1, technical analysis trading model
Technical analysis trading model refers to the trading model that uses market trading data, such as opening price, closing price and trading volume. , and through the computer trading indicators for system search and testing, optimization. Its theoretical basis is mainly based on the existing traditional technology investment theories such as graphic analysis and moving average theory, and it has been tested by a lot of statistical analysis. The biggest advantages of this model are: eliminating the influence of investors' emotions on trading decisions, especially the subjectivity and blindness of judging major events; Avoid analysis errors caused by information asymmetry; Ensure the consistency of transaction analysis; It provides a risk control method for investors.
The following focuses on three trading models in the technical analysis trading model:
1) transaction model based on pattern recognition.
This kind of model is mainly based on traditional classic graphics, such as head and shoulder top, double bottom, triangle and so on. , to capture market trends and trading positions. However, in actual combat, there are still many problems: in terms of risk control, such as head-shoulder top, double bottom, triangle and other trading charts, according to the traditional trading point of view, the investment risk/return ratio is generally 1: 1, and managers will face huge fund net value risks in actual combat; Most of the analysis is based on subjective judgment and lacks objective judgment standards; At present, the number of users of technical analysis in domestic futures market increases, which leads to the increase of false signals in the form of classical charts; The classical chart analysis theory abroad is quite different from that in China. Lack of statistical data.
2) Transaction model based on trend tracking.
This model mainly captures the inflection point of the price according to the designer's data statistics, and then assumes that the trend will continue, and builds positions according to the trend direction, such as MACD, SAR, moving average and so on. This trading model is characterized by not buying at the lowest price or selling at the highest price, giving up the profits before and after the market, and the profits mainly come from capturing the middle part of a big market. Its ability to capture market inflection points varies according to the sensitivity of designers' design. Sensitive trading models respond quickly to the trend reversal, but there are also many wrong signals. The trading model with low sensitivity has slow response to trend reversal, less error signals and more profits before and after giving up. The disadvantage of this trading model is that it produces continuous losses when consolidating the market, which makes investors unacceptable. Therefore, the difficulty in designing a trend-following trading model is not to find a way to capture the trend, but to have a set of perfect trend confirmation and filtering principles in order to avoid risks. In addition, the trend-following trading mode requires futures fund managers to hold positions for a long time, generally more than 2-3 months, so it requires futures fund managers to have a set of psychological control methods suitable for the trend-following trading mode.
3) Trading model based on contrarian.
This model is based on the designer's data statistics, and then assumes that the market needs to be adjusted and open positions in the opposite direction. The difference between it and the trend trading mode is that the trend trading mode can be adjusted automatically, while the counter-trend trading mode often brings incalculable risks because it operates against the main trend, so this trading mode must have a set of stop-loss conditions.
2. Basic analysis of transaction model
The basic analysis trading model refers to a model in which traders use data information outside the market to investigate all the information that affects the basic economic relations, and make quantitative analysis of such factors to establish a database, from which they can judge the market equilibrium price and make investments. The main features of this model are: providing a good analytical basis for large-scale capital entry; Strong theoretical foundation, easy to be accepted by the investment public; It is not helpful for short-term and timing; It is difficult to collect information; Analysis lags behind the market price; This analysis is subjective.
Here are two basic analytical trading models, namely "value evaluation" and "evaluation integral".
1) value evaluation transaction model
Futures prices will have a mutual traction effect on spot prices. According to statistics, in recent 10 years, the correlation coefficient between soybean futures price and spot price in China is 0.9. As for the futures price generated by the futures market, the participants in the futures market include spot traders and speculators, who have their own judgments on the futures price of the same commodity. However, because most of the participants in the mature futures market are speculators, the trading volume of the futures market is often several times or dozens of times that of the spot market, so the futures price is not only determined by the spot price and storage cost, but also includes the capital pricing part. Therefore, as the basic analytical trading model of futures funds, it also includes the speculative factors in the futures market: futures price = (spot price+storage cost) × speculative coefficient. Speculation coefficient is determined according to unexpected events and market speculative funds.
