price patterns! It must be manual and not automated!
2) Therefore it has to be one only pattern among the 4 price patterns (see post 32)
3) To deal with this one only pattern, we may apply simplifiers like , velocity or rate of change of prices, acceleration, support-resistance.
4) Celestial periodicity will give the long-run stability, but it need not be one only frequency or period but a few neighboring frequencies or periods in the spectrum of celestial frequencies or cycles.
5) But most of all the strongest simplifier is that , when measuring the velocity or rate of change , by a stratified sampling hypothesis test, then it has to be an extreme value , which will indicate a reaction or closing of the cycle. This in particular means that we entirely avoid the parametric predictive models of econometry that assume predictability at every time step, as for such to be succsesful they would need to be pod stochastoc coeficients and there practically no such econometric models, and we resort to the more robust and with less assumptions non-parametric statistics and in particular of a single non-parametric measurement of the velocity of the prices, with stratified sampling.
6) It must be a phenomenon tested scientifically with valid quantitative procedures , with sufficient good (intermittent) predictability , for many years.
7) The financial result should be adequate (e.g. >= 1MDS).
8) The financial result, in my case, is to be used not only for economic freedom, but also for a worthy goal e.g. so as to finance my innovative research in the new millennium digital mathematics.
9) The solution to all of the above 8 points leads to one only system: The (solar spikes) super-bubbles system:
(See end of post 44)
The system: MAGNETIC SOLAR SUPER-BUBBLES (MSB)
1) It is the case where we apply a transaction system process e.g. on cross-rate currencies on 4-hours bins, and super-exponential extreme excitation or super-bubbles (Detected in an elementary way by the alligator indicator at H4 in MT4 or at M5 or at M1 (for M1 only if the spread is really tight or we utilize binary options) , in MT4 as Bill Williams describes when the super-bubble makes the alligator open its lines and the prices are also far away from the alligator lines as an extreme super-exponential move, usually with a terminal spike too. We wait till the move stops and starts reversing and we bet with minimum volumes on its reaction. If we fail initially we wait and insist again with higher volumes if the super-bubble completes a bit away and later (see also the post 55 about corrective escalation). In this way even if we say only 60% of the observed bubbles are genuine and reverse, we will have as success rate in the transactions more than 80%. We make sure that the lines of the alligator (that here we may refine to Smoothed moving averages of HL/2 or (H+L+O+C)/4 of periods 12,6,3 and offsets 6,3,1 for the H4 and M5 but the original alligator periods 15,8,5 and offsets 8,5,3 for M1.) are open and not closed thus indicating non-static but sloping averages, and that the candlesticks of the bubble show an acceleration with the few last candlesticks longer, thus indicating a terminal or exhaustion kind of spike or super-bubble. We prefer also not to bet on a super-bubble which is a continuation of the reaction to a previous super-bubble. Both the super-bubble and its reaction each one usually is of 6-12 bars so it lasts for 1-2.5 days at H4 of 4-hours bars or for 20-30-60 minutes at M5 of 5-minutes bars, which is again a solar cycle [see post 5] or even to 1-minute bars M1 in which case the solar cycle is of 5 minutes with half period 2.5 min . Then the system is not based on a constant growth rate, as e.g. the US index funds, and Bitcoin, neither on optimal risk separation rule. Still the maximum exposure rule still holds, and is the only rule , as positions are not kept open for long.(<= 2Day, s<= 5Days). And of course we still use the cycle of the markets called solar magnetic Parker spiral cycle (=month/4 period, thus 5 trading days with half period 2.5 days). The average waiting time for a super-bubble to occur is usually larger than the sum-total duration of the super-bubble and its reaction. But of course if searched among 10-15 instruments such super-bubbles occur one after the other. The super-bubble is of higher predictable reaction, if the previous longer time interval , the market had already a slower trend in the same direction as the super-bubble (so that the spike-super-bubble is an terminal or exhaustion move!)
Alternative way to filter super-bubbles:
THE STATISTICAL HYPOTHESIS TEST ALLIGATOR OSCILLATOR ABOUT THE RATE OF CHANGE OF THE PRICES.
Alternative ways to filter super-bubbles is to convert the alligator indicator to a statistical hypothesis test oscillator about the rate of change per bar of the prices. In other words we calculate a sample for the rate of change R(0)=(Price(n+1)-Price(n))/Price(n) of prices by the maximum period defined by the periods of the alligator, but also the rate of change R(i)=R(1), R(2) , R(3) (i=1,2,3 for the 3 lines of the alligator) per time step for the moving averages of the alligator indicator. Then take the z-scores of it
z(i)=(R(0)-R(i)) /sigma(R(0)), and finally the normal distribution function N(z) of it. By putting critical rejection area levels at 2% failure, we filter a super-bubble if the line of the rate of change of the prices is above 98% or below 2% , but also the lines of the 3 average rate of changes are away from the middle 50%. The Hypothesis test gives when with probability of error . This technique is essentially a stratified sampling hypothesis test about the velocity or rate of change of the prices,or of the velocity of a moving average, implicating a stochastic average rate of return as in the theory of Markovitz, thus necesarily startified or hierachical statistical sampling . If we utilize a Bollinger bundle, then it must have as period the middle period of the alligator, and as width 4 sigma or standard deviations. Then a super-bubble will get out of the channel, and at the same time, the middle line of the channel will have a considerable slope. This in particular means that we entirely avoid the parametric predictive models of econometry that assume predictability at every time step, (as for such to be succsesful they would need to be stochastic coeficients and there practically no such econometric models), and we resort to the more robust and with less assumptions non-parametric statistics and in particular of a single non-parametric measurement of the velocity of the prices, with stratified sampling. This is practically any statistical test of more complicated patterns of averages of prices, or of averages of rates of return of prices: By non-parametric statistical test over these averages as already 2nd order stochastic variables.
The stochastic model that is relevant is again the simplest possible one, e.g. starting from that of the Portfolio Theory of Markowitz, where for the rate of return R we postulate R(t)=R(0)+R(s,t)+e(t) where the R(0) is the constant average rate of return in time of the Markowitz theory of portfolio (constant trend) , R(s,t) is the seasonal part ,with average value non-zero , and on which we apply the above hypothesis test at various frequencies or sampling horizons or with stratified sampling , and e(t) has average value zero , it is normally distributed and is the random excitation part. The stochastic model has no-memory and for the sampling each step gives independent observation. From the above equation we may derive with the exponential function the final stochastic process of the prices and volumes that will be log-normally distributed.
Alternative way to filter super-bubbles:
THE STATISTICAL HYPOTHESIS TEST ALLIGATOR OSCILLATOR ABOUT THE RATE OF CHANGE OF THE PRICES.
Alternative ways to filter super-bubbles is to convert the alligator indicator to a statistical hypothesis test oscillator about the rate of change per bar of the prices. In other words we calculate a sample for the rate of change R(0)=(Price(n+1)-Price(n))/Price(n) of prices by the maximum period defined by the periods of the alligator, but also the rate of change R(i)=R(1), R(2) , R(3) (i=1,2,3 for the 3 lines of the alligator) per time step for the moving averages of the alligator indicator. Then take the z-scores of it
z(i)=(R(0)-R(i)) /sigma(R(0)), and finally the normal distribution function N(z) of it. By putting critical rejection area levels at 2% failure, we filter a super-bubble if the line of the rate of change of the prices is above 98% or below 2% , but also the lines of the 3 average rate of changes are away from the middle 50%. The Hypothesis test gives when with probability of error . This technique is essentially a stratified sampling hypothesis test about the velocity or rate of change of the prices,or of the velocity of a moving average, implicating a stochastic average rate of return as in the theory of Markovitz, thus necesarily startified or hierachical statistical sampling . If we utilize a Bollinger bundle, then it must have as period the middle period of the alligator, and as width 4 sigma or standard deviations. Then a super-bubble will get out of the channel, and at the same time, the middle line of the channel will have a considerable slope. This in particular means that we entirely avoid the parametric predictive models of econometry that assume predictability at every time step, (as for such to be succsesful they would need to be stochastic coeficients and there practically no such econometric models), and we resort to the more robust and with less assumptions non-parametric statistics and in particular of a single non-parametric measurement of the velocity of the prices, with stratified sampling. This is practically any statistical test of more complicated patterns of averages of prices, or of averages of rates of return of prices: By non-parametric statistical test over these averages as already 2nd order stochastic variables.
