Chaos theory and market reality

When a newbie begins to experience problems in their trade, their first reaction is the idea that success in the market he must learn to predict price movements. With a little effort, he finds that the long-term predictions using fundamental analysis, and for short-term - technical. If our new trader examines the history of prices of the market in which he works, he finds what seems to be repeating patterns. For a long time the markets move up and down the long cyclical waves. If he looks closely, he can find some short-term schedule for the figures that are repeated again and again. Once he opens the world of mathematical indicators, he finds that certain combinations of indicators and figures tend to be repeated - often near the major peaks and troughs.

Finding these recurring patterns of different species, he quickly calculates the dimensions of crazy profit, which is possible if someone would take the right action at the right moment.

Not surprisingly, our novice enthusiast will come to the conclusion that this market is something that repeats again from time itself in various ways. And if he or she can learn the characteristic shapes and loops, big profits will flow into the pocket itself. Markets can be organized so that they constantly repeat themselves over and over again in some encrypted form. And if our trader can split this secret code, it will be possible not only to huge profits, but he or she can do to avoid losses.

Our trader begins to pursue this goal, using the available literature. Perhaps the mail would bring him an offer specialized trading systems that profit from certain patterns, certain limited number of qualified professionals. Since these systems are often evaluated in thousands of dollars, our trader would suggest that they should really be working trading tools. Brochures on these systems often contain stories about the legendary trader or, conversely, the traders-recluse, who discovered the secret of the market and made millions on this. Using various tools, the seller has received these secrets and now he is willing to share them with only a few lucky traders - for a fee from them. Such stories reinforce the belief that some traders are making big money, have achieved this success thanks to the fact that they found on the market, something that can only know superprofessional.

However, despite the enormous amount of this kind of prediction methods in books, trading systems and software products, in any year about 75% of traders lose money. If you take a longer period of observation, they lose money about 95% of traders. Yet virtually none of the traders do not ask themselves whether there are in principle suitable for use repeating patterns in prices. I recently saw an article entitled "The magic of graphic figures and projections." This article said nothing about profit, but a lot of "magic".

Its very nature, people are more receptive to the ideas of giving up hope. They believe what they want to believe, despite all the contrary evidence. Fleeting success has led many traders to strengthen believers in their false beliefs. In a recent interview I asked what the most important quality a trader should have. I replied: "There are many different qualities, which must have the trader to be successful. They are all important. But I must stress that there is one quality that is most important - is the ability to perceive reality as it is. "

Unsuccessful traders have a distorted view of the markets themselves, and that they actually do when they sell. For them is very difficult to shake off these distorted images, so they are doomed to defeat in the long run. In this market operates in such a way that strengthens them these misconceptions. This smooths out the contradictions that arise from them.

Introduction to the theory of chaos. Exclude the possibility of studying the history of the market by mathematical and statistical methods and identify the existence of certain repetitive patterns and cycles.

The markets are nonlinear dynamical systems. Chaos theory is a mathematical tool for analyzing this kind of nonlinear dynamic systems. The use of this device shows that market prices are highly random in nature with a slight trend component. The magnitude of this trend component varies from market to market and the value of the time window.

To explain the concept of chaotic systems using fractals. Fractals - are objects having the property of self-similarity, ie such objects, parts of which are similar to the whole object. A popular model for explaining a tree. While the branches get smaller and smaller, but each branch is similar in structure to better order and verve tree as a whole. Similarly, when placing the movement in prices for monthly, weekly, daily and intraday charts the structure of motion remains similar. Like the wildlife - the approach you see more and more detail.

Another characteristic of chaotic markets is the so-called "sensitivity to initial conditions (sensitive dependence on initial conditions)». This is what makes dynamic market systems so difficult to predict. Since we can not exactly describe the current situation and as many errors and inaccuracies in the description of the situation accumulate over time due to the total system - an accurate prediction is impossible. Even if we could accurately predict tomorrow's price change (and we can not), we will still have zero accuracy in predicting even 20 days in advance.

