Quantitative transaction backtest management system and method
By using the strategy expansion unit in the quantitative transaction backtest management system to automatically expand and combine quantitative transaction strategies, the problems of cumbersome editing and difficult strategy optimization in the existing system are solved, and efficient and automated strategy launches and users are achieved decision making.
Patent Information
- Application Number
- CN202410971144.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-08
AI Technical Summary
The existing quantitative transaction backtest management system requires manual participation in strategy editing and modification, which is cumbersome and time-consuming, and it is difficult to find the optimal strategy, resulting in waste of trading opportunities.
Design a quantitative transaction backtest management system, and use the preset algorithm to expand the preset parameter types in the quantitative transaction strategy through the strategy expansion unit, generate multiple extension conditions, and combine them with other conditions to automatically perform the backtest task and output A strategy that meets user preset conditions for users to make decisions.
It realizes automated strategy expansion, improves efficiency and speed, reduces waste of transaction opportunities, improves user experience, and avoids the complexity of manual operations.
Smart Images

Figure CN118917854A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and in particular to a quantitative trading backtesting management system and method. Background Art
[0002] Quantitative trading refers to the process of using computer technology to complete transactions based on formulated quantitative trading strategies. Quantitative trading greatly reduces the impact of investor sentiment fluctuations and avoids making irrational investment decisions when the market is extremely enthusiastic or pessimistic.
[0003] Backtesting refers to the process of using historical real data and simulated transactions to evaluate and verify the effectiveness and feasibility of quantitative trading strategies. Backtesting can provide feedback on quantitative trading strategies and facilitate manual decision-making on strategies.
[0004] At present, in the related technology, after formulating a quantitative trading strategy, the user will enter the backtesting management system to perform the backtesting task and obtain the backtesting results. The backtesting results can indicate the performance of the quantitative trading strategy in the historical environment and in future transactions, which can be reflected in various indicators, including yield, risk indicators, etc. After receiving the backtesting results, the user can make a decision on the strategy. If it is appropriate, the strategy can be deployed in the real market to automatically complete the transaction. If it is not appropriate, it can be manually modified and the backtesting can be continued until a satisfactory strategy is obtained.
[0005] It can be seen that the relevant technology requires manual modification of strategies, which is cumbersome and time-consuming, and it is not easy to find the optimal strategy. In the trading market with very high requirements for trading speed, many trading opportunities may be lost, and the user experience is not good.
[0006] There are also some systems in the related art that allow users to edit multiple strategies in advance, and then use backtesting to recommend the best strategy. This method still requires users to participate in manual strategy editing, which is time-consuming and labor-intensive, and the user experience is not good. Summary of the invention
[0007] In view of this, the present application discloses a quantitative trading backtest management system. The system includes: a control unit, which is used to control a strategy expansion unit, a backtest unit and an output unit to perform their respective tasks; a strategy expansion unit, which is used to respond to the instruction of the control unit, obtain the quantitative trading strategy edited by the user, the quantitative trading strategy includes at least one condition, the quantitative trading strategy is split to obtain the condition, for each of the conditions, when the quantitative parameter included in the condition is a preset parameter type, the quantitative parameter is numerically expanded according to a preset algorithm to obtain multiple extended values, and multiple extended conditions are generated based on the multiple extended values, and each of the extended conditions is combined with other conditions other than the condition to obtain multiple extended quantitative trading strategies; a backtest unit, which is used to respond to the instruction of the control unit, perform a backtest task for each of the extended quantitative trading strategies, and obtain a corresponding backtest result; the backtest result indicates the trading effect of the extended quantitative trading strategy; an output unit, which is used to respond to the instruction of the control unit, and output N extended quantitative trading strategies and their corresponding backtest results that meet the output conditions preset by the user based on the backtest result for the user to make a decision.
[0008] In some embodiments, the preset parameter types include: price type, price fluctuation type, and trading volume type; the strategy extension unit is further used to: when the quantification parameter is a price type, a first numerical range is obtained according to the numerical value of the quantification parameter and a first preset ratio, a first step length is obtained according to the numerical value and a second preset ratio, and multiple extended numerical values are obtained within the first numerical range based on the first step length; the second preset ratio is smaller than the first preset ratio; when the quantification parameter is a trading volume type, a second numerical range is obtained according to the numerical value of the quantification parameter and a third preset ratio, a second step length is obtained according to the numerical value and a fourth preset ratio, and multiple extended numerical values are obtained within the second numerical range based on the second step length; the fourth preset ratio is smaller than the third preset ratio; when the quantification parameter is a price fluctuation type, a third numerical range is obtained according to the numerical value of the quantification parameter and a preset deviation value, and multiple extended numerical values are obtained within the third numerical range based on the preset step length.
[0009] In some embodiments, the control unit is further used to initiate an expansion request to the strategy expansion unit in response to detecting that the user has completed editing the quantitative trading strategy; the strategy expansion unit is further used to expand the quantitative trading strategy in response to receiving the expansion request to obtain multiple extended quantitative trading strategies; the control unit is further used to send a backtesting request to the backtesting unit in response to receiving a backtesting request initiated by the user; the backtesting unit is further used to backtest multiple extended quantitative trading strategies in response to receiving the backtesting request.
[0010] In some embodiments, the backtesting unit corresponds to a distributed computing system; the backtesting unit, according to the number of extended quantitative trading strategies, applies to the distributed computing system for at least the number of computing nodes, and uses the number of computing nodes to execute the backtesting task in parallel to obtain the backtesting results of the quantitative trading strategy.
[0011] In some embodiments, the backtesting unit is further used to obtain historical real data; generate multiple data groups based on the historical real data; the data group includes first historical data and second historical transaction data; the timing of the second historical transaction is located after the first historical data; for each data group, obtain the historical predicted performance of the quantitative trading strategy based on the first historical data, the historical real performance of the quantitative trading strategy based on the second historical data, and obtain the historical performance deviation of the quantitative trading strategy based on the historical real performance and the historical predicted performance; obtain predicted data based on the historical real data, and obtain the future predicted performance of the quantitative trading strategy based on the predicted data; correct the future predicted performance based on the historical performance deviation to obtain the backtesting result of the quantitative trading strategy.
[0012] In some embodiments, the historical performance deviation is the difference between the historical actual performance and the historical predicted performance; the backtesting unit determines the sum of the historical performance deviation and the future predicted performance as the backtesting result of the quantitative trading strategy.
