To finish the piece for a mechanical system


The development of the trading system, we started four years ago in two directions. The first way was based on the correlation theory of random processes and was a dead end. The second is based on the idea of one of the authors previously performed well in the analysis of geophysical, medical, biological and economic processes, so we have no doubt in the success of ...

We walked to her four years ...

It took four years of persistent search, and many months of testing intermediate versions before results were obtained, worthy of attention. Development of the first mechanical trading system (MTS «MAGUS») came together with TruSoft.

To illustrate the effectiveness of the latest version of the figure shows the rate of profit for 2002 to a mechanical system (red line) and the S & P 500 (green line). Data for other years (1999-2001), for comparison, are given in the table.

In calculating the rate of return included commission expenses ($ 20 for the purchase / sale of the package and $ 0.05 for each purchase of shares), and restrictions on the amount of the purchased package - no more than 3% of the number of shares sold in the previous day and the amount of $ 25,000. The distribution of daily necessary capital has a pronounced right asymmetry, ie Mx value must be within a few days. Profit as a percentage of V obtained per transaction. The rate of profit in the figure is obtained, provided that all purchases are implemented, the proposed system (on some days their number may be greater than 10). The figure shows the rate of return over five months in 2002 for worst case, when every day to buy shares of only one company, randomly selected from the daily list that the results obtained in the figure.

Random selection was repeated 5 times, the figure shows the worst case. Despite the fact that the results for this year has been the worst in the last four years, but even in this case, our MTS significantly "replays" the market.

We think that the results are extraordinary, so we decided to introduce traders and financiers with our system. On what principles is MTS.

These four basic principles.

Principle 1. The market is seen as an open complex heterogeneous system, in which there are always fluctuations in the parameters of state of different time scales L0, L1, L2, ..., Ln. Li +1 scale fluctuations due to changes in the parameters of scale fluctuations Li. If conditions at the outer boundary of Li +1 change abruptly at random times tj or naturally over time, but in a random variable. Yj, then change the internal characteristics of Li +1 (subsystems and properties Li) will represent a random process. It is important to distinguish between changes due to the nonlinear nature of the process, from a random process. In the first case, the changes can be very fast, but it is possible to predict the state of the system at time tj +1, if its condition prior to tj. In the second case, the prediction is impossible in principle.

Principle 2. To describe and predict the market to use recursive method description [1]. This means that the description of Li +1 is sufficient to know the properties of subsystems Li. Model of the system with a higher hierarchical level can be, and often more simple than the model of the lower level.