The stochastic model that is relevant is again the simplest possible one, e.g. starting from that of the Portfolio Theory of Markowitz, where for the rate of return R we postulate R(t)=R(0)+R(s,t)+e(t) where the R(0) is the constant average rate of return in time of the Markowitz theory of portfolio (constant trend) , R(s,t) is the seasonal part ,with average value non-zero , and on which we apply the above hypothesis test at various frequencies or sampling horizons or with stratified sampling , and e(t) has average value zero , it is normally distributed and is the random excitation part. The stochastic model has no-memory and for the sampling each step gives independent observation. From the above equation we may derive with the exponential function the final stochastic process of the prices and volumes that will be log-normally distributed.
The volumes measurement do provide better forecasting. The true rules of volumes are
1. A (statistical) momentum acceleration is a true acceleration, if the volumes are increasing too.
2. A (statistical) momentum deceleration is a true deceleration, if the volumes are decreasing too.
Of course the converse does not hold: An acceleration can be true, even if the volumes are not increasing. But if they are increasing we are sure it is true acceleration. The same with the deceleration.
The Log-Periodic Power Law Singularity (LPPLS) applies as the bubble reaches it point of crashing. In other words faster and faster waves appear with smaller amplitudes. (
see http://www2.math.su.se/matstat/reports/serieb/2009/rep7/report.pdf
https://warwick.ac.uk/fac/sci/maths/research/events/2013-2014/statmech/ght/programme/sornette_2.pdf
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0165819
https://www.sciencedirect.com/science/article/pii/S187538921000249X
https://en.wikipedia.org/wiki/Didier_Sornette
)
The Log-Periodic Power Law Singularity (LPPLS) applies as the bubble reaches it point of crashing. In other words faster and faster waves appear with smaller amplitudes. (
see http://www2.math.su.se/matstat/reports/serieb/2009/rep7/report.pdf
https://warwick.ac.uk/fac/sci/maths/research/events/2013-2014/statmech/ght/programme/sornette_2.pdf
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0165819
https://www.sciencedirect.com/science/article/pii/S187538921000249X
https://en.wikipedia.org/wiki/Didier_Sornette
)
2) Extreme excitation or super-bubbles are stronger , emotionally at the side of negative emotions , that is fear, panic etc. So it is exploiting by healing and counter acting fear, panic with the goal of reducing inequalities. If the bubble is a packet of real business transactions ordered in the big banks, then the reaction is by the hedge-market makers and liquidity providers other banks. The intuitive image that one gets from the idea of super-bubbles as they appear in the supply-demand of the markets, is the next: Imagine the balance of demand-supply in markets like the surface of a lake, then a super-bubble is a stone (of various sizes) as it falls on the surface of the lake, and the reaction to the super-bubble is the reaction waves that the stone creates.
3) At this system, the re-opened positions, may reach and usually start having exposure which is equal to the maximum exposure as traced on the Balance of the funds. There is no optimal separation ratio, and the nominal or effective leverage (see below in the definitions of terms) is much less (e.g. 5-10-20). E.g. if the elementary position may have exposure 5% at the begining , and there may be 10 such positions of effective leverage 5, (most of them insured my Stop Loss to zero loss, and in total of exposure <= 10%, the total nominal or effective leverage LE is about 10*5=50)
4) In conclusion this system has more difficult front office risk considerations and conduction, but easier back-office risk considerations and conduction compared to the Constant growth ratio system. As this MSB system is conducted at H4 scale (or H1 sometimes) it is even for this only reason more difficult to conduct.
5) Theoretically if given sufficient time in some cases the escalation of the positions may reach the optimal risk separation ratio (about 80% of the balance) , therefore of high effective or nominal leverage. Then it will become a faster capitalization , but also in total much more difficult conduction.
6) Demo practice 2 or 3 times in Alpari broker, during 2010-2011 , at various sizes accounts, as presented in 3 cases below, has shown a capitalization rate of 1MDS (doubling within one month) , but with maximum exposure in some cases reaching 20% (if the optimal Kelly criterion allows, after mainly subjective discretionary risk assessments). Also there is a demo practice in Strerling Gent broker below again with 1 MDS capitalization speed, but at higher frequency at M5 with more than 130 transactions (5 minutes bars, again super-bubbles detected by the alligator indicator). Nevertheless the spread in this demo account was really small, and probably this might work in real accounts with binary options where the spread can be avoided.
7) The system at h4 time frame , can be applied with binary options too, but the expiration must be 1 or 2 days after of the opening position, while the assessment that we are in a super-bubble, is done as mentioned above at 4-hours bars charts with the alligator indicator on them. In addition there must be applied an interactive-corrective escalation of positions too as described e.g.in post 55 .
8) Here are some pictures and examples of super-bubbles as detected by the alligator indicator, from the book by Bill Williams "Trading Chaos" 2nd edition, chapter 9. The superexponetial-bubble is the simplifying "eye" that turns the chaos of price movements in to sequence of randomly occurring simple pattern (that of super-bubble) which thus allows for randomly intermittent predictability. The average waiting time for a super-bubble to occur is usually larger than the duration of the sum-total of the super-bubble and its reaction. But of course if searched among 10-15 instruments such super-bubbles occure one after the other.
https://www.scribd.com/doc/177005696/Bill-Williams-Trading-Chaos-Second-Edition
9) Finally here is a ted.com video by Didier Sornette that explains why super-bubbles are unique, universal but predictable behavior of the markets.
https://www.ted.com/talks/didier_sornette_how_we_can_predict_the_next_financial_crisis
THE SUPER-BUBBLES DO NOT HAVE ONLY THE EASIEST PATTERN RECOGNITION (E.G. WITH PROGRAMMABLE STATISTICAL HYPOTHESIS TEST) BUT ALSO THE HIGHEST PREDICTABILITY OF NEXT EVOLUTION, AMONG THE OTHER MOE COMPLICATED PATTERNS.
7) The system at h4 time frame , can be applied with binary options too, but the expiration must be 1 or 2 days after of the opening position, while the assessment that we are in a super-bubble, is done as mentioned above at 4-hours bars charts with the alligator indicator on them. In addition there must be applied an interactive-corrective escalation of positions too as described e.g.in post 55 .
8) Here are some pictures and examples of super-bubbles as detected by the alligator indicator, from the book by Bill Williams "Trading Chaos" 2nd edition, chapter 9. The superexponetial-bubble is the simplifying "eye" that turns the chaos of price movements in to sequence of randomly occurring simple pattern (that of super-bubble) which thus allows for randomly intermittent predictability. The average waiting time for a super-bubble to occur is usually larger than the duration of the sum-total of the super-bubble and its reaction. But of course if searched among 10-15 instruments such super-bubbles occure one after the other.
https://www.scribd.com/doc/177005696/Bill-Williams-Trading-Chaos-Second-Edition
9) Finally here is a ted.com video by Didier Sornette that explains why super-bubbles are unique, universal but predictable behavior of the markets.
https://www.ted.com/talks/didier_sornette_how_we_can_predict_the_next_financial_crisis
THE SUPER-BUBBLES DO NOT HAVE ONLY THE EASIEST PATTERN RECOGNITION (E.G. WITH PROGRAMMABLE STATISTICAL HYPOTHESIS TEST) BUT ALSO THE HIGHEST PREDICTABILITY OF NEXT EVOLUTION, AMONG THE OTHER MOE COMPLICATED PATTERNS.
