Order book reconstruction method and order simulation matching method using same

Abstract

The invention provides an order book reconstruction method and an order simulation matchmaking method using the same, and the order book reconstruction method comprises the steps: carrying out the matchmaking of a transaction reconstruction order book from the forward direction according to the transaction rule of the exchange through employing the item-by-item entrustment and item-by-item transaction data, and restoring each transaction in real transactions, thereby constructing a high-precision matchmaking environment. And on the basis of the reconstructed order book, three external simulation entrustment order simulation matching modes, namely an overprice transaction mode, an independent mode and an interaction mode, are provided, and high-precision matching is provided for external simulation entrustment.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to an order book reconstruction method and an order simulation matching method using the same. Background Art

[0002] The types of matching methods available on the market are limited, and their performance and accuracy vary widely, making it difficult to meet the current stringent requirements of high-frequency trading for data integrity, matching accuracy, processing efficiency, and flexibility.

[0003] For order book reconstruction, most systems rely on the 3-second Level 2 snapshots of ten levels of buy and sell volume and price data to serve as the market quotes, or use a method of reconstructing the order book by eliminating each transaction and each order. These methods have delays and information loss when constructing the order book, and cannot reflect the latest market dynamics in real time.

[0004] Since reconstructing the order book and matching orders requires transaction and order data for each transaction, the Shanghai and Shenzhen Stock Exchanges receive billions of stock transaction data every day. The existing system is slow when using this data to reconstruct the order book and match orders. It cannot meet the speed requirements of high-frequency strategy backtesting, nor can it meet the speed requirements of receiving real-time transaction data to reconstruct the order book and match orders in real time.

[0005] Therefore, it is necessary to provide an order book reconstruction method and an order simulation matching method using the same to solve the above problems. Summary of the Invention

[0006] In view of the problems and shortcomings of the prior art, the present invention provides an order book reconstruction method and an order simulation matching method using the same.

[0007] The present invention solves the above technical problems through the following technical solutions:

[0008] An embodiment of the present invention provides an order book reconstruction method, comprising:

[0009] Utilize the order and transaction data of each transaction to reconstruct the order book from forward matching transactions according to the exchange's trading rules, and restore every transaction in the real transaction;

[0010] The exchange trading rules include call auction and continuous auction trading rules, and the call auction includes opening call auction, suspension and resumption call auction and closing call auction.

[0011] Preferably, the call auction period includes:

[0012] All orders are original orders. After obtaining all original orders, they are matched and executed according to the call auction transaction rules of the exchange to obtain the transaction results of each order. During this period, each original order is added to the order book to obtain a corresponding new order book. After the call auction is completed, the final order book is obtained.

[0013] The call auction transaction rules of the exchange include: the price at which the maximum transaction volume is achieved; the price at which all buy orders above the price and sell orders below the price are executed; and the price at which at least one buyer or seller at the same price has all transactions completed.

[0014] Preferably, the call auction period for suspension and resumption of trading includes:

[0015] When the stock price meets the following conditions, a temporary trading halt will be implemented:

[0016] The intraday trading price of a stock with no price limit increases or decreases by 30% or more compared to the opening price of the day for the first time; and

[0017] The intraday trading price of a stock without price fluctuation limits rises or falls by 60% or more compared to the opening price of the day for the first time;

[0018] Among them, the temporary suspension time during a single trading session is 10 minutes.

[0019] Preferably, the continuous bidding period includes:

[0020] During the continuous auction period of the Shanghai Stock Exchange, the transaction data sent first is converted into order data, which is then added to the order book for matching transactions. The remaining new order data is then added to the order book for subsequent transactions.

[0021] During the continuous auction period of the Shenzhen Stock Exchange, the amount corresponding to the market order is removed from the subsequent transaction data. Then, the subsequent transaction data is processed accordingly. During the continuous auction period, the order book corresponding to each transaction and each order is obtained.

[0022] The transaction rules during the continuous bidding period include:

[0023] If the highest bid price is the same as the lowest ask price, the transaction price will be that price.

[0024] If the bid price is higher than the lowest bid price immediately disclosed, the lowest bid price immediately disclosed shall be the transaction price;

[0025] If the selling price is lower than the highest buying price immediately disclosed, the highest buying price immediately disclosed shall be the transaction price.

