CN112330446A

CN112330446A - Futures quantitative transaction platform

Info

Publication number: CN112330446A
Application number: CN202010299362.0A
Authority: CN
Inventors: 吴凡
Original assignee: Shanghai Zexun Asset Management Co ltd
Current assignee: Shanghai Zexun Asset Management Co ltd
Priority date: 2020-04-16
Filing date: 2020-04-16
Publication date: 2021-02-05

Abstract

The invention provides a futures quantitative trading platform which comprises a variety attribute module, a strategy source code module, a real-disk trading module, a historical data return testing module and a graph display module, wherein the strategy source code module is used for storing a variety of a futures; the strategy source code module comprises a K line synthesis strategy and a multi-space signal algorithm strategy; the real-disk transaction module comprises a real-time data processing submodule, a bin position calculating submodule and an order processing submodule, wherein the real-time data processing submodule is used for initializing a Tick memory pool after reading local historical data and receiving real-time Tick data through a transaction stream; the historical data back testing module reads the local data and the stored Tick data, performs strategy historical data back testing and gives a test report; the graphic display module is used for displaying the information content of the real disk transaction module and the historical data testing module. The invention is used for solving the technical problems that the existing quantitative transaction software is low in running speed, is difficult to deal with some complex mathematical models, limits a plurality of functions and brings a plurality of limitations.

Description

Futures quantitative transaction platform

Technical Field

The invention relates to the crossing field of finance and computer software, in particular to a futures quantitative trading platform.

Background

In the 20 th century and the 80 th century, with the popularization of personal computers and the subsequent rise of the internet, many second-class market investors utilize computer programs to perform automatic investment on securities, futures and other financial derivatives, and compared with the traditional investment mode, the method has the main characteristic that a computer automatically determines selling points to perform level selling. However, many financial investors lack the level of computer programming, and thus, many companies specialized in the development and sale of trading platform software are gradually emerging, and many popular quantitative trading software are developed correspondingly.

Among them, the more widely known are TradeStation, Amibroker, MetaTrader, TradeBlazer, Wen Hua finance, pyramid, etc. The common characteristic of these software is that after purchasing, a customer can use the scripting language provided by the customer or the common simple programming language (such as C #, Python) on the market to write out the transaction idea quickly, and then the software has the capability of historical data backtesting and real disk transaction, and although the scripting language and the C # language are written conveniently, the running speed is generally much slower than that of the C + + language (the difference can reach an order of magnitude). And due to the grammatical limitation of the script language, the user often catches the elbow when writing a complex mathematical model. Meanwhile, a general quantitative trading software adopts a one-to-one corresponding rule that a strategy corresponds to a hand position, and the setting is very unfavorable for investors with small fund amount.

Disclosure of Invention

In view of the above-mentioned shortcomings of the prior art, the present invention aims to provide a futures quantitative trading platform, which is used for solving the technical problems that the existing quantitative trading software is slow in operation speed, difficult to deal with some complex mathematical models, limited in many functions and limited in many limitations.

The invention provides a futures quantitative trading platform which comprises a variety attribute module, a strategy source code module, a real-disk trading module, a historical data return testing module and a graph display module, wherein the strategy source code module is used for storing a variety of a futures;

the variety attribute module is used for storing the attribute value and the set value of each futures variety,

the strategy source code module comprises a K line synthesis strategy and a multi-space signal algorithm strategy, wherein the K line synthesis strategy is used for synthesizing Tick data into K lines of each minute level;

the real disk transaction module comprises a real-time data processing sub-module, a bin position calculating sub-module and an order processing sub-module,

the real-time data processing submodule initializes the Tick memory pool after reading the local historical data and receives the real-time Tick data through the transaction stream;

the bin position calculating submodule is used for calculating the integral holding quantity of the required leveling bin;

the order processing sub-module performs a warehouse opening action according to the integral warehouse holding amount, and processes transaction report information and orders needing to withdraw or deliver the orders;

the historical data back testing module reads the local data and the stored Tick data, performs strategy historical data back testing and gives a test report;

the graphic display module is used for displaying the information content of the real disk transaction module and the historical data testing module.