QUANTITIES-MAGNITUDES OF RISK-MANAGEMENT OF ESCALATION AND THEIR SYMBOLS.
LT : Effective (or nominal) leverage of the Portfolio. That is the ratio of the value of the open position by the portfolio and the balance (closed positions) of the funds
L0 , l0 : Effective (or nominal) leverage of the Portfolio of single elementary position.
LM : Margin leverage of the instrument in this account.
M0 : margin of elementary position P0
M : Size of the portfolio in number of elements or positions
P0 : Elementary position in volume size of contracts
V(P0): Value (nominal) of the open elementary position P0 .
Ind: Index fund or instrument in general.
Pr(Ind): Market price of the Index fund or instrument in general.
Mmax : maximum number of elementary positions in the portfolio, by escalation till optimal separation ratio reached.
Rs or Xs: Optimal separation ratio for the instrument and the chosen bins (usually days).
N: average true range of a bar (day) in price units.
RN: The previous N, as percentage of the price of the index or instrument.
F0 : Initial balance of the funds for transactions.
m: Contract size, at the symbol specifications of the instrument or constant multiplier to convert price changes into money for the instrument.
CS: Constant contract size in forex
e0: Maximum exposure percentage of the Balance of the funds , per elementary position P0 , usually 2%.
HYPOTHESES
1) Escalation spacing N of the grid, which remain constant during the escalation
2) Stop-Loss 2N
3) e0: corresponds to 2N exposure.
4) We assume funds remaining constant for the procedure escalation
5) We assume leverage l0 constant during the escalation.
For FX instead of index funds Contact size/Pr(ind)=m(t) which is variable while for index it is constant.
If the starting size of the elementary position (for indexes) is based on having the effective leverage equal to 1, then instead of the formula P(0)=e(0)F(0)/(2Nm) as above we use another by solving the equation P(0)mPr(Ind)/F(0)=1, which gives
P(0)=F(0)/mPr(Ind)
P(0)=F(0)/mPr(Ind)
<2 change="" font="" is="" not="" of="" prices="" rate="" the="" zero.="">
<2 change="" font="" is="" not="" of="" prices="" rate="" the="" zero.="">THE SUPER-BUBBLES DO NOT HAVE ONLY THE EASIEST PATTERN RECOGNITION (E.G. WITH PROGRAMMABLE STATISTICAL HYPOTHESIS TEST) BUT ALSO THE HIGHEST PREDICTABILITY OF NEXT EVOLUTION, AMONG THE OTHER MOE COMPLICATED PATTERNS.
THUS THE ORDER OF BETTER PREDICTABILITY IS
1) LONG TERM PERMANENT TREND
2) A CYCLE IN THE ORDER OF PREDICTABILITY DESCRIBED ABOVE AND REALIZED AS REACTION TO SUPER-EXPONENTIAL TERMINAL MOVE (OR SPIKE).
1) LONG TERM PERMANENT TREND
2) A CYCLE IN THE ORDER OF PREDICTABILITY DESCRIBED ABOVE AND REALIZED AS REACTION TO SUPER-EXPONENTIAL TERMINAL MOVE (OR SPIKE).
Here in the chart below which is with the prices after taking the logarithm, we may watch deviations from the linear moves (super-exponential moves) and their highly predictable reaction lasting in the average 3.75 years. during 30 years!
TIME SCALES AND PRICE PATTERNS OCCURRENCE
THE OVER ALL STATISTICAL BEHAVIOR OF PRICE PATTERNS
We notice that although the patterns are essentially 4 categories in details there are 6 distinct statistical patterns of the random or statistical posItion, velocity and acceleration of the prices.
A class of stochastic processes can be defined as the behavior of the markets based on these 6 basic patterns P1, P2, P3, P4, P5, P5, P7. The Pi i=1,2,3,4,5,6,7 are essentially random statistical patterns with random and variable parameters of size duration, and relative analogies that define them. E.g. we may have a Pareto distribution of the duration and price height of the patterns because of the inequalities in the markets. The 7th pattern P7 of stationary behavior we may call intermittency pattern. We may then assume a class of Markov processes with random and variable transition probabilities, where each random type of pattern occurs and then a next one occurs. Bu the transition probabilities are not arbitrary! E.g. spikes occur usually at the begging (initial spikes) and the end (terminal spikes) of trends, up and down trends with stationary channels in between them, shape cycles, that in the average of stable period, and some times of fixed beginning and end.
The probability that a type of pattern occurs changes also according to the time scale. In 2, 5 or more years annual bars time frames, the non-waving trend pattern is dominating, while say in 5-minutes bars time-frame flat patterns are dominating.
We notice that, by utilizing only pattern P3, of non-waving trend, and intermittency P7 we may derive all, other patterns with appropriate patterns of transition probabilities of the P3! I have coded a simulator of such a class of stochastic processes, superimposed on many time frames called Multi-time scales Rainbow Walks stochastic processes.
This is the overall behavior of patterns of the markets, and there are some invariant properties like
1) Cyclic behavior as alternation with up and down trend patterns with flat channels at the bottoms and tops
2) Statistical momentum conservation (see post 10) where the 1st time-step partial correlation of a price is almost always positive,
3) A Pareto distribution of the duration and height of the patterns, due to the inequalities of the enterprises in the economic system (see post 10, 25, 57,63 )
We notice that a common sub-rule for all the 4 price patterns is the
Buy at a support line, sell at a resistance line.
The support-resistance levels are the poles on which the 3 modes of a) Domination b) Competition c) Cooperation (see post 22) , start and end so as to form the 4 price patterns. Therefore any system based on the support-resistance levels is very basic, fundamentally simple and statistically successful. One of the best ways to detect the support-resistance levels (poles or attractors) is through the mid-levels of the reversals or the continuation or stationary patterns(waving, diminishing, expanding or constant amplitude. The best tool here is the statistical, e.g. price-activity histograms).
SUPPORT-RESISTANCE TRADING METHODS In particular a very simple and high probability of success trading strategy with support-resistance is the Single Support-Resistance System. This system is meaningful for cross-rates like EURUSD or cross-rates of currencies closely correlated e.g. in the same continent etc where the balance and stationarity of the market is more important and often rather than any continuing trend. Because in this system we bet for the stationarity-stability of the market rather than any kind of continuing growth. According to this method, we buy and sell simultaneously, AT a single (statistical or maxima-minima, or decimal xx00, xx50, or better mid-level of a continuation or reversal pattern or boundary of a flat channel, etc) support-resistance level. Then the statistical behavior of the market is that with high probability it will hit the support-resistance it will deviate a little and will return to hit it again. When it deviates, we close the position-direction with profit (putting also a reasonable stop-loss, e.g. equal to the targeted profit, to the other part of the position) , and when it will return, (even not completely to hit the support-resistance) we close the losing position-direction. The total is a small profit, which nevertheless it happens with high probability and very often. A variation of it, is the double or ending support-resistance system, in which we close the winning position only if it reaches a next support-resistance, so that we have time for the losing position to deviate or retrace. And in this variation we may open the positions at a support-resistance which is convenient as the market stays there for long, or anywhere (e.g. after a spike even if it seems ending ) , as long as we end the positions on a support-resistance. Because it it independent from where we open the pair o positions, but it does matter how we close around a support-resistance. As a support-resistance is also a statistical one, in other words if we calculate online-real time the histogram of price activity, of how often the price is found in particular price-bin (pip or tick) , the support resistance will be a (local) maximum of the histogram, we have also a mathematical proof why such a system would be winning when repeating it. (Here the conservation in adequate time duration of zero-momentum of the market at a support-resistance is required for the proof).
