The system of "dynamic portfolio speculator"

This article presents the results of a statistical study of the dynamics of stock prices in the U.S. market, describes a device semiautomatic trading system and reviewed the results of its tests in the period from September 2000 to September 2001

That we have researched and found out that

To better understand the processes occurring in the market, and identify patterns that can be used to build a trading system, we have investigated the following:

- Occurrence, duration and amplitude of the fluctuations in the price of quasiregular (swings);

- Filters for regular scan of the market;

- The possibility of price spikes (gaps) and their direction in relation to the probability of the prior trend;

- Determination of the historical phases of market activity on the values of swings;

- Correlation of stocks in the major indexes (such as S & P 500).

As a result, it was found:

1. At any given time there are hundreds of stocks, the dynamic characteristics of price changes which are very stable for several months. These shares may be collected and, based on this selection, you can build a trend-tracking trade system.

2. These shares may be divided into groups according to the riskiness and potential profitability of speculation with them. For high-income group is characterized by a change in prices by more than 50% in the quarter, the least risky group - by 20-30%.

3. Filter to regularly scan the market include restrictions on the average volume of trading, the share price, the limit value of intraday changes. The parameters of the optimal filter for the selection of candidates vary little over time, and most of all depends on the phase of the market. Thus, there is reason to believe that the filter can produce the required number of shares for a long time.

4. Gaps (gaps) significantly increases the risks of working with stocks. Almost every action in her stories are significant (over 10%) jumps prices. To reduce these risks and reduce the impact of gaps in the portfolio's value is reasonable to apply the diversification (10-20 stocks) and to adhere strictly to the rules: do not have a position against the trend on the daily scale. So the risks resulting gepami may be reduced to a reasonable size.

5. State of the stock market can be divided into phases of the activity, ie, by the characteristic (not average, how is the construction of indices) largest price changes. Typically, in a phase of high activity of the market price of the shares come in varying degrees of regular, often correlated or anti-correlated, motion. During such periods of speculation are most effective. Risky speculation on trends in the market will reverse when the direction is not yet determined. All this gives grounds to construct an algorithm of capital in the different phases of the market.

6. Correlation of stock prices to market indices (eg, stocks in the known index S & P 500) is quite interesting. We investigated the change of the composition of the list and correlated stocks. It turned out that they depend on the phase of market activity (high activity correlated percentage share increases). Secondly, it is found that the correlation with the stock index is not constant over time. It can be concluded that the use of market indices to determine the position of a particular stock does not make sense - you need to focus on the trend inherent in the stock itself.