[0013] In some embodiments, the historical real data includes market data; the backtesting unit is further used to obtain forecast data by using a first forecasting model constructed based on Transformer when the market data indicates large market fluctuations; and to obtain forecast data by using a second forecasting model constructed based on LSTM when the market data indicates obvious market trends.
[0014] In some embodiments, the historical real data is multimodal data; before using the prediction model to obtain the prediction data, the backtest unit uses the multimodal processing model to process the multimodal data according to the time sequence of the multimodal data to obtain structured data with time sequence; when the multimodal data is a picture, the multimodal processing model is a SIFT model; when the multimodal model is text, the multimodal processing model is an NLP model; when the multimodal model is audio, the multimodal processing model is an ASR model; the structured data is used to input the prediction model to obtain the prediction data.
[0015] In some embodiments, the output conditions include maximum rate of return, minimum loss, and lowest risk; in response to the user configuring the maximum rate of return, the output unit is used to output the top N extended quantitative trading strategies with the maximum rate of return and their corresponding backtest results for the user to make a decision; in response to the user configuring the minimum loss, the output unit is used to output the top N extended quantitative trading strategies with the minimum loss and their corresponding backtest results for the user to make a decision; in response to the user configuring the lowest risk, the output unit is used to output the top N extended quantitative trading strategies with the lowest risk and their corresponding backtest results for the user to make a decision.
[0016] In the schemes recorded in any of the aforementioned embodiments, the strategy expansion unit can use a preset algorithm to expand the conditions containing preset type parameters in the quantitative trading strategy to obtain multiple extended conditions and combine them with other conditions to obtain multiple extended quantitative trading strategies, and then use the backtesting unit to backtest each extended quantitative trading strategy and output N strategies that meet the output conditions preset by the user for the user to make a decision. In this way, the backtesting management system can be used to realize automated strategy expansion and launch N strategies that meet user needs for the user to select the optimal strategy, without the need for the user to perform complex operations, with high efficiency, speed and good effect, and can avoid the waste of excessive trading opportunities caused by modifying the strategy.
[0017] The present application proposes a quantitative trading backtest management method. The method includes: obtaining a quantitative trading strategy edited by a user, wherein the quantitative trading strategy includes at least one condition; splitting the quantitative trading strategy to obtain the conditions, for each of the conditions, when the quantitative parameter included in the condition is a preset parameter type, numerically expanding the quantitative parameter according to a preset algorithm to obtain multiple extended values, and generating multiple extended conditions based on the multiple extended values, combining each of the extended conditions with other conditions other than the condition to obtain multiple extended quantitative trading strategies; performing a backtest task for each of the extended quantitative trading strategies to obtain a corresponding backtest result; the backtest result indicates the trading effect of the extended quantitative trading strategy; outputting N extended quantitative trading strategies that meet the preset output conditions and their corresponding backtest results for the user to make a decision.
[0018] In the scheme recorded in the aforementioned method embodiment, the preset algorithm can be used to expand the conditions containing the preset type parameters in the quantitative trading strategy to obtain multiple extended conditions and combine them with other conditions to obtain multiple extended quantitative trading strategies, and then backtest each extended quantitative trading strategy and output N strategies that meet the output conditions preset by the user for the user to make a decision. In this way, automated strategy expansion can be achieved and N strategies that meet the user's needs can be launched for the user to select the optimal strategy, without the user having to perform complex operations, with high efficiency and good speed, and can avoid the waste of excessive trading opportunities caused by modifying the strategy.
[0019] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] In order to more clearly illustrate the technical solutions in one or more embodiments of the present application or related technologies, the drawings required for use in the embodiments or related technology descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in one or more embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0021] The following is a brief introduction to the drawings required for use in the embodiments or related technical descriptions.
[0022] Figure 1 This is a schematic diagram of the structure of a quantitative trading backtesting management system shown in this application.
[0023] Figure 2 This is a schematic diagram of the structure of a quantitative trading backtesting management system shown in this application.
[0024] Figure 3 This is a structural diagram of a quantitative trading backtesting management system illustrated in this application.
[0025] Figure 4 This is a structural diagram of a quantitative trading backtesting management system illustrated in this application.
[0026] Figure 5 A flowchart of a quantitative trading backtesting management method shown in this application.
[0027] Figure 6 The figure is a flowchart of a method for expanding a numerical value shown in the present application.
[0028] Figure 7 The figure is a flowchart of the backtesting method for quantitative trading strategies. DETAILED DESCRIPTION
[0029] The exemplary embodiments will be described in detail below, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are only examples of devices and methods consistent with some aspects of the present application as detailed in the attached claims.
[0030] The terms used in this application are only for the purpose of describing specific embodiments, and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in this article refers to and includes any or all possible combinations of one or more associated listed items. It should also be understood that the word "if" used in this article, depending on the context, can be interpreted as "at the time of" or "when" or "in response to determining".
[0031] The present application proposes a quantitative trading backtest management system. In the system, the strategy expansion unit can use a preset algorithm to expand the conditions containing preset type parameters in the quantitative trading strategy to obtain multiple extended conditions and combine them with other conditions to obtain multiple extended quantitative trading strategies, and then use the backtest unit to backtest each extended quantitative trading strategy and output N strategies that meet the output conditions preset by the user for the user to make a decision. In this way, the backtest management system can be used to realize automated strategy expansion and launch N strategies that meet user needs for users to select the optimal strategy, without the need for users to perform complex operations, with high efficiency, speed and good effect, which can avoid the waste of excessive trading opportunities caused by modifying the strategy.
[0032] The quantitative trading backtesting management system is explained below with reference to the accompanying drawings.
[0033] See also Figure 1 , Figure 1 This is a schematic diagram of the structure of a quantitative trading backtesting management system shown in this application. Figure 1 As shown, the quantitative trading backtesting management system (referred to as the system) 100 includes a control unit 110, a strategy expansion unit 120, a backtesting unit 130, and an output unit 140 that are communicatively connected.
[0034] The control unit 110 is a control module of the system, which is used to issue some instructions to control other units to perform corresponding operations, similar to a hub. The unit can be implemented by deploying software code on hardware.
[0035] The strategy expansion unit 120 is what distinguishes the system from other quantitative trading systems. This unit can be implemented by deploying software code on hardware.
[0036] The unit can obtain the quantitative trading strategy edited by the user in response to the instruction of the control unit, wherein the quantitative trading strategy includes at least one condition.