A considerable number of thoughtful traders and experts suggest that intraday trading, such as a 5-minute bars, there is an attempt to trade a random noise and, therefore, is a waste of time. Over time, those who trade the noise, are doomed to defeat because of the cost of trading (commission, overhead, etc.). At the same time, these same experts say that long-term price movements are not random. Traders can trade successfully on the basis of daily or weekly charts if they follow trends.

A natural question arises - how, at one and the same market may be that short-term price movements are random, whereas those composed of short-term movements are random preddeterminirovanny character?

In fact, such a paradox can exist. The system can be random in the short term and long-term preddeterminirovannoy be. An example of such systems in nature are human bronchi.

There is no short-term patterns and repetitive cycles, which would have been significant with a predicative point of view. Patterns of prices and the indicators that traders use to trade, can easily be found in any set of random numbers. Thus, the chance to predict future prices in the short time frame with the help of technical analysis are approximately equal to the probability of predicting loss of a game of roulette.

In the book, Edgar Peters was a four-year history of research on every tick S & P. It was concluded that, although short-term fluctuations are not completely random, but preddetermenirovanny component in them is vanishingly small. Therefore, "it is extremely unlikely that frequently sells (for a short time interval), the trader can actually make a profit for a long time." It was also found to lack of data on the intraday cycles.

As far as I know the literature, it is not just a personal opinion. This is a scientific fact. Traders ignore it putting themselves at risk finansovumu. Does this mean that the market fluctuates randomly, and all traders are doomed to lose In consequence the cost of trading? No.

Traders can use a long-term trend component of the market for a statistical advantage. This is exactly what they are doing trend-tracking systems. This explains why good trend-tracking systems that are traded on markets, diversified, profitable from year to year, while the traders, who work within the day, suffer losses in the long run. To be a successful trader you need to put yourself in the position of the owner of the casino. At any rate the owner of the casino has a statistical advantage. Although casinos may suffer short-term losses, the more players place their bets, the more chance of winning the casino. If you trade approach, which has a statistical advantage, and if you follow it strictly (a big if), as a casino, you can not suffer losses in the long run.

My estimations regarding trader success give the following approximate result - one third of the success depends on the system, one-third of the choice of markets and one-third of the discipline with which the trader should be on your system. We can never know in advance, at which point our system will bring its statistical advantage. The best thing we can do is to create it without adjustment under the historical data and test it. If the system under perepodognana historical data, the test is useless and even harmful.

An easy way to avoid the fitting - this is to use the same rules for all markets and to test the system on the maximum number of possible markets. If the system makes a profit on many markets for a long time, it is likely that such a system is not fitted.

To increase your chances and risk reduction is very important to optimize your system's portfolio and account size. As a statistical advantage is obtained from following the trend, you can increase your chances of concentrating on markets with a more pronounced trend component. I wrote the book «Trendiness in the Futures Markets», which is a comprehensive study of 29 popular tendency of markets to trend in a range of time windows from 5 to 85 days.

Although we can measure the propensity of various markets to trend in a historical perspective, but we can not know for sure what the market will have the most pronounced trend in the next half-year year. Therefore, to reduce the risk of short-term, we need to diversify - as well as concentrate. My own research has shown that the optimal portfolio for a trend-follower of the trading system should contain from 10 to 20 markets. Personally, I trade 19 different markets with different trend-tracking systems.

Another aspect of risk management is to control the ratio of potential losses to the size of the account. As the losses and profits are closely related, to reduce losses to get a smaller profit. It is impossible to estimate the optimal composition of the portfolio as long as you do not define a common historical losses when trading a portfolio. I think that the start size of the portfolio should be no less than twice the size of the maximum losses on the portfolio plus a margin.

An alternative to a strict scientific approach is the way of trade, used by most traders. They have no idea what gives them a way to trade a statistical advantage. They assume that if something is written in books or come from the renowned guru or costs a lot of money, then it must be good. They use vysokosubektivnye methods which in principle can not be are tested. They are too lazy to create any hard and fast rules and to correct their testing. If you found yourself in this description, do not be surprised that trade brings you a loss. For you may be happy to trade, but do not forget that the pleasure has to pay.