In an embodiment of the present invention, the attribute values and setting values of the futures varieties include contract names, names of exchanges where contracts are located, contract point differences, contract multipliers, whether contracts allow for parallel and current warehouses, maximum daily opening hands of contracts, and the like, and data in the variety attribute module is called by the other four modules.

In an embodiment of the present invention, the bin calculation submodule waits for the multi-empty signal of each algorithm strategy for the first n seconds per minute, and an integer amount of taken positions required to open the bin can be obtained by using a bin stacking function.

In an embodiment of the present invention, the historical data retrieval module includes a local data reading processing sub-module, a simulation bin calculating sub-module and a simulation order processing sub-module,

the local data reading processing submodule is used for verifying the integrity of the local data and reading the local data into the memory;

the simulated bin position calculation submodule is used for calculating a multi-space signal of each algorithm strategy, and can obtain the integral quantity of taken bins of the account due to the simulated total bin position and the number of opening hands needing to be increased and decreased by adopting a bin position superposition function;

and the simulated order processing submodule calculates the quantity obtained by the simulated bin position calculating submodule to obtain a return report after calculation.

In an embodiment of the present invention, the principle of the bin stacking function is: and (3) superposing the multi-space signals of all the algorithm strategies at the same moment into an overall signal through weight, multiplying the overall signal by the maximum allowable position opening amount to obtain the due floating point position holding number of the account, comparing the floating point position holding number with the existing certificate position holding of the account, and rounding the floating point position holding number if the position difference is more than 0.75 to obtain the new due integer position holding amount.

In an embodiment of the present invention, the process of the real disk mode corresponding to the real disk transaction module is as follows:

step 1: reading a configuration file of the real disk transaction, wherein the configuration file comprises account information: account name, password, futures trader code, corresponding sentiment flow and transaction flow address of the futures trader;

step 2: logging in a market flow to obtain the real-time market of the current day;

and step 3: logging in a transaction flow, and calculating bin position details of an account, wherein the bin position details comprise the current empty bin position of the account, the current number of opening hands, the number of closing hands and the number of withdrawing single strokes of the account, and the current amount of hanging single of the account;

and 4, step 4: the real-time data processing submodule reads local historical data, reads the local historical data, and combines the local historical data and the current-day real-time quotation into a continuous quotation stream to be stored in the memory;

and 5: removing all the hang lists, and checking whether the initialization of the Tick memory pool is completed;

step 6: loading and running a thread corresponding to each K line synthesis strategy and the multi-empty signal strategy, and determining whether the thread needs to be started or suspended through a configuration file;

and 7: entering a program main loop, wherein the main loop is cycled at intervals, each cycle is required to check whether the configuration file and the account information are changed, and if the configuration file and the account information are changed, the contents required to be updated and displayed are fed back to the graphic display module;

and 8: the main cycle starts a stacking mode or a non-stacking mode of the bin according to the configuration file, and an integer holding capacity of the account which should simulate the total bin is obtained by the bin calculation submodule;

and step 9: and the order processing submodule performs a warehouse opening action according to the integral warehouse holding quantity, and processes transaction report information and orders needing to withdraw or deliver the orders.

As described above, the present invention has the following advantageous effects:

1. the software of the transaction platform is completely compiled by using C + + as a language, so that the running speed of the software is increased and the running stability of the software is ensured while the user can freely compile complex strategies.

2. Compared with other quantitative transaction software, the invention adopts the original strategy bin position overlapping technology, and the strategies can be overlapped and used in real time no matter how many strategy numbers are more than the maximum allowable bin positions, so that a lot of risks can be eliminated to a great extent, the frequent rate of opening the leveling bin in the oscillation market is effectively reduced, and the good effect which is not possessed by other software is achieved.