The CORRECTIVE ESCALATION from a support-resistance level, trading method
According to this method it is not so important the pattern recognition method and the pattern recognition in other time frames, or the focus on the 3rd wave in the current time frame, as it is the corrective adjustment and escalation of the position. Less than 20% of the success is in the initial high probability forecasting based on charts, technical analysis, indicators, and pattern recognition, and more than 80% of the success is on the corrective escalation technique after the Kelly criterion. We may start at a support-resistance level, (which may be a continuation pattern at its channel boundary or mid level) and we open position at the most probably direction of little size only, if we do not predict well and the prices go to the opposite direction we increase and open in the opposite direction , and we repeat it so as to have always a net position size towards where the prices are going. We do not use stop-loss as the role of the stop loss is the level where we open the opposite position. When the market breaks out finally to a direction we escalate (pyramid) and increase further the position. Typical sequences of position sizes are (1,-2,3, 5,4,3,2) or (-1,2,-3,5,4,3,2 ) We close by trailing or to the next support resistance level. Or we close at the predicted (by Babson median) end of a wave-vector between the 1st and 3rd (in Elliot counting).The starting leverage is usually 1 and increases to 2, 3 etc. While the percentages f of risked funds, start with 1% and in the average is is safe if it is at the 5%-6%. In general if p1 is the probability to win a trade, and p2 the probability to lose it (p1+p2=1), and the reward-to-risk ratio is equal to b, then the percentage f of funds risked in the trade should be about f=(b*p1-p2)/b. This is the Kelly theorem (see http://en.wikipedia.org/wiki/Kelly_criterion ). Nevertheless we prefer to risk in the average only 5%-10%, of what the Kelly formula suggests, that is about 2%, and the reason is that after simulations (see post 43), the paths with 2% exposure, are much more smooth and psychologically bearable in their volatility and draw-downs, compared to the exposure suggested by the Kelly formula! Let us say that with pattern recognition we have a success rate of 65% and the reward-to-risk ratio is 70% , then by the Kelly formula we must not exceed risking (0.7*0.65-0.35)/0.7=15% If it was a pattern recognition with 70% success rate (something very difficult) then the Kelly exposure would be 39% ! With the CORRECTIVE ESCALATION let us say that we start say with risking 1%, and then we correct in the opposite direction with 2% and then we re-correct in the initial direction with 5% (in total 1%-2%+5%=4% in the correct direction or -1%+2%-3%+5%=3%),then we continue pyramiding with 4%, then 3%, then 2% and no more. Thus in total 4%+4%+3%+2%=13% thus less than the Kelly exposure.Such sequences of trades are called excursions, and permit the definition of hierarchical sampling, where while at the level of trades the success rate may be say 55% at the sample layer of excursions it may be 75% or higher. With the CORRECTIVE ESCALATION we have the chance, even if the probability of an up or down move to be 50% to increase the success rate of the trades, to 70% - 75% with the above sequence (given of course the average momentum conservation).
The above show why trading based on invariants like support-resistance lines is probably the most powerful method.
We may summarize the trading of the 3 patterns (flat channel, trending channel, spike) as follows
1) There is external trading of channels and there is internal trading of channels. External trading of flat channels is better (of lower risk and higher performance) than internal trading, only at the end of the flat diminishing (contracting triangle) channel , when it is ready for a breakout to a spike. Otherwise the 3rd wave internal trading of flat channels is better.
2) The internal trading is of flat channels and of trending channels. The internal trading of trending channels is considered better, than of flat channels.
3) The external trading of flat channels, is of flat channels after spikes or not. External trading of flat diminishing channels before spikes is considered the best opportunity, for spikes as it is pro-active.
Also expanding flat channels are the worse for external trading, the best (among flat channels) for internal trading. Contracting flat channels, are the worse for internal trading, but good for external trading.
A tree of questions to make a pattern recognition and make trading decisions is the next.
1) Is it a) a flat-channel? (constant width or diminishing, expanding width)
b) a trending waving channel?
c) a spike?
2) If it is a spike, is initial or terminal? Internal or external? (if terminal / internal we prepare for re-action trading, if initial/external we prepare for 3rd wave continuation)
3) If it is a trending waving-channel, are we on the start of its 3rd wave? Or at the start of its (2n+1)-wave? (for internal one-sided trading)
4) If it is a flat-channel, are we on the start of its 3rd wave? (for internal trading). If not is the channel diminishing or expanding leading to a new trend? (for external trading).
5) Check the acceptable background time-frames T1, T2, T3 (etc) to find a tonal pattern there. We are interested even better in particular for a spike or 3rd wave tonal pattern. as this has high predictability and low risk. Then rank the T0 (focal) time frame pattern in relation to the direction/phase/timing suggested by the tonal pattern and intermediate background patterns.
The global multi-time-frame filtering may let hidden probabilities to be better predicted on the patterns of the focal time-frame especially if the tonal pattern is a spike or 3rd wave, like the next
1) If a trending channel will more probably decelerate and stop trending, or even reverse.
2) If a flat channel will more probably breakout up or down.
3) If a spike more probably will react to a counter-spike or continue.
After answering the above 5 questions, we then chose among the next four modes of trading
1) Trade only spikes, posterior with the 3rd-wave rule
2) Trade only with the 3rd-wave rule (this includes trend-channels, continuation flat-channels and reversal patterns).
3) Trade internally with the nth-wave for flat channels, or one-sided with (2n+1) wave for trending channels
4) Trade all the above and also externally, at the end of continuation diminishing flat-channels.
The choice of any of the above modes, which are in order from lower risk to higher risk, is also very much depending on having found among the background time-frames a tonal pattern which is spike or 3rd wave, as this has high predictability and low risk.
A standard way to apply the pattern-trading, is to choose a simple indicator, at the focal time frame which gives many signals, called the beat of the trading, but then execute only those signals that are in congruence with the pattern rules both at the focal time frame but also to the background time frames, and in particular the tonal-background time frame. The beat of the trading usually depicts the inner waves of flat or trending channels at the focal time frame. This will create the necessary intermittency and Pareto 20-80 rule on the opportunities. For example in the daily time frame or the 1-minute bars time frame we may choose the cross of moving average of 2 bars and 5 bars, or the force-index of 2 bars, or the Bollinger bands of 5 bars, and then select signals according to the daily, weekly and monthly time frame.
Classical technical analysis patterns
1) The classical technical analysis of continuation patterns are the flat channel of constant amplitude and the flat channel of diminishing amplitude (triangle) or flat channel expanding amplitude, when before them was a trending pattern. In other words are composite patterns based on the above enumeration of the basic patterns. They are ideal for external trading. They are usually after trends without channel or spikes and most often continue the price movement rather than reversing it. The flat channel of increasing amplitude, is supposed to breakout as reversal rather than as continuation (and in the securities and commodities market usually breaks-out downwards).
2) The reversal patterns of classical technical analysis are very short-term rather flat-like channels that reverse a channeled trend to a new opposite direction channeled trend. But a simple way to detect and predict a reversal is the next: When after a trend, the prices are stationary, and the channel pattern is not one of the constant width flat-channel, or triangle-channel , then it is a reversal pattern, and the prices will reverse trend. They are ideal for internal trading of the 3rd wave towards the new trend (Blow vector) (see 3rd wave trading rule below) .
