The system of "dynamic portfolio speculator"

Special Technique allows the trader to follow a list of 100 candidates, and a portfolio that includes up to 20 open positions.

The worst enemy of the trader - the trader

The worst enemy of the trader - is himself. Who knew it - he is not a lost cause. People by nature are impulsive, greedy and cowardly. So they often do silly things. And wean make nonsense can not everyone. But we must try to do it at least for those who want. Well, those who are "just bad luck", who are willing to admit our weaknesses, we can help. To systematically work through the system, which we preach, do not need much: patience and perseverance. These simple qualities are known as "the golden rules of the market." And for practical work is required and a good broker with the appropriate service. Russian investors, we recommend (and use ourselves) brokers with Russian-speaking staff.

Testing the system "Dynamic portfolio speculator" was held in a rather complicated for the exchange period. Transition to decimal quotes changed intraday market and killed many deytreyderskie technology. This is all against a background of falling market, long-term change in trend growth. And the nature of the fall during this time also changed and alternated with periods of local growth. However, the system and traders passed this test.

A little explanation about the purpose of testing and its process. The purpose of testing was to show the performance of the proposed semi-automatic system in different phases of the market, when many thought that the vibrations reached the frequency and chaotic that the work on the daily scale should not give positive results.

The testing process was: trader model performed surgery on a demo account, carefully following the instructions. When testing has been verified the entire processing chain to ensure system health - from the selection of shares prior to the opening of specific items. As a result, the high degree of reproducibility of the model operations. During the whole process of the trader reflected on a special website.

During the tests, the program was made only one addition: in November 2000, the unit was improved calculation of capital - to reduce the risk imposed greater restrictions on the use of credit (shoulder). Thus, we can say that we have been able to maintain the system in working condition for the duration of testing.

With this program, the trader showed results similar to those expected. Stocks chosen for the work, which provided for testing performance is no less than 30% in the quarter, as a result of the trader in our program - about 200% per year, which, considering the reinvestment, modeled by getting the first 60% of the profit corresponds to the potential of selected candidates. The quantitative results are proof of the effectiveness of the control technique used for selecting candidates.

The distribution of income for the year came uneven for a simple reason: the trader twice was on vacation for two weeks or once a service break for almost a month. At each position is fully closed, and a pipeline to begin anew. In addition, the work program and Trader clearly had periods of stagnation, caused by a sharp reversal of the market.

A few words about the statistical properties of the result. The system did not show high efficiency. The rate of return (Profit factor) was only 1.5% at 60 losers (Table 1), which, we think, is quite typical of the trend-tracking systems used for statistical trade.

Income, as can be seen from the table, obtained mainly in short, rather than long positions (respectively, 159.000 and 26.000). Given the graphics market index during the same period, this is not surprising.

Scan the market and updating the list of shares for the time of testing were conducted three times, with the result list is updated almost completely. Trading was fairly active and massive: all positions were opened to 229 shares held in 1644 deals, a complete turn in the money to $ 26 million; portfolio consisted of a maximum of 23 stocks (on average 10-15).

In parallel with the testing we train traders. They are pretty easy and quickly learned how to work with "Dynamic Portfolio speculator." Despite the seemingly complex flowchart diagram of the control signals and dual circulation of lists of candidates, the technology of all participants in the process is quite simple and can be debugged at a technical level in a month. Duration of training all comers - 1-2 months. Experience for self-management can be purchased for 1-6 months.

Conclusions and

How long can it work? It seems that the "Dynamic Portfolio speculator" will work for a long time without any fundamental changes, which can not be said of the other, more automatic programs. After failed to link in one portfolio shares with different rates of change of prices, we find another interesting application of the system. It may help with efforts to restore the fallen value of investment portfolios. An interesting result was the selection of shares is not to maximize revenue and to further reduce risk. Stocks, of course, others were selected, change the setting of the program and the speed of work on it, but as a result of fluctuations in portfolio value decreased, and overall management has become more reliable.

Thus, we believe that we have managed to create a trading system to verify the practical ability of the effective work during the year. We expect to continue to work on it to improve and adjust the settings in the system software and all the technology in response to changing market conditions. We do not plan to modify the system in a more automatic. We're just going to create other systems.