[0037] The quantitative trading strategy is split to obtain the conditions. For each of the conditions, when the quantitative parameter included in the condition is of a preset parameter type, the quantitative parameter is numerically expanded according to a preset algorithm to obtain a plurality of extended numerical values, and a plurality of extended conditions are generated based on the plurality of extended numerical values. Each of the extended conditions is combined with other conditions other than the condition to obtain a plurality of extended quantitative trading strategies.
[0038] The control unit 110 may issue an instruction to the policy extension unit upon receiving a backtest request initiated by the user or in response to the user completing policy editing.
[0039] The preset parameter type library can be maintained in the strategy extension unit. After receiving the instruction, the edited strategy can be obtained, and several conditions for splitting the strategy can be obtained. After obtaining the conditions, they can be matched one by one with the parameter types in the preset parameter type library. The matching indicates that the condition is a condition that needs to be extended. Then, the quantitative parameters contained in the condition can be expanded to obtain multiple extended conditions, which can be combined with other conditions to obtain multiple extended quantitative trading strategies.
[0040] The preset parameter types can be set according to the requirements, and can be some numerical parameters that are convenient for expansion. For example, the preset parameter types can include price types (including opening price, current price, market price, market volume, etc.), price fluctuation types (including the highest daily increase, the average daily increase, etc.), and transaction volume types (including transaction volume, transaction amount, volume ratio, etc.).
[0041] The preset algorithm is not limited in this example, and can achieve the purpose of expanding the value. For example, a preset number of addition, subtraction, multiplication, and division operations can be performed on the value to obtain multiple expanded values, or a reasonable range of values can be determined based on the value, and multiple expanded values can be obtained within a reasonable range of values, etc. Subsequent embodiments will propose some better algorithms that are more in line with the logic and habits of trading strategy formulation, improve the expansion effect, and enhance user experience.
[0042] For example, the quantitative strategy is: condition 1: the current price reaches 10 yuan, condition 2: MACD golden cross, buy 500 shares. The preset parameter types include: price type, change type, and trading volume type.
[0043] The parameter of condition 1 is that the current price belongs to the preset parameter type. Then according to the algorithm, 10 yuan can be expanded to get 6 values of 9.7, 9.8, 9.9, 10, 10.1, 10.2, and 10.3. The current price can be used to get 6 expansion conditions for the 6 values, and then combined with the MACD golden cross to get 6 extended quantitative trading strategies.
[0044] It should be noted that if a quantitative strategy has multiple conditions containing preset parameter types, they need to be expanded one by one in the order of the conditions.
[0045] For example, the strategy includes condition 1 that the current price reaches 10 yuan, condition 2 that MACD forms a golden cross, and condition 3 that the trading volume reaches 100,000, and buy 500 shares.
[0046] The parameters of condition 1 and condition 3 are of the preset parameter type.
[0047] First, condition 1 is expanded to obtain 6 extended quantitative trading strategies, and then condition 3 in the 6 extended quantitative trading strategies is expanded separately to obtain 6*6, a total of 36 extended quantitative trading strategies.
[0048] If more conditions are included, the expansion method is similar, and eventually multiple extended quantitative trading strategies can be obtained.
[0049] The backtesting unit 130 is used to respond to the instruction of the control unit, execute the backtesting task for each of the extended quantitative trading strategies, and obtain the corresponding backtesting results.
[0050] The backtest results indicate the trading effect of the extended quantitative trading strategy. The backtest results can indicate the performance of the quantitative trading strategy in a historical environment and in future transactions, which can be reflected in various indicators, including yield, risk indicators, etc.
[0051] The backtesting task is not specifically limited in this example, and can refer to the backtesting method in the related art. In the subsequent embodiments of this application, a backtesting method combining the historical performance deviation of the quantitative strategy to correct the backtesting results will be proposed, which can improve the accuracy of the backtesting results, and then backtest the backtesting results that are more in line with the actual situation, thereby improving the user experience.
[0052] In the backtest unit, the backtest results of each extended quantitative trading strategy can be obtained based on historical real data and predicted data based on historical real data. For example, based on the previous example, the backtest results of the extended 6 strategies can be obtained.
[0053] The output unit 140 can respond to the instruction of the control unit and output N extended quantitative trading strategies and their corresponding backtest results that meet the output conditions preset by the user based on the backtest results for the user to make a decision. The N can be set according to the needs. For example, 3 or 5.
[0054] Before initiating a backtest, the user can configure some output conditions, such as the highest rate of return, the lowest risk, or the smallest loss, etc. The output unit can output N extended quantitative trading strategies that meet the output conditions and their corresponding backtest results through a display unit that interacts with the user based on the backtest results.
[0055] Users can choose M (M is less than or equal to N) strategies from N extended quantitative trading strategies and deploy them in real trading to achieve automatic trading. For example, in the previous example, the user has selected 1 to 3 strategies to deploy.
[0056] In some embodiments, the output conditions include maximum yield, minimum loss, and lowest risk;
[0057] In response to the user configuring the maximum rate of return, the output unit is used to output the top N extended quantitative trading strategies with the maximum rate of return and their corresponding backtest results for the user to make a decision;
[0058] In response to the user configuring the minimum loss, the output unit is used to output the top N extended quantitative trading strategies with the minimum loss and their corresponding backtest results for the user to make a decision;
[0059] In response to the user configuring the lowest risk, the output unit is used to output the top N extended quantitative trading strategies with the lowest risk and their corresponding backtesting results for the user to make a decision.
[0060] For example, based on the previous example, assuming that N is 3 and the output condition is the highest rate of return, the three strategies with the highest rate of return among the six strategies can be output.
[0061] The quantitative trading backtest management system illustrated in the embodiment of the present application can expand the conditions containing preset type parameters in the quantitative trading strategy by using a preset algorithm through the strategy expansion unit to obtain multiple extended conditions and combine them with other conditions to obtain multiple extended quantitative trading strategies, and then use the backtest unit to backtest each extended quantitative trading strategy and output N strategies that meet the output conditions preset by the user for the user to make a decision. In this way, the backtest management system can be used to realize automated strategy expansion and launch N strategies that meet user needs for the user to select the optimal strategy, without the need for the user to perform complex operations, with high efficiency and good speed, and can avoid the waste of excessive trading opportunities caused by modifying the strategy.
[0062] In some embodiments, some better expansion algorithms can be proposed, which are more in line with the logic and habits of trading strategy formulation, to improve the expansion effect and enhance the user experience.