[0026] An embodiment of the present invention also provides an order simulation matching method using the above-mentioned order book reconstruction method. Based on the reconstructed order book, three external simulated entrustment order simulation matching modes are implemented. The three external simulated entrustment order simulation matching modes include over-price transaction mode, independent mode and interactive mode.

[0027] Preferably, when in over-price execution mode, all external simulated orders for the same stock symbol are matched and executed based on the serial transactions in the same order book. When an external simulated order passes the legitimacy check and comes in, a confirmation response is first returned and added to the external simulated order pool. Then, based on the latest serial transaction data, it is determined whether the external simulated order can be executed immediately.

[0028] If the current external simulated order is a buy order, the order price is greater than or equal to the latest transaction price and the order time is less than the latest transaction time, or if the current external simulated order is a sell order, the order price is less than or equal to the latest transaction price and the order time is less than the latest transaction time, then all orders will be executed and a complete transaction confirmation return will be returned;

[0029] Otherwise, no transaction report is returned, and the system waits for the next transaction to arrive before determining whether the external simulated order in the external simulated order pool can be executed.

[0030] Preferably, when in independent mode, different parent orders of the same stock are independent of each other, each consuming its own order book copied from the same base order book. Different external simulated orders of the same price and direction under the same parent order are independent of each other when consuming the pending orders at the same price level in the same order book. After one external simulated order has consumed the pending orders at a certain level, another external simulated order with the same price and direction can continue to consume the same amount of pending orders at the same price level.

[0031] Preferably, when in interactive mode, different parent orders for the same stock are independent of each other, each consuming its own copy of the order book from the same base order book. Different external simulated orders of the same price and direction on the same parent order affect each other when consuming pending orders at the same price level in the same order book. After one external simulated order consumes all pending orders at a certain level, another external simulated order at the same price and direction cannot consume the same amount of pending orders at that price level and can only wait for new pending orders to appear at that price level.

[0032] Preferably, a unified context object MatchContext is constructed, which is used to maintain and update the buy and sell order books, external entrustment pools, and transaction status in real time.

[0033] Preferably, a lightweight, high-performance MQ middleware is built to achieve efficient decoupling and asynchronous communication between modules.

[0034] Compared with the prior art, the technical solution of the embodiment of the present invention has the following beneficial effects:

[0035] The order simulation matching method provided by the embodiment of the present invention includes: using the order and transaction data of each transaction to reconstruct the order book from the forward matching transaction according to the exchange trading rules, restore each transaction in the real transaction, and thus construct a high-precision matching environment;

[0036] Furthermore, by faithfully restoring every detail of all trading stages, including the opening call auction, suspension and resumption call auction, and closing call auction, it is possible to output snapshots of each trading stage, frequency, and gear.

[0037] Furthermore, by providing three matching modes: over-price transaction mode, independent mode, and interactive mode, a variety of high-precision simulated matching is achieved;

[0038] Furthermore, by building the context MatchContext and a lock-free asynchronous communication mechanism, the data processing process is optimized and the processing speed is improved, so that the system can remain stable and efficient when processing large amounts of data. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A flowchart of an order book reconstruction method provided by one embodiment of the present invention;

[0040] Figure 2 A flowchart of an order book reconstruction method provided by another embodiment of the present invention;

[0041] Figure 3 A flowchart of an order simulation matching method provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0043] The technical solution of the present invention is described in detail below with reference to specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0044] Based on the problems existing in the existing technology, such as Figure 1As shown, the present invention provides an order book reconstruction method, comprising:

[0045] Step S101: Using the order and transaction data of each transaction, the order book is reconstructed from the forward matching transaction according to the exchange trading rules to restore each transaction in the real transaction;

[0046] Step S102: The exchange trading rules include call auction and continuous auction trading rules. The call auction includes opening call auction, suspension and resumption call auction, and closing call auction.

[0047] Specifically, reconstructing the order book is an important part of the matching system, and its correctness determines the quality of the matching system. This application uses transaction-by-transaction data to reconstruct the order book from the forward direction based entirely on the exchange's trading rules, and can restore every detail of the exchange's real transactions.

[0048] During the Shanghai Stock Exchange's continuous auction period, individual order data represents the remaining volume released after the initial matching process. Therefore, order data is reconstructed based on the previously transmitted transaction data. On the Shenzhen Stock Exchange, when an individual order is a market order, since it lacks information about the market order type, the corresponding volume is filled based on the subsequent individual transaction data. This allows the order book to be reconstructed by forward matching transactions using individual order data according to exchange rules. Securities auction transactions are matched based on the principles of price priority and time priority.

