1. For the establishment of the largest single and overall trading positions, I think it is more appropriate to determine the fixed proportion of total funds by combining the average oscillation amplitude of a single variety.
2. Timing of opening positions. Now that we have decided to follow the trend, the price should be our most concerned indicator. We can use the price to break through a certain high point as a signal to open a position, but this signal to open a position can be filtered out by other different conditions, such as the distance between the price and the moving average.
3. Selection of the number and timing of adding positions. As for jiacang, it is a difficult operation step to master. I initially think that adding positions should be carried out after the trend is established. Can this pressure or support be directly used for technical analysis? Will this cause great conflict to the whole programmed trading system? Because this is subjective after all, and it is difficult to quantify before operation, it deviates from the original intention of the programmed trading system, and whether it is feasible is debatable.
4. Setting of stop loss position. Stop loss is an important part of the whole trading system. Similar to determining the maximum position, we can set the stop loss as a fixed proportion and treat it differently according to the different characteristics of different varieties, and adjust the stop loss point after adding positions.
5. Close the profit position. As for the liquidation, because we follow the trend, we can't let small price fluctuations affect the overall operation, but it is still difficult to define whether it is a small callback or a big adjustment or a price reversal in quantity and energy. Of course, using statistical methods, we can calculate the fluctuation of historical data and its corresponding price performance, but the significance of such historical data for future operation deserves further verification.
Generally speaking, the success rate of a trading system designed with the idea of following the trend must not reach 50%, because the movement characteristics of the market will inevitably lead to such a result, but when the trend is clear, the profit rate is likely to be high. If the transaction cost is not considered, the futures market itself is a zero-sum market, and the money earned by those who make money is the same as the money lost by those who lose money. For a long time, it has been controversial whether such a market can formulate a trading system, because historical experience has proved that no matter how successful traders are, they will repeat themselves in this market, and as long as they trade, the market will inevitably carry out. Anything can happen before traders quit the market, and the trading system that can make money now may not work well in the future. Therefore, there is no trading system that can be tried and tested at any time and under any circumstances in this market, so it is very important to constantly verify and correct the trading system.
Finally, the biggest advantage of programmed trading is that it can help system users overcome the weakness of human greed and fear to the greatest extent. At present, the main products that support the programmable trading of financial products such as stock futures and foreign exchange are: tradeestablishment (master trader) international first-class trading software, but the cost is relatively expensive, and the software usage fee is 298$/ month. If you want to study historical materials, you have to buy them separately.
MT4 (to be upgraded to MT5 soon) is powerful and free to use. Because it is free, the number of its users has increased rapidly, and there is a tendency to surpass tradestation.
Esignal is more popular internationally.
Amibroker's Old Trading Software (1) Value Discovery Type
(2) Trend-chasing type
(3) Types of high-frequency transactions
(D) Low latency arbitrage (I) Input-output analysis of system design
(2) Principles of system design
1. Accuracy.
2. stability.
3. simple.
(3) Steps of system design
1. Suggestions on trading strategy.
2. Programmatic trading strategy.
3. Testing of programmed trading system.
4. Optimization of programmed trading system. The domestic securities market has only a history of 20 years, and the market maturity is still far behind that of Europe, America and other developed areas in Asia and the Pacific. Moreover, some characteristics of the domestic market also limit the development of programmed transactions in China.
First of all, the domestic trading places are relatively single, and stocks are only traded on exchanges. In foreign countries, a lot of liquidity exists outside the exchange, such as large and small dark pools, ECN and so on. Even the same stock will be traded on multiple exchanges. At present, only about 25% of NYSE stock trading volume is executed through NYSE. Many foreign execution algorithms are tailored to this situation, such as various intelligent routing algorithms, but these algorithms are useless in China.
Secondly, the T+ 1 delivery system in the domestic market cannot implement a large number of intraday trading strategies, let alone high-frequency trading strategies. In addition, short selling is not allowed in the stock market, the market maker system is lacking, the products available for trading are single, and the trading instructions are not perfect, which is not conducive to the formulation of programmed trading strategies.
Nevertheless, we are optimistic about the development prospect of programmed trading in China. Since the official launch of Shanghai and Shenzhen 300 stock index futures in April this year, a large number of programmed arbitrage strategies have been released and created amazing trading volume. Figure 14 shows the comparison of monthly trading volume between Shanghai and Shenzhen 300 stock index futures and S&P index futures (including ordinary contracts and mini contracts). As can be seen from the figure, if the exchange rates of the two currencies are converted, the domestic stock index futures trading volume is basically at the same level as the S&P index. Standard & Poor's index futures are market-oriented, while stock index futures are not yet open to institutional investors in China. You can imagine the potential of the domestic market after the market is fully opened in the future. Combined with the current situation in China, we believe that programmatic trading is expected to flourish for several reasons:
1. Arbitrage trading between stock index futures and ETF needs more algorithm support, because similar trading strategies involve the execution of a basket of stocks, and effective algorithms can greatly reduce the execution risk.
2. There are few services for domestic brokers to execute algorithms. At present, in the domestic stock market, institutional investors directly place orders through the direct entry channel provided by brokers, but brokers do not provide large-scale algorithm additional services, and there is still broad room for development in the future.
3. Other potential markets. Commodity futures, warrants and other markets also implement the T+0 delivery system, which is also a potential market for programmatic trading. In fact, many investors engaged in short-term trading (trend tracking and reversal) have developed various programmatic trading platforms and strategies, but the specialization and scale need to be improved. The programmatic trading of commodity futures and stock index futures is also the focus of the development of programmatic trading.
4. Talent advantage. Programmatic trading usually requires talents with solid mathematical foundation and excellent programming ability, and there is a good talent reserve in China. More and more foreign quantitative funds set up branches in China to hire talents to research and develop algorithms and strategies locally, which also proves this point.