[0063] The preset parameter types include: price type, price increase or decrease type, and transaction volume type;
[0064] The policy extension unit 120 is further configured to:
[0065] In the case where the quantization parameter is of price type, a first value range is obtained according to the value of the quantization parameter and a first preset ratio, a first step length is obtained according to the value and a second preset ratio, and a plurality of extended values are obtained within the first value range based on the first step length;
[0066] In the case where the quantization parameter is a volume type, a second value range is obtained according to the value of the quantization parameter and a third preset ratio, a second step length is obtained according to the value and a fourth preset ratio, and a plurality of extended values are obtained within the second value range based on the second step length; the fourth preset ratio is smaller than the third preset ratio;
[0067] In the case where the quantization parameter is of the fluctuation type, a third value range is obtained according to the value of the quantization parameter and a preset deviation value, and a plurality of extended values are obtained within the third value range based on a preset step size.
[0068] In the case where the quantitative parameter is a price type, the first value range can be obtained based on the value by using the lower limit value = value * (1-first preset ratio) and the upper limit value = value * (1+first preset ratio). The first step length can be obtained based on the value by using the first step length = value * second preset ratio.
[0069] The second preset ratio is smaller than the first preset ratio, so that multiple extended values can be obtained within the first value range. The first preset ratio can be set based on experience, such as 3% or 5%. The second preset ratio can be set based on experience, such as 1% or 2%.
[0070] For example, assuming that the first preset ratio is 3%, the second preset ratio is 1%, and the condition is that the current price reaches 10, the numerical range can be 9.7-10.3, the first step length is 0.1, and then 9.7, 9.8, 9.9, 10, 10.1, 10.2, 10.3, a total of 6 values, can be obtained, that is, 6 expansion conditions for the current price to reach these 6 values, which can be combined with other conditions to obtain an expansion strategy.
[0071] Therefore, when the quantitative parameter is price-related, no matter whether the price value is in the single digit, tens, hundreds, or thousands, a reasonable expansion range can be determined by using the ratio method, and a reasonable expansion value can be obtained to meet the logic of strategy optimization. It is easy to expand a better quantitative trading strategy, help users maximize their benefits, and improve user experience.
[0072] When the quantitative parameter is trading volume, the relevant calculation logic can refer to that of price.
[0073] For example, assuming that the third preset ratio is 3%, the fourth preset ratio is 1%, and the condition is that the transaction amount reaches 100,000, the numerical range is 97,000-103,000, and the second step is 1,000, so we get 97,000, 98,000, 99,000, 100,000, 101,000, 102,000, and 103,000, a total of 6 numerical values. We can get 6 expansion conditions for the transaction amount to reach these 6 numerical values, which can be combined with other conditions to get the expansion strategy.
[0074] Therefore, when the quantitative parameter is trading volume, no matter whether the trading volume value is in the thousands, tens of thousands, hundreds of thousands or more, a reasonable expansion range can be determined by proportion and a reasonable expansion value can be obtained to meet the logic of strategy optimization. It is easy to expand better quantitative trading strategies, help users maximize their benefits and improve user experience.
[0075] In the case where the quantitative parameter is of the increase or decrease type, the third value range is obtained based on the value by using the lower limit value = value - preset deviation value and the upper limit value = value + preset deviation value.
[0076] The preset deviation value and the preset step length can be set based on experience, for example, the preset deviation value is 0.5% or 0.3%, and the preset step length is 0.1% or 0.2%.
[0077] For example, assuming the preset deviation value is 0.3%, the preset step size is 0.1%, and the condition is an increase of 5%, the numerical range is 4.7%-5.3%. Based on the preset step size of 0.1%, we get 6 values of 4.7%, 4.8%, 44.9%, 5%, 5.1%, 5.2%, and 5.3%, which are a total of 6 values. We can get 6 expansion conditions for the increase to reach these 6 values, which can be combined with other conditions to get the expansion strategy.
[0078] Therefore, when the quantitative parameter is trading volume, a reasonable expansion range can be determined regardless of the size of the increase or decrease, and a reasonable expansion value can be obtained to meet the logic of strategy optimization, making it easy to expand a better quantitative trading strategy, help users maximize their benefits, and improve user experience.
[0079] In some embodiments, the backtesting unit and the strategy expansion unit are asynchronously operated and decoupled from each other, which can improve backtesting efficiency, save time for users, and thus improve user experience.
[0080] The control unit 110 is further configured to initiate an extension request to the strategy extension unit 120 in response to detecting that the user has completed editing the quantitative trading strategy;
[0081] The strategy extension unit 120 is further configured to extend the quantitative trading strategy in response to receiving the extension request to obtain a plurality of extended quantitative trading strategies;
[0082] The control unit 110 is further configured to send a backtest request to the backtest unit in response to receiving a backtest request initiated by a user;
[0083] The backtesting unit 130 is further configured to perform backtesting on the plurality of extended quantitative trading strategies in response to receiving the backtesting request.
[0084] See also Figure 2 , Figure 2 This is a schematic diagram of the structure of a quantitative trading backtesting management system shown in this application. Figure 1 Compared with the present invention, an interaction unit 150 for interacting with the user and a storage unit 160 for storing data are added. The interaction unit 150 can be similar to a client, can receive user operations, and can display some information to the user. The storage unit 160 can be any type of storage unit.
[0085] In this example, after the user completes editing, uploading or confirming the quantitative trading strategy in the interactive unit 150, the user can notify the control unit that the user has completed editing the quantitative trading strategy, and the unit can initiate an expansion request. After receiving the request, the strategy expansion unit can expand and send it to the storage unit 160 to store the expanded strategy.
[0086] The user needs to perform strategy backtesting at a certain point in time. For example, after editing many strategies, the user can trigger strategy backtesting through the interaction unit 150. The interaction unit 150 can notify the control unit that the customer needs backtesting. The control unit can send a backtesting request to the backtesting unit 130. After receiving the backtesting request, the backtesting unit 130 can obtain each extended strategy from the storage unit 160 and complete the backtesting.
[0087] This operation can use the idle time between the user initiating the backtest and completing the strategy editing to perform early strategy expansion, so that there is no need to expand the strategy during the backtest task, which improves the backtest efficiency, saves time for users, and improves the user experience.
[0088] In some embodiments, a distributed parallel computing method may be used in the backtesting unit to improve the execution efficiency of the backtesting task, save time for users, and improve user experience.
[0089] See also Figure 3 , Figure 3 This is a schematic diagram of the structure of a quantitative trading backtesting management system for this application. Figure 1 Compared with the backtest unit 130 and the distributed computing system 200. Figure 4 , Figure 4 This is a schematic diagram of the structure of a quantitative trading backtesting management system for this application. Figure 2 Compared with the backtest unit 130 and the distributed computing system 200 .
[0090] Figure 3 or Figure 4 The distributed computing system 200 shown in FIG. 2 provides distributed computing capabilities, and backtesting tasks are deployed in these nodes.
[0091] The backtesting unit 130 applies for at least the number of computing nodes from the distributed computing system 200 according to the number of extended quantitative trading strategies, and uses the number of computing nodes to perform backtesting tasks in parallel to obtain backtesting results of the quantitative trading strategies.
[0092] For example, if the number of extended quantitative trading strategies is 36, at least 36 computing nodes can be applied to the distributed computing system 200, and the 36 strategies can be sent to these nodes respectively. It should be noted that in order to improve computing efficiency, a backtest of a strategy can be performed through a combination of multiple nodes. For example, a backtest task can be calculated through 3 nodes, so in the case of 36 strategies, 108 nodes need to be applied to perform the task.
[0093] This can greatly improve the backtesting efficiency for multiple expansion strategies through distributed parallel computing, greatly saving users' time and improving user experience.
[0094] In some embodiments, a backtesting method is proposed that combines the historical performance deviation of the quantitative strategy to correct the backtesting results, which can improve the accuracy of the backtesting results, and then backtest the backtesting results that are more in line with the actual situation, thereby improving the user experience.
[0095] The backtest unit 130 is further used to obtain historical real data;
[0096] Generate multiple data groups based on the historical real data; the data groups include first historical data and second historical transaction data; the time sequence of the second historical transaction is after the first historical data;
[0097] For each data group, obtaining a historical predicted performance of the quantitative trading strategy based on the first historical data, a first historical actual performance of the quantitative trading strategy based on the second historical data, and obtaining a historical performance deviation of the quantitative trading strategy based on the historical actual performance and the historical predicted performance;
[0098] Obtaining forecast data based on the historical real data, and obtaining the future forecast performance of the quantitative trading strategy based on the forecast data;
[0099] The future predicted performance is corrected based on the historical performance deviation to obtain the backtest result of the quantitative trading strategy.
[0100] The historical real data can be set according to needs, including transaction data, news, weather, market environment data (such as market size, market growth rate, market development trend, etc.), stock data (price-earnings ratio, price-to-book ratio, market value, economic indicators (GDP growth rate, inflation rate CPI, etc.), technical indicators MACD, KDJ, CCI, etc.).
[0101] Historical real data is time series data, with a start time and an end time, and the time series is relatively long, for example, 1 year, 3 years, 5 years, etc. These data can be divided into multiple data groups, each of which includes two adjacent groups of historical data, which are referred to as first historical data and second historical data in this application to distinguish between the two. For example, the historical real data of 1 year can be grouped by month, and 11 groups are obtained, namely, January and February, February and March, March and April, until October and November, November and December.
[0102] For a certain data group, the possible future data can be predicted based on the first historical data, and then the possible future data can be used to use the backtesting method in the relevant technology to obtain the historical predicted performance of the strategy. Then the real second historical data can be used to use the backtesting method in the relevant technology to obtain the historical real performance of the strategy. Then the difference between the historical real performance and the historical predicted performance is used to obtain the historical performance deviation of the data group. Then the weighted sum of the historical performance deviations corresponding to each data group can be used to obtain the final historical performance deviation.
[0103] For example, for January and February, the prediction model can be used to predict the January data to obtain the possible February data, and then the predicted February data can be used for backtesting to obtain the historical prediction performance of the strategy. The performance includes at least the rate of return and several risk indicators. Then the real February data is used for backtesting to obtain the historical real performance of the strategy. By making a difference, the historical performance deviation of the strategy for January and February can be obtained. By analogy, the deviations of the 11 data groups are obtained and the average value is calculated to obtain the final historical performance deviation.
[0104] Then, the prediction model can be used to predict the historical real data to obtain the prediction data, and the future prediction performance of the strategy backtest can be completed based on the prediction data. After that, the sum of the historical performance deviation and the future prediction performance can be determined as the final backtest result of the quantitative trading strategy.
[0105] Therefore, the backtesting method that combines the historical performance deviation of the quantitative strategy to correct the backtesting results can improve the accuracy of the backtesting results, and then backtest the backtesting results that are more in line with the actual situation, thereby improving the user experience.
[0106] In some embodiments, there may be multiple prediction models used to predict future data based on historical data, and a suitable model may be selected for prediction based on market conditions to improve prediction effectiveness and efficiency, thereby improving backtesting accuracy and user experience.
[0107] The historical real data includes market data; the backtesting unit is further used to obtain forecast data using a first forecasting model constructed based on Transformer when the market data indicates that the market fluctuates greatly;
[0108] When the market data indicates that the market trend is obvious, the second prediction model built based on LSTM is used to obtain the prediction data.
[0109] For example, market data may include the market size of stocks, market growth rate, market development trends, etc. Historical real data also includes transaction data. These data are obtained through historical records and are used to train and test prediction models.
[0110] The first prediction model built on Transformer:
[0111] Transformer is a powerful neural network architecture that is particularly good at processing sequence data. In the trading market, historical transaction price sequences can be used as input data, and then trained using the Transformer model to predict the trend of transaction prices in the future. Transformer's self-attention mechanism can help the model better capture long-distance dependencies in sequence data, thereby improving the accuracy of predictions.
[0112] The second prediction model built based on LSTM:
[0113] LSTM (Long Short-Term Memory) is a recurrent neural network suitable for processing sequence data. In the trading market, LSTM can be used to learn the trend and periodic changes of historical trading prices and predict future price trends based on them. Through the memory unit and forget gate of LSTM, the model can better capture the long-term dependencies in the sequence data, while alleviating the problem of gradient disappearance, so that the model can better adapt to the complex changes in the trading market.
[0114] Therefore, in the backtest management system, when market data indicates that the market is volatile, the system will automatically select the prediction model built on Transformer, using its powerful sequence modeling capabilities to adapt to the ever-changing market; and when the market trend is obvious, the system will switch to the prediction model built on LSTM to better capture the changes in market trends. In this way, the appropriate model can be selected for prediction based on market conditions, improving prediction results and efficiency, thereby improving backtest accuracy and user experience.
[0115] In some embodiments, the prediction model can also process multimodal information, which increases the scope of application of the prediction model and enables predictions to be made based on more information, thereby improving effectiveness and efficiency, thereby improving backtesting accuracy and user experience.
[0116] The historical real data is multimodal data; before using the prediction model to obtain the prediction data, the backtest unit uses the multimodal processing model to process the multimodal data according to the time sequence of the multimodal data to obtain structured data with time sequence; when the multimodal data is a picture, the multimodal processing model is a SIFT model; when the multimodal model is text, the multimodal processing model is an NLP model; when the multimodal model is audio, the multimodal processing model is an ASR model; the structured data is used to input the prediction model to obtain the prediction data.
[0117] For example, historical facts include not only transaction data, but also multimodal data related to transactions, such as pictures, text, and audio. Before using the prediction model to make transaction decisions, the system needs to process these multimodal data to obtain structured time series data, and use these data to input the prediction model to obtain prediction data.
[0118] Assume that multimodal data related to transactions is collected. For image data, the SIFT (Scale Invariant Feature Transform) model can be used to process it, extract key points and feature descriptors, and then convert these features into time-series structured data for use as input for the prediction model. For text data, the NLP (Natural Language Processing) model can be used to process it, converting text information such as transaction-related news and analysis reports into structured time-series data. For audio data, the ASR (Automatic Speech Recognition) model can be used to convert transaction-related audio data into text, and then the NLP model can be used to process it and convert it into time-series structured data.
[0119] When the system is backtesting, before using the forecast model to obtain forecast data, the system will use these processed time series structured data as input, and determine the input to the forecast model built based on Transformer or the forecast model built based on LSTM according to market fluctuations to obtain forecast data for future trends. In this way, the system can make full use of the information of multimodal data and combine different processing models to more comprehensively understand the market situation, improve forecasting effect and efficiency, thereby improving backtesting accuracy and user experience.
[0120] Corresponding to any of the above embodiments, the present application also proposes a quantitative trading backtesting management method. Figure 5 , Figure 5 The following is a flow chart of a quantitative trading backtesting management method shown in this application. Figure 5 As shown, the method may include S502-508.
[0121] S502, obtaining a quantitative trading strategy edited by a user, wherein the quantitative trading strategy includes at least one condition.
[0122] S504, splitting the quantitative trading strategy to obtain the conditions, for each of the conditions, in the case where the quantitative parameter included in the condition is of a preset parameter type, numerically expanding the quantitative parameter according to a preset algorithm to obtain a plurality of extended numerical values, and generating a plurality of extended conditions based on the plurality of extended numerical values, combining each of the extended conditions with other conditions other than the condition, to obtain a plurality of extended quantitative trading strategies.
[0123] S506, performing a backtesting task for each of the extended quantitative trading strategies to obtain a corresponding backtesting result; the backtesting result indicates a trading effect of the extended quantitative trading strategy.
[0124] S508, outputting N extended quantitative trading strategies that meet preset output conditions and their corresponding backtesting results for the user to make a decision.
[0125] The explanation of the steps can refer to any of the previous embodiments and will not be described in detail here.
[0126] Through the embodiments recorded in S502-508, the conditions containing preset type parameters in the quantitative trading strategy can be expanded by using a preset algorithm to obtain multiple extended conditions and combined with other conditions to obtain multiple extended quantitative trading strategies, and then each extended quantitative trading strategy is backtested and N strategies that meet the output conditions preset by the user are output for the user to make a decision. In this way, automatic strategy expansion can be achieved and N strategies that meet the user's needs can be launched for the user to select the optimal strategy, without the user having to perform complex operations, with high efficiency, speed and good effect, and can avoid the waste of excessive trading opportunities caused by modifying the strategy.
[0127] In some embodiments, the preset parameter types include: price type, price increase or decrease type, and transaction volume type. Figure 6 , Figure 6 The figure is a flowchart of a method for expanding a numerical value shown in the present application. Figure 6 This is an explanation of S504. Figure 6 As shown, the method may include S602-S606.
[0128] The step of numerically expanding the quantization parameter according to a preset algorithm to obtain a plurality of extended values includes:
[0129] S602, when the quantization parameter is of price type, obtaining a first value range according to the value of the quantization parameter and a first preset ratio, obtaining a first step length according to the value and a second preset ratio, and obtaining a plurality of extended values within the first value range based on the first step length; the second preset ratio is less than the first preset ratio;
[0130] S604, when the quantization parameter is a volume type, obtaining a second value range according to the value of the quantization parameter and a third preset ratio, obtaining a second step length according to the value and a fourth preset ratio, and obtaining a plurality of extended values within the second value range based on the second step length; the fourth preset ratio is smaller than the third preset ratio;
[0131] S606, when the quantization parameter is of the fluctuation type, a third value range is obtained according to the value of the quantization parameter and a preset deviation value, and a plurality of extended values are obtained within the third value range based on a preset step size.
[0132] In some embodiments, the obtaining of the quantitative trading strategy edited by the user, wherein the quantitative trading strategy includes at least one condition, comprises obtaining the quantitative trading strategy edited by the user in response to detecting that the user has completed editing the quantitative trading strategy, wherein the quantitative trading strategy includes at least one condition.
[0133] The backtesting task is executed for each of the extended quantitative trading strategies to obtain the corresponding backtesting result, including executing the backtesting task for each of the extended quantitative trading strategies in response to receiving a backtesting request initiated by the user to obtain the corresponding backtesting result.
[0134] In some embodiments, a backtesting task is performed for each of the extended quantitative trading strategies to obtain corresponding backtesting results, including:
[0135] According to the number of extended quantitative trading strategies, apply to the distributed computing system for at least the number of computing nodes;
[0136] The backtesting task is executed in parallel using the number of computing nodes to obtain the backtesting result of the quantitative trading strategy.
[0137] In some embodiments, see Figure 7 , Figure 7 The figure is a flowchart of the backtesting method for quantitative trading strategies. Figure 7 This is an explanation of S506. Figure 7 As shown, the method includes S702-710.
[0138] S702, obtaining historical real data;
[0139] S704, generating a plurality of data groups based on the historical real data; the data groups include first historical data and second historical transaction data; the time sequence of the second historical transaction is after the first historical data;
[0140] S706, for each data group, obtaining a historical predicted performance of the quantitative trading strategy based on the first historical data, a historical actual performance of the quantitative trading strategy based on the second historical data, and obtaining a historical performance deviation of the quantitative trading strategy based on the historical actual performance and the historical predicted performance;
[0141] S708, obtaining forecast data based on the historical real data, and obtaining future forecast performance of the quantitative trading strategy based on the forecast data;
[0142] S710, correcting the future predicted performance based on the historical performance deviation to obtain a backtest result of the quantitative trading strategy.
[0143] In some embodiments, the historical performance deviation is the difference between the historical actual performance and the historical predicted performance; the future predicted performance is corrected based on the historical performance deviation to obtain the backtest result of the quantitative trading strategy, including: determining the sum of the historical performance deviation and the future predicted performance as the backtest result of the quantitative trading strategy.
[0144] In some embodiments, the historical real data includes market data. The obtaining of forecast data based on the historical real data includes: obtaining forecast data using a first forecast model constructed based on Transformer when the market data indicates that the market fluctuates greatly;
[0145] When the market data indicates that the market trend is obvious, the second prediction model built based on LSTM is used to obtain the prediction data.
[0146] In some embodiments, the historical real data is multimodal data. Before using the prediction model to obtain the prediction data, the method further includes: according to the time sequence of the multimodal data, using the multimodal processing model to process to obtain structured data with time sequence; when the multimodal data is a picture, the multimodal processing model is a SIFT model; when the multimodal model is text, the multimodal processing model is an NLP model; when the multimodal model is audio, the multimodal processing model is an ASR model; the structured data is used to input the prediction model to obtain the prediction data.
[0147] In some embodiments, the output conditions include maximum yield, minimum loss, and lowest risk; the output of N extended quantitative trading strategies that meet the preset output conditions and their corresponding backtest results for the user to make a decision includes: in response to the user configuring the maximum yield, outputting the top N extended quantitative trading strategies with the maximum yield and their corresponding backtest results for the user to make a decision;
[0148] In response to the user configuring the minimum loss, outputting the top N extended quantitative trading strategies with the minimum loss and their corresponding backtesting results for the user to make a decision;
[0149] In response to the user configuring the lowest risk, the top N extended quantitative trading strategies with the lowest risk and their corresponding backtest results are output for the user to make a decision.
[0150] Those skilled in the art will appreciate that one or more embodiments of the present application may be provided as a method, system or computer program product. Therefore, one or more embodiments of the present application may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, one or more embodiments of the present application may take the form of a computer program product implemented on one or more computer-usable storage media (which may include but are not limited to disk storage, CDROM, optical storage, etc.) containing computer-usable program code.
[0151] The term "and / or" in the present application means at least one of the two. For example, "A and / or B" may include three options: A, B, and "A and B".
[0152] Each embodiment in the present application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the data processing device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0153] The above describes a specific embodiment of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0154] The subject matter described in this application and the embodiments of the functional loading can be implemented in the following: digital electronic circuits, tangibly embodied computer software or firmware, computer hardware that can include the structure disclosed in this application and its structural equivalents, or a combination of one or more of them. The embodiments of the subject matter described in this application can be implemented as one or more computer programs, that is, one or more modules in the computer program instructions encoded on a tangible non-temporary program carrier to be executed by a data processing device or control the loading of a data processing device. Alternatively or additionally, the program instructions can be encoded on an artificially generated propagation signal, such as a machine-generated electrical, optical or electromagnetic signal, which is generated to encode information and transmit it to a suitable receiver device for execution by a data processing device. The computer storage medium can be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or a combination of one or more of them.
[0155] The processes and logic flows described in this application can be performed by one or more programmable computers executing one or more computer programs to perform corresponding functions by loading according to input data and generating output. The processes and logic flows can also be performed by special-purpose logic circuits, such as FPGAs (field programmable gate arrays) or ASICs (application-specific integrated circuits), and the apparatus can also be implemented as special-purpose logic circuits.
[0156] A computer suitable for executing a computer program may include, for example, a general and / or special purpose microprocessor, or any other type of central processing unit. Typically, the central processing unit will receive instructions and data from a read-only memory and / or a random access memory. The basic components of a computer may include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, a computer will also include one or more large-capacity storage devices for storing data, such as a disk, a magneto-optical disk, or an optical disk, or the computer will be loadably coupled to this large-capacity storage device to receive data from it or to transmit data to it, or both. However, a computer does not necessarily have such a device. In addition, a computer may be embedded in another device, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device such as a universal serial bus (USB) flash drive, to name a few.
[0157] Computer readable media suitable for storing computer program instructions and data may include all forms of non-volatile memory, media and memory devices, such as semiconductor memory devices (e.g., EPROM, EEPROM and flash memory devices), magnetic disks (e.g., internal hard disks or removable disks), magneto-optical disks, and CD ROM and DVD ROM disks. The processor and memory may be supplemented by, or incorporated in, special purpose logic circuits.
[0158] Although the application includes many specific implementation details, these should not be interpreted as limiting any disclosed scope or the scope of protection claimed, but are mainly used to describe the features of specific disclosed specific embodiments. Certain features described in multiple embodiments in the application can also be implemented in combination in a single embodiment. On the other hand, the various features described in a single embodiment can also be implemented separately in multiple embodiments or implemented with any suitable sub-combination. In addition, although the feature can work in some combinations as described and even initially claim protection, one or more features from the claimed combination can be removed from the combination in some cases, and the claimed combination can point to a variation of a sub-combination or a sub-combination.
[0159] Similarly, although loads are depicted in a particular order in the accompanying drawings, this should not be understood as requiring that these loads be executed in the particular order shown or executed sequentially, or requiring that all illustrated loads be executed to achieve the desired results. In some cases, multitasking and parallel processing may be advantageous. In addition, the separation of various system modules and components in the described embodiments should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product, or packaged into multiple software products.
[0160] Thus, specific embodiments of the subject matter have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the particular order or sequential order shown to achieve the desired results. In some implementations, multitasking and parallel processing may be advantageous.
[0161] The above are only preferred embodiments of one or more embodiments of the present application and are not intended to limit one or more embodiments of the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of one or more embodiments of the present application should be included in the scope of protection of one or more embodiments of the present application.