[0049] Based on the problems existing in the existing technology, such as Figure 2 As shown, the present invention also provides an order book reconstruction method, comprising:

[0050] Step S201: Using the order and transaction data of each transaction, the order book is reconstructed from the forward matching transaction according to the exchange trading rules to restore each transaction in the real transaction;

[0051] Step S202: The exchange trading rules include call auction and continuous auction trading rules. The call auction includes opening call auction, suspension and resumption call auction, and closing call auction.

[0052] Step S203: The call auction period includes:

[0053] All orders are original orders. After obtaining all original orders, they are matched and executed according to the call auction transaction rules of the exchange to obtain the transaction results of each order. During this period, each original order is added to the order book to obtain a corresponding new order book. After the call auction is completed, the final order book is obtained.

[0054] The call auction transaction rules of the exchange include: the price at which the maximum transaction volume is achieved; the price at which all buy orders above the price and sell orders below the price are executed; and the price at which at least one buyer or seller at the same price has all transactions completed.

[0055] Step S204: The call auction period for suspension and resumption of trading includes:

[0056] When the stock price meets the following conditions, a temporary trading halt will be implemented:

[0057] The intraday trading price of a stock with no price limit increases or decreases by 30% or more compared to the opening price of the day for the first time; and

[0058] The intraday trading price of a stock without price fluctuation limits rises or falls by 60% or more compared to the opening price of the day for the first time;

[0059] Among them, the temporary suspension time during a single trading session is 10 minutes.

[0060] Specifically, this application is based on the stock price strictly following these two conditions to conduct precise intraday temporary suspension and resumption of trading, and then conduct a collective auction for resumption of trading to ensure that every detail of the real transaction is restored.

[0061] Step S205: The continuous bidding period includes:

[0062] During the continuous auction period of the Shanghai Stock Exchange, the transaction data sent first is converted into order data, which is then added to the order book for matching transactions. The remaining new order data is then added to the order book for subsequent transactions.

[0063] During the continuous auction period of the Shenzhen Stock Exchange, the amount corresponding to the market order is removed from the subsequent transaction data. Then, the subsequent transaction data is processed accordingly. During the continuous auction period, the order book corresponding to each transaction and each order is obtained.

[0064] The transaction rules during the continuous bidding period include:

[0065] If the highest bid price is the same as the lowest ask price, the transaction price will be that price.

[0066] If the bid price is higher than the lowest bid price immediately disclosed, the lowest bid price immediately disclosed shall be the transaction price;

[0067] If the selling price is lower than the highest buying price immediately disclosed, the highest buying price immediately disclosed shall be the transaction price.

[0068] Based on the problems existing in the existing technology, such as Figure 3As shown, the present invention also provides an order simulation matching method using the above-mentioned order book reconstruction method, comprising:

[0069] Step S301: Using the order and transaction data of each transaction, the order book is reconstructed from the forward matching transaction according to the exchange trading rules to restore each transaction in the real transaction;

[0070] Step S302: The exchange trading rules include call auction and continuous auction trading rules. The call auction includes opening call auction, suspension and resumption call auction, and closing call auction.

[0071] Step S303: Based on the reconstructed order book, three external simulated entrustment order simulation matching modes are implemented, wherein the three external simulated entrustment order simulation matching modes include an over-price transaction mode, an independent mode, and an interactive mode.

[0072] Specifically, when an external simulated order enters the matching system, it will first be checked by risk control to see if the order information is legal. If it is illegal, a rejection return will be returned. If it is legal, an order confirmation return will be returned and the order will be added to the external simulated order pool. Then, it will be determined whether the order can be executed immediately. If it can be executed, a partial execution return or a full execution return will be returned, and the data corresponding to the external order pool will be deleted. The remaining partially executed orders and orders that cannot be fully executed immediately will wait in the external order pool for subsequent market conditions to update the order book to trigger the execution.

[0073] When multiple external systems call the same matching system, it is desirable that external simulated orders for the same stock sent into the matching system are not affected by external simulated orders sent by other external systems. Therefore, the matching system of the present invention provides an interface for registering a parent order. After registering a parent order for a particular stock, the corresponding external simulated order can be sent to the matching system. External simulated orders between different parent orders for the same stock are independent of each other. Based on whether different external simulated orders with the same price and direction under the same parent order compete for the pending order volume at each level on the order book, this application proposes two matching modes: independent and interactive.