2) Overall evaluation transaction model
The main disadvantage of the basic analysis transaction model is that the information asymmetry caused by the difficulty of information collection lags behind the market price and is subjective. However, with the development of information technology and the improvement of trading system, the fair sharing of information will further narrow the information asymmetry, and it is relatively easy to obtain the latest information. The difficulty is how to distinguish authenticity and priority in information processing and overcome the influence of excessive subjective judgment. The main steps of integral evaluation trading mode are as follows:
First, determine the analysis factors
In order to maintain the comprehensiveness of analysis and statistical factors, the number of analysis factors, whether long or short, should not be too small, generally not less than five. For example, supply and demand analysis factors, taking soybean futures as an example, include: predicted planting area and actual planting area factors; Predicted yield and actual yield factors; Import and export volume of soybeans; Soybean crushing capacity; Inventory factors; Emergency factors, etc.
Another example is the cycle analysis factor, taking soybeans as an example. Periodic analysis factors include: March-April-Sino-American soybean sowing date, planting area prediction factors. At the same time, new South American soybeans began to go on the market, and the price was at the bottom. May-August-The weather and yield of soybeans in China and the United States are the main analysis and prediction factors. With the arrival of the peak consumption season, the price rose slowly compared with the previous period. In July and August, soybeans were affected by fluctuations such as bluish yellow and high temperature weather, and the price reached the annual peak. Around September-165438+1October-the actual soybean harvest in China and the estimated soybean planting area in South America. After June-10, due to the listing of new soybeans in China and the United States, the price once again fell back to the lowest price area of the year.
B, determine the time period of analysis
No matter which transaction model analysis method, it needs enough statistical analysis sample data to ensure the reliability of statistical results, so it needs to go through more than one cycle, such as the growth cycle of agricultural products and the economic cycle of metals. , which should include unexpected events or political factors to test the ability of the transaction analysis model to cope with and control risks.
C, determining the fractional value
The methods to determine the score can be ordinary positive and negative score method, weighted score percentage method and so on. The bullish factor score is positive, the bearish factor score is negative, and the factors without obvious bullish or bearish tendency are taken as 0.
D. Calculate the score result
Accumulate the scores of each influencing factor to get the scoring result. If the score is positive, the market trend is mainly upward. If the score is negative, the market trend is mainly downward; If the score is 0 or close to 0, the market will be in a consolidation state.
E, score tracking system
With the occurrence of different events and the change of time, various factors have different effects on the price. For example, the impact of unexpected events on prices will gradually fade with the changes of events, so it is necessary to constantly adjust the score values of various factors, determine the score results, and adjust the decision-making results of trading models.
3. Mathematical measurement transaction model
Mathematical econometric trading model refers to the model that designers make a lot of statistical analysis of historical trading data according to modern investment theory, and find out certain laws from it, and invest in the market when there are deviations or special circumstances, such as arbitrage trading model and gap trading model.
From the user's point of view, there are mainly the following two types: one is an analytical trading model, and the other is an operational trading model. There are great differences between the technical analysis trading model and the basic analysis trading model:
1, the analysis of the trading model focuses on foresight, which is an advanced analysis of market trends; The operational trading model focuses on the reactive trading decisions that should be taken when a certain price appears in the market.
2. Analytical trading model pays attention to individual income, requires high accuracy for a certain market, and ignores the analysis of unfavorable market conditions; The operational trading model pays attention to the overall benefit in actual combat, and requires the trading model to evaluate the income results produced by all situations in the market as a whole.
3. The biggest difference between the two is that the actual operators have to face pressure from all sides, including the market, investors and fund managers themselves. Therefore, the design of the model should also include how to control psychological stress factors through some methods and effectively implement the signals sent by the trading model.