The internal trading of a flat channel can be considered a trading between two parallel support-resistance levels. One of the best ways to do it , is to open a double neutral position (buy and sell of the same size), somewhere in the mid of the channel [and usually immediately after we detect a reversal or continuation or stationary pattern ], and close the winning position after a reflection on any of the two boundaries (support-resistance), while close the losing, at the opposite boundary, or anywhere the profit is positive and after at least 1/3 of the width of the channel. This type of the initial neutral double position trading, can result in to external trading too, if the prices breakout from one of the boundaries. In that case we close the losing position, and we go one with the winning at least till we break-even. The advantage is that we get a trading no matter which direction the market chooses, and also psychologically is appealing for a start of the trading. We call it the initially neutral parallel support-resistance levels trading.
Spikes Trading:
3rd wave (selective to opportunities) trading rule:
Among the various ways to trade the above patterns, we highlight the the 3rd wave towards the new trend (Blow vector) which a very selective and safe trading method among the patterns. In this mode of trading, we focus on waving patterns only (flat, continuation, or reversal, or waving-trend), and we want to take the waving pattern from its start. The best way to do it, is to detect the 1st wave (called Lightning) of the channel (usually a spike if its a reversal and a new waving channel is forming), then wait for the retrace of the 1st wave , that is the 2nd wave (called thunder), and enter, at the beginning of the 3rd-wave (called Blow). We may or may nor pyramid along the 3rd wave. The expected level that the 3rd-wave will reach is calculated with the median line of Babson. Since the channel (flat or trending) requires at least two waves to be shaped, we realize that trading the 3rd wave is essentially trading the earliest possible wave after the channel has been shaped.
This method, if the 1st wave is a spike (essentially initial and external) is also a spike-trading method. And as we may accept skew support-resistance likes, and the 2nd-wave starts and ends in the support-resistance or boundaries of the channel, it is also a support-resistance-to-support-resistance method. We may notice that this method applies not only to waving trending channels, but also to flat waving channels! And it may also apply to non-waving trends, if we look to the larger time-frame and analyze if, there, the non-waving trend is the 3rd wave. So we realize that the selective under risk-management 3rd wave rule can essentially trade all patterns.
3rd waves are detected mainly at changes of the market patterns. There are mainly 4 types of changes (another term could be mutations)
1) A channel (waving) trend changes in to a flat channel (continuation pattern)
2) A channel (waving) trend changes on to a reverse channel (waving) trend (reversal pattern)
3) A flat channel changes in to a channel (waving) trend.(Break out trend starting)
4) A flat channel changes in to a flat channel , higher or lower from the previous.(Break out flat channel starting)
All changes occur with a 1st wave (very often spike) , and we may wait for the 3rd wave.
Here are some videos with details of how to trade the 6 price patterns
http://www.youtube.com/watch?v=1ByzZ6b4ET8
http://www.youtube.com/watch?v=A49xzKYGyg4
http://www.youtube.com/watch?v=6YZ4ORz-UJ0
http://www.youtube.com/watch?v=hyELHtxTp7I
http://www.youtube.com/watch?v=4bIiQyfqW2U
http://www.youtube.com/watch?v=SqCFhhou3SM
http://www.youtube.com/watch?v=TzqDzQwPJnE
http://www.youtube.com/watch?v=_0-SJNnnHFM
http://www.youtube.com/watch?v=h8BP24tDJhA
http://www.youtube.com/watch?v=cf80pXwlda0
http://www.youtube.com/watch?v=J5fc25zRkWI
http://www.youtube.com/watch?v=V1VdCj57Zas
Even if a market would be a sequence of spikes alternating with flat continuation channels (that is a market without trends) , with 50% probability of the next spike up or down (a random walk of spikes) , still an external trading of the breakouts of the flat intermediate channels would be a very profitable trading.
By involving the 7 main factors of front-office in trading:
a) Support-Resistance lines and channel (which may require to involve the W. Babson media lines recursive forecasting, and channel width-zones or width-phases rules)
b) Deceleration (or divergence)
c) Volumes (especially as periodicity in the 3 sessions in forex. Or e.g. with the indicator On-Balance-volumes which is an excellent smoothing, the decelerations is also easy to read) Low volumes means also low volatility, and it is not good for external trading of channels but it may be good for internal trading of channels, as long as the channel has sufficient width.
d) Correlation of start-end of trends with spikes
e) Pyramiding (as Pareto trend-duration optimality)
f) Adjusting (or anti-pyramiding, as optimal position size adjustments relative to random fluctuations, during constant trends. See posts 3 and 33. In optimal adjusting we increase the position when losing and it decrease when gaining, due to random fluctuations. It is best aplied to the equity curve, which in a successful system, has it own constant trend , but it can apply also to the spot market during constant trends.)
We may also define an additive score for the above factors a), b), c),d),g) and classify opportunities and instruments according to their score.
g) The focal frequency of periodicity and one background periodicity (periodicity of the volatility. All relevant parameters of the indicators of the focal periodicity, are powers of 2
[=(1/2)^n)] submultiples of the focal period. The same with the background period. We may take as focal the month=20days (star spin), and as background the half-year=120days. We could as well take focal the day=24 hours (planet spin) and as background the month=20days, but this would require automation almost 100%. Furthermore we could utilize as focal the day=24 hours (planet spin) with double background , the month=20days (star spin), and as 2nd background the half-year=120days.I call it the tri-angular system.).
[=(1/2)^n)] submultiples of the focal period. The same with the background period. We may take as focal the month=20days (star spin), and as background the half-year=120days. We could as well take focal the day=24 hours (planet spin) and as background the month=20days, but this would require automation almost 100%. Furthermore we could utilize as focal the day=24 hours (planet spin) with double background , the month=20days (star spin), and as 2nd background the half-year=120days.I call it the tri-angular system.).
It is hardly possible not to have a very successful trading
The rest of the back-office factors are the exposure and leverage rule, and the reinvestment and capitalization growth rules.We may also define an additive score for the above factors a), b), c),d),g) and classify opportunities and instruments according to their score.
A system protocol to trade trend-patterns, based on the previous factors that creates the "initiating setup" would be
1) We check that the focal trend just started and it is the same direction with the background trend(s).
2) We make sure that there is still acceleration and deceleration has not started. We confirm it also with the volumes, to be sure that it is a genuine acceleration.
3) We look also for an initiating spike to confirm the trend start
4) We pyramid each time by opening more positions, near the support or channel backward zone. Because adjustment and pyramiding are superimposed pyramiding is diminishing as the trend proceeds (e.g. like the numbers 5,4,3,2,1.Obviously the size 5 corresponds to the "Blow" Elliot subwave)
5) We adjust optimally during the trend by partially closing near the resistance , or channel forward zone. We may trail also as rule of optimal adjustment of the position size relative to the random fluctuations.
6) If deceleration appears, then the partial closing of 5) becomes a complete closing of all positions.