Drawings

Fig. 1 is a schematic diagram of a module structure disclosed in the present invention.

FIG. 2 is a flow chart of a main routine disclosed in the present invention.

Fig. 3 is a flow chart of the real disk mode disclosed in the present invention.

Fig. 4 shows a flow chart of the overlay mode disclosed in the present invention.

FIG. 5 is a flow chart illustrating a backtesting mode according to the present disclosure.

Detailed Description

The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.

It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.

Referring to fig. 1, the present invention provides a system including a variety attribute module, a policy source code module, a real disk transaction module, a historical data review module, and a graph display module;

the strategy source code module comprises a K line synthesis strategy and a multi-space signal algorithm strategy, wherein the K line synthesis strategy is used for synthesizing Tick data into K lines of each minute level, and the multi-space signal algorithm strategy is specifically realized through a programming language and a mathematical model;

the graphic display module is used for displaying the information content of the real disk transaction module and the historical data retrieval module, the graphic display module adopts QT to realize GUI (graphical interface), and the real disk module and the historical data retrieval module can call the module.

Based on the above embodiment, the attribute values and setting values of the futures varieties comprise contract names, names of exchanges where contracts are located, contract point differences, contract multipliers, whether contracts allow for parallel and current warehouses, maximum daily opening numbers of contracts and the like, and data in the variety attribute module is called by other four modules.

Based on the above embodiment, the bin calculating submodule waits for the first n seconds per minute (n is set to 13) of the multi-empty signal of each algorithm strategy, and can use a bin stacking function to obtain the integral taken position quantity of the required open-level bin.

Based on the above embodiment, the historical data back-testing module comprises a local data reading processing sub-module, a simulation bin position calculating sub-module and a simulation order processing sub-module,

the simulation order processing submodule obtains a return report after calculation by using the quantity obtained by the simulation position calculation submodule, and the return report comprises: open flat records, daily monthly account equity statistics, etc.

Based on the above embodiment, the principle of the bin stacking function is as follows: superposing all the multi-space signals of the algorithm strategy at the same moment into an overall signal through weight, multiplying the overall signal by the maximum allowable position opening amount to obtain the due floating point position holding number of the account, comparing the floating point position holding number with the existing certificate position holding of the account, rounding the floating point position holding number if the position difference is more than 0.75 to obtain the new due integer position holding amount;

the disadvantage that the total bin frequently changes due to the change of few algorithm strategies can be overcome by adopting the bin stacking function, and a plurality of existing algorithm strategies are combined and stacked to be matched with various funds, including the operation of small funds, so that the user loss of the small funds is avoided.

Referring to fig. 2, the implementation method of the present invention is: firstly, whether the same program at the same position is started in the memory or not is confirmed, and if the situation exists, the opening of the bin is mistaken; then starting a program, and selecting a real-time mode and a retest mode;

referring to fig. 3, the flow of the real-disk mode is as follows, corresponding to the real-disk transaction module, and implemented by the real-time data processing sub-module, the bin calculating sub-module, and the order processing sub-module:

and 7: entering a main loop of the program, wherein the main loop makes a loop at intervals (a common time interval is set to be no more than 500 milliseconds, namely a time interval between two adjacent ticks), and each loop needs to check whether the configuration file and the account information have changes, for example: the user can pause or start part of the strategies, modify the weight of the number of hands, pause account transaction after the number of wrong single strokes of the account reaches a certain number, and the like in the disk, and if the number of wrong single strokes of the account changes, the content needing to be updated and displayed is fed back to the graphic display module;

Specifically, the non-overlap mode of the bin in step 8 refers to: each strategy is completely independent, and one strategy corresponds to one hand of position or fund.