Claims
1. A quantitative trading backtesting management system, characterized in that: The system comprises: A control unit, used to control the strategy expansion unit, the backtest unit and the output unit to perform their respective tasks; A strategy expansion unit is used to obtain a quantitative trading strategy edited by a user in response to an instruction of the control unit, wherein the quantitative trading strategy includes at least one condition, The quantitative trading strategy is split to obtain the conditions. For each of the conditions, when the quantitative parameter included in the condition is a preset parameter type, the quantitative parameter is numerically expanded according to a preset algorithm to obtain a plurality of extended values, and a plurality of extended conditions are generated based on the plurality of extended values. Each of the extended conditions is combined with other conditions other than the condition to obtain a plurality of extended quantitative trading strategies. A backtesting unit, configured to execute a backtesting task for each of the extended quantitative trading strategies in response to an instruction of the control unit, and obtain a corresponding backtesting result; the backtesting result indicates a trading effect of the extended quantitative trading strategy; The output unit is used to respond to the instruction of the control unit and output N extended quantitative trading strategies and their corresponding backtest results that meet the output conditions preset by the user based on the backtest results for the user to make a decision.
2. The quantitative trading backtesting management system according to claim 1, characterized in that: The preset parameter types include: price type, price increase or decrease type, and transaction volume type; The policy extension unit is further used to: In the case where the quantization parameter is of price type, a first value range is obtained according to the value of the quantization parameter and a first preset ratio, a first step length is obtained according to the value and a second preset ratio, and a plurality of extended values are obtained within the first value range based on the first step length; the second preset ratio is smaller than the first preset ratio; In the case where the quantization parameter is a volume type, a second value range is obtained according to the value of the quantization parameter and a third preset ratio, a second step length is obtained according to the value and a fourth preset ratio, and a plurality of extended values are obtained within the second value range based on the second step length; the fourth preset ratio is smaller than the third preset ratio; In the case where the quantization parameter is of the fluctuation type, a third value range is obtained according to the value of the quantization parameter and a preset deviation value, and a plurality of extended values are obtained within the third value range based on a preset step size.
3. The quantitative trading backtesting management system according to claim 1, characterized in that: The control unit is further configured to initiate an extension request to the strategy extension unit in response to detecting that the user has completed editing the quantitative trading strategy; The strategy extension unit is further configured to extend the quantitative trading strategy in response to receiving the extension request to obtain a plurality of extended quantitative trading strategies; The control unit is further configured to send a backtest request to the backtest unit in response to receiving a backtest request initiated by a user; The backtesting unit is further configured to perform backtesting on the plurality of extended quantitative trading strategies in response to receiving the backtesting request.
4. The quantitative trading backtesting management system according to claim 1, characterized in that: The backtesting unit corresponds to a distributed computing system; the backtesting unit applies for at least the number of computing nodes from the distributed computing system according to the number of extended quantitative trading strategies, and uses the number of computing nodes to execute backtesting tasks in parallel to obtain the backtesting results of the quantitative trading strategy.
5. The quantitative trading backtesting management system according to claim 1, characterized in that: The backtesting unit is further used to obtain historical real data; Generate multiple data groups based on the historical real data; the data groups include first historical data and second historical transaction data; the time sequence of the second historical transaction is after the first historical data; For each data group, obtaining a historical predicted performance of the quantitative trading strategy based on the first historical data, a historical actual performance of the quantitative trading strategy based on the second historical data, and obtaining a historical performance deviation of the quantitative trading strategy based on the historical actual performance and the historical predicted performance; Obtaining forecast data based on the historical real data, and obtaining the future forecast performance of the quantitative trading strategy based on the forecast data; The future predicted performance is corrected based on the historical performance deviation to obtain the backtest result of the quantitative trading strategy.
6. The quantitative trading backtesting management system according to claim 5, characterized in that: The historical performance deviation is the difference between the historical actual performance and the historical predicted performance; The backtesting unit determines the sum of the historical performance deviation and the future predicted performance as the backtesting result of the quantitative trading strategy.
7. The quantitative trading backtesting management system according to claim 5, characterized in that: The historical real data includes market data; the backtesting unit is further used to obtain forecast data using a first forecasting model constructed based on Transformer when the market data indicates that the market fluctuates greatly; When the market data indicates that the market trend is obvious, the second prediction model built based on LSTM is used to obtain the prediction data.
8. The quantitative trading backtesting management system according to claim 7, characterized in that: The historical real data is multimodal data; before using the prediction model to obtain the prediction data, the backtest unit uses the multimodal processing model to process the multimodal data according to the time sequence of the multimodal data to obtain structured data with time sequence; when the multimodal data is a picture, the multimodal processing model is a SIFT model; when the multimodal model is text, the multimodal processing model is an NLP model; when the multimodal model is audio, the multimodal processing model is an ASR model; the structured data is used to input the prediction model to obtain the prediction data.
9. The quantitative trading backtesting management system according to claim 1, characterized in that: The output conditions include maximum yield, minimum loss and lowest risk; In response to the user configuring the maximum rate of return, the output unit is used to output the top N extended quantitative trading strategies with the maximum rate of return and their corresponding backtest results for the user to make a decision; In response to the user configuring the minimum loss, the output unit is used to output the top N extended quantitative trading strategies with the minimum loss and their corresponding backtest results for the user to make a decision; In response to the user configuring the lowest risk, the output unit is used to output the top N extended quantitative trading strategies with the lowest risk and their corresponding backtesting results for the user to make a decision.
10. A quantitative trading backtesting management method, characterized in that: The method comprises: Acquire a quantitative trading strategy edited by a user, wherein the quantitative trading strategy includes at least one condition; The quantitative trading strategy is split to obtain the conditions. For each of the conditions, when the quantitative parameter included in the condition is a preset parameter type, the quantitative parameter is numerically expanded according to a preset algorithm to obtain a plurality of extended values, and a plurality of extended conditions are generated based on the plurality of extended values. Each of the extended conditions is combined with other conditions other than the condition to obtain a plurality of extended quantitative trading strategies. Performing a backtesting task for each of the extended quantitative trading strategies to obtain a corresponding backtesting result; the backtesting result indicates the trading effect of the extended quantitative trading strategy; Output N extended quantitative trading strategies that meet the preset output conditions and their corresponding backtest results for the user to make decisions.
Citation Information
Patent Citations
Full-market multi-variety financial management system based on intelligent dimension element strategy platform
CN110533540A
Bulk commodity price tracking method based on intelligent algorithm
CN115409540A
Long-term price prediction system and method for electricity market
CN116703461A
Asset portfolio recommendation method and device
CN117474685A
System and method of trading securities and program
JP2006189994A
Cited By
Transaction service processing method and device, program product and electronic equipment
CN119671732A