To apply the above protocol to the flat waves, and stationarity-ranging patterns also , we do not pyramid and adjust at each channel boundary (in other words along the trend or expedition), but during the movement that crosses transversally the channel (in other words along the subwave or excursion), and we may do it up only or down only or up and down, according the background periodicity, three states (up, down, neutral)
There are indicators for each of the above factors.
a) For the Support-Resistance lines it can be used, an indicator that gives a zig-zag with s/r lines , and/or pitchfork lines. We may also utilize a statistical histogram to define horizontal support-resistance. And we may also use the extrapolator channel indicator, so that its 5-line channel gives non-horizontal s/r . We may use also, the Ketler channel (simple or hull moving average) for curvilinear , non-horizontal support-resitance lines.
b) For the deceleration we may use the On-Balance-Volume indicator, or the Awesome oscillator of B. Williams, or the MACD, plus the pitchfork lines as in a). We may also use the extrapolator indicator which is a sinusoidal function fit at 30 days, thus it gives deceleration-acceleration.
c) For the volumes we may utilize an volumes-oscillator if we are interested for sessional periodicity, or the On-balance-volume indicator if we are interested for he deceleration or divergence.
d) For the spikes, a custom oscillator for spikes, or the spike-runs indicator or the aligator of B. Williams, or the ATR(1) with a moving average of more than 100 bars.
e) For the frequency of periodicity, we may focus on the star-spin (monthly) periodicity, with daily time frame, that is once per day monitoring. Thus the ketler channel as in a) should be at 20 days,or we may utilize the aligator of B. Williams which is at the periodicity of star-spin.. For background periodicity, the planet-orbital, that is 6 months or 3 months harmonic of the annual periodicity. We may use a EMA 50/100 days cross, or a bollinger bundle at 80 days and 2s.
f) For the pyramiding and adjusting and trailing with stop loss and take profit, we should program a grid trader to do it 100% automatically (I call such a robot a "weaver"). Then "drive" manually this robot, based on the manual pattern recognition of the trends. If we run the planet-spin system (focal periodicity=24 hours) then the "driving" of the grid-trader based on the hourly pattern recognition of the trend should be automatic too, which makes it more difficult that the star-spin system (focal periodicity=1month=20 days). As pattern recognition is best done manual, while pyramiding and adjusting, and trailing is best done automatically, the star-spin system (focal periodicity=1month=20days) is more effective at the present state of the art of technology.For both the flat-wave pattern and the stationarity pattern, works also a trading mode based on the Maximum Likelihood Price-Level principle. In other words as the mid-level of the channel is also as a "center-of-gravity" of the prices , it holds that the probability is maximum for the price to be there. Therefore it is optimal to open positions say 1/3 of the channel width around and away from the mid-level and close the position at the mid-level. In this way the success rate of the trades is maximized. (see also post 24)
5) We may include also as pattern the no-pattern at all, that is a case not falling in to any of the above 4 cases. Such no-patterns are not traded al all.
Sometimes trading the Spikes and Trends is called trend-following, while trading the wave-patterns and stationarities is called counter-trend trading (and scalping, if it is in fast time frames like 1 hour , 5 minutes etc).
Other times couter-trend trading is also called the trading of trends where we enter in a reversal a bit earlier when the deceleration and divergence appears, before a clear break-out to the new trend.
There are trading systems specialized to one only or some only of the trading patterns. E.g. the famous 50 years old, turtle trading system, trades trends (excursions and expeditions) and spikes only. Or e.g. scalpers trade mainly the ranging and waving patterns.
There is a rule of time-scales and the 4 patterns: The larger the time scale the more often and the clearer the trend without noise. So directional (only up or only down) trading is the best. The shorter the time scale, the more often and the clearer the ranging pattern. So bi-directional (both up and down) trading is the best. Spikes and waves appear clearer in special intermediate time scales, like 48-hours, or one month. This is a consequence of the law of attraction , the power law of economic inequalities and the power distribution of the trends and volumes,
Notice, that according to this approach, the assumed self-similarity of the market behaviour in different time scales is simply a myth. Although there is restricted self-similarity in different time-scales, there is also systematic deviation of the behaviour of the market in different time scales.
The smart duality in trading is not detecting the up or down direction, but detecting when to trade and when not, and detecting when the market is in the ranging/waving patterns (transversal trading) and when in the trend/spike patterns (longitudinal trading) .We must remember that in a good trading, more than 80% of the time the market is highly uncertain about going up (or down), and less than 20% of the time, the market has significant decidable probability that it will go up (or down)
These 4-price patterns create an optimal portfolio of 4 trading strategies specialized to trade respectively the 4 price patterns. The largest percentage of the portfolio, is allocated to the strategy that trades the spikes, less percentage is allocated to the strategy that trades usual trends, less percentage is allocated to the strategy that trades waves, and finally the least percentage is allocated to the strategy that trades stationarities. happily because each patterns excludes the occurrence of the other patterns, the same funds are used each time , and the percentages have the meaning of exposure, leverage, and percentage of funds risked each time.Notice that the more rare the pattern, the higher the exposure and capitalization speed and the lower the risk because the longer the time that we are outside the market. It may seem strange to the common sense, but the highest speed of capitalization is concentrating on spikes where also is the longest time that you are outside of trading.Really God's speed in capitalization.
We remind that as a summary the ways that optimal portfolio theory can apply in trading is through at least 3 perspectives
1) Optimal portfolio of instruments ( post 14)
2) Optimal portfolio of strategies on the 4 basic price patterns (post 32)
3) Optimal portfolio of time scales or characteristic frequencies. (post 37)
So we could state a Pareto-rule (20-80 rule) for trading the 4 price patterns:
"More than 80% of the profits are obtained from the trading of less than 20% of the opportunities among the 4-price patterns"
And obviously the best profits are obtained from the rare spikes , and the 3rd wave trading rule, which is a local rule to apply the Pareto 20%-80% rule of filtering opportunities . The global filtering rule is of course the filtering based on many background time-frames and patterns on them. The way to do is to search among the accepted background patterns at what time-frame the pattern is most clear and strong, and considered it as tonal pattern. Obviously half only of the opportunities at the focal time frame are accepted, those that agree in direction timing and phase with the tonal pattern. But we may also rank all acceptable opportunities of the focal time-frame of interest according to if or not the intermediate patterns from the total till the focal are favorably agreeing in direction, phase and timing. In this way much less of half the opportunities of the patterns of the focal time-frame are selected to trade. In this way by filtering locally with the 3rd wave rule and globally with the nested time frames after the tonal time-frame, the success-rate increases from say 52% to 80% or more!
It seems simple to describe or draw the above 4 price patterns and the trading of them. But any 100% automation of them (in the present state of the art of software and platforms 2011) with a robot or EA, would be unsatisfactory. It would be inadequate, even if it was done at the daily focal frequency (hourly time frame) or monthly focal frequency (daily timeframe) where there is low noise, or even if instead of indicators it was utilized readily available pattern recognition algorithms. Where the coding fails is at the stochastic pattern recognition. Most of the current examples (2011) of such automations succeed to have 50% only performance of the corresponding manual execution performance. Human senses inspection and pattern recognitions is always better, and it is adequate and satisfactory. The (manual) Bill Williams trading, is an example (described in 3 of his books) where over daily bars and by trading partly only and few only of the cases of the above 4-patterns, for the last 50 years (with 15 minutes only once per day inspection)he succeeded in having after the leverage an average annual 300% ( or 25% monthly) rate of return. Other example of such trading over daily bars, is the commodex protocol (which contrary to Bill Williams method, is not fully revealed, rather automated, and is sold only as signals, by Philip Gotthelf) which the last 50 years has an average annual rate of return of 130% (or 10.8% monthly).
The pyramiding is theoretically justifiable as optimal in the trading, after the Pareto duration of the above 4 patterns.
Upon the above rules, we must add the rule of constant percentage of funds risked per trade and for all open trades simultaneously and the rule of optimal adjusting , which extends and gives "weaving" (or bulkiness) to the classical trading of the patterns . (see also post 33). The combination of optimal adjusting and pyramiding, boosts the performance of classical trading of the 4 patterns, by a factor of about 7 times! (See also post 24). Of course all resulted rates of return of such a trading, if it were to be recorded by the bank accounting and published, would be figures divided by 100, as the usual maximum nominal leverage in forex is 100. So all this effort would be equivalent to achieve what is the usual annual return of the markets (10%-12%) but with 100 times less variance of the equities curve, so that nominal maximum draw down would not exceed the 0.33%.