Referring to fig. 4, specifically, the implementation steps of the overlay mode in step 8 are as follows:

1. strategy initialization: calculating transaction signals of each strategy from the first day of historical data (generally, the current date is advanced by 10 to 20 transaction days), and further synthesizing a historical total position holding table, wherein the position superposition function is required to be called in the synthesizing process;

2. checking whether the monthly balance is needed, if yes, sequentially balancing the old contracts and then opening new contracts with the same bin positions;

3. entering a polling part per minute:

3.1, time check is carried out, if the current time is not in the transaction time or the reading time of the multi-empty signal of the algorithm strategy (currently, the current setting is that the first 13 seconds per minute is the reading time), the later part is skipped to wait for entering the next polling;

3.2, checking the bill delivery state, checking whether a bill is delivered to the exchange but the return information of the exchange is not received, and waiting for next polling if the bill is delivered to the exchange;

3.3, when all the multi-space signals are updated in the minute or the reading time limit of the multi-space signals is reached, delivering all the multi-space signals to a bin superposition function;

and 3.4, generating an order according to the position difference, pushing the order into an order queue, and then enabling the delivery line to poll and process (the delivery line is established when an account is initialized, and the delivery line processes delivery, delays delivery, withdrawals and then delivers the delivery line and other delivery line related operations through polling).

Referring to fig. 5, the retest mode is implemented by a local data reading and processing sub-module, a simulated bin position calculating sub-module and a simulated order processing sub-module, corresponding to the real disk transaction module:

step 1: reading a test configuration file to obtain information such as test starting time and test ending time setting;

step 2: checking the date of the local Tick data text, checking whether the data text is complete and continuous, and stopping the retest operation if the data text is not qualified;

and step 3: reading each Tick from a local Tick data text file in a for loop respectively, calling a K line synthesis strategy to synthesize the latest K line, and calling each algorithm strategy to calculate a multi-space signal;

and 4, step 4: the back test mode can also adopt a non-superimposed mode or a superimposed calculation bin position mode;

and 5: and after the daily data are calculated, judging whether the month needs to be changed or whether the last day needs to be retested, and if so, leveling all the bins. And when the data in all the retest days are completely retested, counting the account total rights and interests and all transaction sheet records and writing the records into a retest report text, and meanwhile, counting the transaction records and the strategy rights and interests of each algorithm strategy and writing the strategy retest report text.

In conclusion, the invention can improve the platform efficiency, and solves the problem that a large number of strategies are operated on small funds simultaneously by adopting the strategy bin stacking technology. Therefore, the invention effectively overcomes various defects in the prior art and has high industrial utilization value.

The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.

Claims

1. A futures quantitative trading platform is characterized by comprising a variety attribute module, a strategy source code module, a real-disk trading module, a historical data return testing module and a graph display module;

2. The futures quantitative trading platform of claim 1, wherein: the attribute values and the set values of the futures varieties comprise account names, passwords, futures trader codes, corresponding market flow and trading flow addresses of the futures traders, contract names, names of exchanges where contracts are located, contract point differences, contract multipliers, whether the contracts allow for flat and straight silos or not and maximum daily open hand numbers of the contracts.

3. The futures quantitative trading platform of claim 1, wherein: and waiting for the multi-space signal of each algorithm strategy in the first n seconds of each minute by the bin position calculation submodule, and obtaining the integral position holding quantity of the required open-level bin by adopting a bin position superposition function.

4. The futures quantitative trading platform of claim 1, wherein: the historical data remeasurement module comprises a local data reading processing submodule, a simulation bin position calculation submodule and a simulation order processing submodule,

5. The futures quantitative trading platform according to claim 3 or 4, wherein: the principle of the bin stacking function is as follows: and (3) superposing the multi-space signals of all the algorithm strategies at the same moment through weights to form an overall signal, multiplying the overall signal by the maximum allowable open-position amount to obtain the due floating point position number of the account, comparing the floating point position number with the existing integer position of the account, and rounding the floating point position number if the position difference is more than 0.75 to obtain the new due integer position number of the account.

6. The futures quantitative trading platform of claim 1, wherein: the flow of the real disk mode corresponding to the real disk transaction module is as follows: