To finish the piece for a mechanical system

Kent watson welcomes to share his passion of casino game with everybody.
- A few months - long positions.

Here, as in technical analysis, wave period is the time between two adjacent extrema. This is convenient because it is usually only interested in bull phase increases, and bears - the phase of the wave decay in the conventional sense.

5. Determination of oscillations of different time scales are sequentially from the smallest scale. Algorithm and the methods of analysis of market movements are directly linked with the above assumptions. Defining here, obviously, is the method of determining the oscillations of different time scales. The effectiveness of the solution of this question depends on the efficiency of the whole problem solving.

Analysis methods skolzyaschego medium and filtering methods using digital filters showed that the use of them is entirely futile, since they are effective only in the case where the period and amplitude of the vibrations required almost no change over time. Therefore, for this purpose we have developed an original numerical heuristic method, which we call the "method of soft smooth." The same method is used to estimate the errors of the empirical values of wave processes.

Feature of the method is that it slightly distorts the natural and does not generate false laws component. The effectiveness of the method is independent of the parameters of the stationary conditions of the wave process. The smaller discrete measurements, the higher the efficiency of the method. Smoothing the sequence of random numbers with a constant average number of breeds with the autocorrelation coefficient of about 0.15 (ie almost random series) and the variance is 3 times smaller than the variance of the original series. Consequently, the smoothing procedure can be repeated, and then we get a series with minor fluctuations around zero.

In comparison, using a single random number smoothing moving average by three points, we get a series with the same variance, but with the autocorrelation coefficient 0.65, ie it will contain a false wave process. What it can lead in practice, it is hardly explain. To assess the state of wave processes (sredneperiodnyh, and long-wave Elliott), and to receive support and resistance lines are used regression models containing a small number of harmonic functions. Models are known by the inclusion of variables provided a statistically significant reduction in the residual variance. Certain periods of the harmonics is by analogy with the spectral method, but not identical with it. Harmonic functions are used to determine the initial approximation of the process state.

It turns out that in the next approximation is almost always more effective is either a model that includes a small number of empirical, generally - not orthogonal functions, or finite-difference analog of the nonlinear differential equation of low degree. These models are particularly useful for the assessment of long-period oscillations.

Assistant Trader

We call our system "Assistant Trader." It can be used in three modes - Mechanical (in this case, the trader has only a mute function) polumehanicheskom (trader works on a strictly defined algorithm) and offline (trader uses the system only as a tool for decision-making).

We assume that the completion of our methodological principles of the trading system will improve the result of a few times, ie get get 300-400% per annum. And it is - only a small part (about 10%) of the profits that could be obtained by a mechanical system of "hindsight", ie having future value prices. Although hardly the real income can be raised significantly above 300-400% - obstacle to this is significant error empirical values of wave processes, which can not be reduced by any modern methods of treatment. So we do not face the prospect of a win-win trade system, which would soon become obsolete itself. Obviously, a good mechanical system on the market can only work in one piece. If the operation manual or polumehanichesky, it is most likely that the decisions taken by different traders, almost will not overlap if the number of traders using this system does not exceed the number Nt = Px / P, where Px - potentially available income (which can be equal to 3000%), P - the maximum profit of the system in the mechanical mode.

Our trading system can work in a wide range of free capital - from $ 100 thousand to $ 10 million system provides a stable income (at least 50% per year), even under the most unfavorable conditions, when the portfolio contains of one denomination. Finally, perhaps the most important thing is that set out a systematic approach to the full can be used for long-term investment objectives and in general for a wide range of economic issues, including the assessment of the macroeconomic situation in the country, region or the world. Effectiveness of the solution of such problems is tested by analyzing income and expenses in Russia in the early 90's.