( The reason it is boosted so much is similar to the reason that the measurement of the periphery of an island is boosted very much if we utilze fractals to measure it, instead of visible smooth curves on the map. This is also the reason that some traders consider it the "holy grail" of trading as it converts the wealth creation by classical investment and trading that requires initial capital, to wealth creation without initial capital. Although the optimal adjustment gives you "Godspeed" in wealth creation, it seems to me that any attempt to push the market further to faster speed of wealth creation than this, that the 4 price patterns, optimal frequency, pyramiding, and adjustment permit, in a standard chart-based and not inside-information-based trading, would be greediness and lead to destruction. And this is because, if the above optimal frequencies of the 4 price patterns of demand-supply and the mathematical optimal solutions of trading them with pyramiding and adjusting is all that the science and mathematics can define and solve for the systematic regularities of the markets, then trying to get something faster, is simply a worsening of the optimal solution. )
The optimal adjusting acts opposite to pyramiding: While optimal adjusting decreases the position in forward movement, pyramiding increases it. Also combined optimal adjustment and pyramiding under constant percentage of funds risked, increases the new position size in backward movement while it decreases the new position size in forward movement. Still it is not a "martingale technique" as it closes the position as soon as it reaches the stop loss. The stop loss is not defined by the acceptable percentage of funds risked, but by the minimum distance level that the pattern prediction is no more valid. From this level and the risk percentage is defined the size of the position.
The optimal adjusting acts opposite to pyramiding: While optimal adjusting decreases the position in forward movement, pyramiding increases it. Also combined optimal adjustment and pyramiding under constant percentage of funds risked, increases the new position size in backward movement while it decreases the new position size in forward movement. Still it is not a "martingale technique" as it closes the position as soon as it reaches the stop loss. The stop loss is not defined by the acceptable percentage of funds risked, but by the minimum distance level that the pattern prediction is no more valid. From this level and the risk percentage is defined the size of the position.
The optimal adjusting is done best with a grid trader, and should be at the maximum resolution that the spreads permit (e.g. 5 pips). As a contrast, the detection of the 4 patterns of price movements, is best when done at the largest possible time frame as compensated with their frequency of appearance (e.g. over daily bars, and over the monthly periodicity) where the noise is minimum.
So here is resolved a common fallacy: Many traders try to detect the above price patterns at intraday time frames, and trade them. It is not the optimal as the noise is higher , at intraday time frames (worse Sharpe ratio) than daily time frame. The real intraday action is only the optimal adjustment (also combined with pyramiding), which includes backwards as well as forwards, to the trend, iterative reopening of positions and trailing-outs or take-profits.
A good pattern recognition of the 4 basic patterns, as above for trading them, does require specially coded indicators and algorithms.
Personally,
1) I detect the ranging pattern through flat least square channels with a threshold on the slope and bounded approximation error.
2) I detect the waving patterns with zig-zags that give also support-resistance and also trigonometric (sinusoidal or harmonic) functions approximation , forecasting and Fourier analysis. For the support-resistance there is also the statistical method of peak of histograms of the price action.
3) I detect the trend patterns with straight trend lines over zig-zag, the Andrew's Pitchfork tool , Babsonian medians or with polynomials approximation.
4) And I detect the spikes with special coded indicators that measure the level of the speed and the acceleration, or with approximation of wavelet-impulses or with exponential functions approximation.
The 3 patterns waving, trends, spikes are very well modelled with sinusoidal, polynomial, and exponential functions that are the only solutions of linear recursive equations.
But once this is done and locked in a template, then the best way to detect the patterns is manually by subjective observation and not automated in a robot or EA. happily this can be done with 15-20 minutes only once per day, as it is optimal to detect the 4 patterns in larger scale over daily bars or daily monitoring frequency. On the other hand the (Pareto continuation) pyramiding and adjusting cannot be done manually as it is optimal to be executed in the fastest possible frequency and time frame. It requires automation with a robot or EA. And the above also is my concept of optimal combination of automation and manual conduction, for best results in performance and risk management. No manual alone or 100% automated alone can be better than the above combination.
A good pattern recognition of the 4 basic patterns, as above for trading them, does require specially coded indicators and algorithms.
Personally,
1) I detect the ranging pattern through flat least square channels with a threshold on the slope and bounded approximation error.
2) I detect the waving patterns with zig-zags that give also support-resistance and also trigonometric (sinusoidal or harmonic) functions approximation , forecasting and Fourier analysis. For the support-resistance there is also the statistical method of peak of histograms of the price action.
3) I detect the trend patterns with straight trend lines over zig-zag, the Andrew's Pitchfork tool , Babsonian medians or with polynomials approximation.
4) And I detect the spikes with special coded indicators that measure the level of the speed and the acceleration, or with approximation of wavelet-impulses or with exponential functions approximation.
The 3 patterns waving, trends, spikes are very well modelled with sinusoidal, polynomial, and exponential functions that are the only solutions of linear recursive equations.
But once this is done and locked in a template, then the best way to detect the patterns is manually by subjective observation and not automated in a robot or EA. happily this can be done with 15-20 minutes only once per day, as it is optimal to detect the 4 patterns in larger scale over daily bars or daily monitoring frequency. On the other hand the (Pareto continuation) pyramiding and adjusting cannot be done manually as it is optimal to be executed in the fastest possible frequency and time frame. It requires automation with a robot or EA. And the above also is my concept of optimal combination of automation and manual conduction, for best results in performance and risk management. No manual alone or 100% automated alone can be better than the above combination.
The trading systems can be classified also according the various combinations and sequences of the 4 price patterns (e.g. Trend->Trend reversed, Trend->stationarity, Spike->trend, Trend->stationarity->trend etc). In addition if we inspect 2 different time-frames one focal, and one background, then the classification, is of pairs of the above combinations. I have tried it, but it becomes quite complicated. The pairs or triples of the 4 patterns in a single focal time-frame seem to me adequate.
A good trading system usually does not trade all the time, so that the account is not exposed all the time. There should be time intervals that the system does not trade, as the conditions are nor favourable for forecasting or opening positions. We call this phenomenon intermittency in trading. It is a quite common idea in trading that by measuring the trend in different time frames, and requiring some aggregate condition on them, intermittency is created.
But we shall discuss here a quite uncommon type of intermittency. It is the intermittency created by concentrating and detecting only spikes from the 4 price patterns above. The reason is that spikes are correlated with longer duration of the subsequent trend. And this gives better probabilities and opportunities for trades that will close profitably rather that at the stop loss or in loss. We may call such systems the Spike-based trading systems. Notice as a contrast, that martingale systems are stationarity-based or ranging market-based or wave-based trading systems.
Spike-based trading systems permit very reliable results where the average profit of a trade is double or more times larger compared to the average loss of a trade.
The maximum draw-down of the equity is also surprisingly small.
The emotional impact of such systems is also peculiar: There are small repeated losses and then something big which is a good surprise and is a large profit. Notice as a contrast the emotional impact for example of a common and unsophisticated martingale where you have small repeating gains, and then a big surprise which is a big loss. Although both modes can be profitable in a systematic way.
According to my tastes such spike-based trading systems are my preference. Bill Williams has dedicated a whole book on a spike-based system, where he trades the retrace and reaction of the spike. A spike is almost always a breakout. But a breakout is not always a spike. There are different approaches where a) The spike itself is traded (Lighning) b) The retrace and reaction is traded (Thunder) c) The continuation is traded (Blow), d)Any combination of the a),b),c).
The main difficulty of spike-based trading systems is how to detect a spike early enough in a reliable way. It is a difficulty that the usual linear indicators hardly can solve. My approach is to design and use non-linear indicators with appropriate exponents that solve the problem in a neat and reliable way.
There are mainly two methods to detect a spike: a) The static b) the dynamic. The static detects it as a breakout from a support-resistance or non-horizontal trend level. Although a spike is a breakout, a breakout is not always a spike. Therefore a special logic on the straight lines should be involved so as to restrict to some breakouts only. Alternatively we may involve the dynamic method, through the growth dynamics of momentum or acceleration.
There are mainly two methods to detect a spike: a) The static b) the dynamic. The static detects it as a breakout from a support-resistance or non-horizontal trend level. Although a spike is a breakout, a breakout is not always a spike. Therefore a special logic on the straight lines should be involved so as to restrict to some breakouts only. Alternatively we may involve the dynamic method, through the growth dynamics of momentum or acceleration.
The benefit of discovering how to detect spikes is worth the design effort on indicators and logic on lines.
My extensive back-tests show that among all the possible rainbow frequencies (see post 5) or time frame spikes, the optimal is at the color blue rainbow frequency; in other words the sessional-spikes that are usually of the size of the average daily range, but occur in less than half of the daily duration.
The correlation of spikes with the duration of the subsequent trend is of course a consequence of the law of attraction as in the post 9. See also post 11 and 25. That is, longer trends are shaped by volumes of transactions inherited by larger organizations or economies, and their packets are larger creating larger "populations" of abrupt demand-supply and therefore spikes.
The same law (of attraction) leads to the optimality of escalation practice (pyramiding).
The law of rythms , the law of action , the law of polarity, and the law of attraction are probably the most important in the quest for a "holy grail" of a trading system of long lasting success, or for the quest of a "golden thread" for a long lasting prosperity.
My extensive back-tests show that among all the possible rainbow frequencies (see post 5) or time frame spikes, the optimal is at the color blue rainbow frequency; in other words the sessional-spikes that are usually of the size of the average daily range, but occur in less than half of the daily duration.
The correlation of spikes with the duration of the subsequent trend is of course a consequence of the law of attraction as in the post 9. See also post 11 and 25. That is, longer trends are shaped by volumes of transactions inherited by larger organizations or economies, and their packets are larger creating larger "populations" of abrupt demand-supply and therefore spikes.
The same law (of attraction) leads to the optimality of escalation practice (pyramiding).
The law of rythms , the law of action , the law of polarity, and the law of attraction are probably the most important in the quest for a "holy grail" of a trading system of long lasting success, or for the quest of a "golden thread" for a long lasting prosperity.
Once we have a trading system good enough to be successful with trades where StopLoss<=TakeProfit we may apply a reinvestment technique based on the Kelly formula:
f=(bp-(1-p))/b . where f=optimal percentage of funds to risk, b=TakeProfit/StopLoss (it has to be >=1) and p=probability of success of the trade. (see http://en.wikipedia.org/wiki/Kelly_criterion)
For example let us say that we have an account of 1000$, and our trading system has a general success rate of 66% (after back-tests with constant lot size). Let us assume that we take a signal for a trade, that its stop-loss SL=20 and take-profit TP=30 have ratio TP/SL=1.5
Then the Kelly formula can be used to define the lot size: The optimal percentage to risk is f=(1.5*0.66-0.33)/1.5=0.44 or 44%. In other words we may risk 440$, and for the 20 pips stop loss this is converted in to 440/20=22$/pip or 2.2 standard lots! Seems quite a big number indeed. In practice we should trade with smaller lot sizes, (e.g. a fixed percentage of what Kelly formula gives , my experiments show that 5%-25% of the Kelly percentage is a safer percentage) because the Kelly formula is derived with the assumption of continous subdivision of the funds, and constant success rate p, while in practice, the minimum lot size, does not permit such a procedure, and the success rate p of the trades is variable. The latter leads to utilizing not the average succes rate of trades average(p), but a cut-off tail of its distibution after an appropriate hypothesis test, thus leading with a probability 92% of not overating the success rate p, to only say 10% of it. Thus also a 10%-12% only of the Kelly percentage.
If we continue to apply this formula, if the account grows, larger, more lots are opened so it is both a de-investment (when the results go bad) and re-investment (when the results go well) money management system.
We may also derive directly the optimal leverage from the above Kelly formula and the SL, as
OptimalLeverage=((f*Equity)/SL*10)*100=(((((TP/SL)*p-(1-p))/TP/SL)*Equity)/SL*10)*100 or
OptimalLeverage=(((((TP/SL)*p-(1-p))/TP/SL)*Equity)/SL*10)*100
In practice, my backtests show that one should take a constant smaller percentage over the Kelly variable percentage.
We may also derive directly the optimal leverage from the above Kelly formula and the SL, as
OptimalLeverage=((f*Equity)/SL*10)*100=(((((TP/SL)*p-(1-p))/TP/SL)*Equity)/SL*10)*100 or
OptimalLeverage=(((((TP/SL)*p-(1-p))/TP/SL)*Equity)/SL*10)*100
In practice, my backtests show that one should take a constant smaller percentage over the Kelly variable percentage.
Upon this money management system we may superimpose the optimal adjustment management system, described in another post that somehow does the reverse.
Traders usually fix the maximum draw down and the risk of funds per trade based on psychological criteria. E.g. They say I am willing to accept obly 1% risk per tarde or only 6% risk per trade. But the right approach is to derive it based on the kelly criterion (which is derived from the goal of optimal maximum capitalization speed and minimum probability of rashing the account). Therefore the procedure should be as follows:
1) We backtest the system with a rather reasonable but otherwise arbitrary risk per trade (e.g. between 1%-6%). Then we take note of the success rate p of the trades and the average trade profit/average trade loss as average risk/reward ratio b
2) We utilize the kelly formula f=(bp-(1-p))/b to find the optimal risk per trade f
3) We re-run the backtest with the optimal risk per trade f, to see the final features and parameters of the trading system like absolute initial and relative maximum draw down.
Traders usually fix the maximum draw down and the risk of funds per trade based on psychological criteria. E.g. They say I am willing to accept obly 1% risk per tarde or only 6% risk per trade. But the right approach is to derive it based on the kelly criterion (which is derived from the goal of optimal maximum capitalization speed and minimum probability of rashing the account). Therefore the procedure should be as follows:
1) We backtest the system with a rather reasonable but otherwise arbitrary risk per trade (e.g. between 1%-6%). Then we take note of the success rate p of the trades and the average trade profit/average trade loss as average risk/reward ratio b
2) We utilize the kelly formula f=(bp-(1-p))/b to find the optimal risk per trade f
3) We re-run the backtest with the optimal risk per trade f, to see the final features and parameters of the trading system like absolute initial and relative maximum draw down.
Applying a fixed system on different instruments or groups of instruments and markets leads to different success rate p of the trades. For example the current financial crisis (2008-2011 etc) makes the gold and silver prices have remarkable persistent high intensity trends, which leads to an increased success rate p of the trades. Consequently, the % of the funds to risk according to the above (modified) Kelly formula as well as the optimal leverage is considerably higher.
Probably the best instantaneous rewarding "why?", of manual trading is the joy and satisfaction in playing, among the situations of higher or lower uncertainty of what will happen in the global economy and markets, so as to plan and conduct a strategy that lets you know on occasions what will happen with acceptable low uncertainty.
Probably the best instantaneous rewarding "why?", of manual trading is the joy and satisfaction in playing, among the situations of higher or lower uncertainty of what will happen in the global economy and markets, so as to plan and conduct a strategy that lets you know on occasions what will happen with acceptable low uncertainty.