[0074] Step S304: When in over-price execution mode, all external simulated orders for the same stock symbol are matched and executed based on the serial transactions in the same order book. When an external simulated order passes the legitimacy check and is accepted, a confirmation response is first returned and added to the external simulated order pool. The order is then determined to determine whether it can be executed immediately based on the latest serial transaction data.

[0075] If the current external simulated order is a buy order, the order price is greater than or equal to the latest transaction price and the order time is less than the latest transaction time, or if the current external simulated order is a sell order, the order price is less than or equal to the latest transaction price and the order time is less than the latest transaction time, then all orders will be executed and a complete transaction confirmation return will be returned;

[0076] Otherwise, no transaction report is returned, and the system waits for the next transaction to arrive before determining whether the external simulated order in the external simulated order pool can be executed.

[0077] Specifically, with over-the-price execution, assuming a simulated external order has a small order size, if a real order at that price has already been filled during real trading, it can be assumed that the simulated order has also been fully filled at that price. Over-the-price execution does not require registering a parent order; all simulated external orders for the same stock symbol are matched and executed based on the individual transactions in the same order book.

[0078] Step S305: When in independent mode, different parent orders of the same stock are independent of each other, and each consumes its own order book copied from the same basic order book. Different external simulated orders of the same price and direction of the same parent order are independent of each other when consuming the pending orders at the same price level of the same order book. After an external simulated order consumes the pending orders at a certain level, another external simulated order with the same price and direction can continue to consume the same amount of pending orders at the same price level.

[0079] Step S306: When in interactive mode, different parent orders for the same stock are independent of each other and each consumes its own order book copied from the same basic order book. Different external simulated orders of the same price and direction of the same parent order affect each other when consuming the pending orders at the same price level of the same order book. After an external simulated order consumes the pending orders at a certain level, another external simulated order with the same price and direction cannot consume the same amount of pending orders at the same price level and can only wait for new pending orders to appear at the same price level.

[0080] Specifically, the over-the-price execution mode determines whether an external simulated order can be executed based on the latest transaction price. It does not consider the price or volume of the counterparty's order book. It is a simple and intuitive matching method. The independent mode and interactive mode take into account both the pending order price and the pending order volume of the order book level. The execution is more stringent than the over-the-price execution mode. The execution strictness of the three modes is as follows: over-the-price execution mode, independent mode and interactive mode.

[0081] In specific implementation, this application provides two methods of accessing market data. One is to automatically read local market data, including reading local file market data and receiving real-time market data, reconstructing the order book, and matching external simulated orders sent by external systems; the other is for the external system to send market data to the matching system to reconstruct the order book, and at the same time send external simulated orders for matching transactions. In this way, the external system can accurately control the order book status before the external order arrives, accurately analyze the transaction logic of the external simulated order, and the sending speed of market data is controlled by the external system.

[0082] In addition, to support the needs of simulated real-time strategy backtesting, a real-time replay control mechanism is provided. The read market data time is compared with the local system time in real time, achieving real-time mapping of the 09:15–15:00 trading period. A second round of simulated real-time replay is also supported from 15:15–21:00, allowing users to conduct complete strategy backtesting and simulation verification during non-trading hours.

[0083] In the specific implementation, a unified context object MatchContext is constructed to implement a memory-centric data carrying method. The context object is used to maintain and update the buy and sell order books, external entrustment pools, and transaction status in real time.

[0084] Specifically, the full-memory architecture effectively reduces system latency and improves execution efficiency through the following mechanisms:

[0085] (1) Avoid frequent I / O operations and data loading

[0086] All core data, including buy and sell order books, external entrustment pools, transaction status, etc., are maintained and updated in real time in memory, avoiding the overhead of accessing disks, databases or external storage, significantly reducing data reading latency, and improving the overall response speed of the system.

[0087] (2) Build a unified memory context structure to achieve in-place data processing

[0088] By building a globally unified in-memory data structure called MatchContext, core data related to matching, including buy and sell order books, external entrustment pools, and transaction status, is integrated into the same memory space. Each module directly operates on the context memory object to update data and transfer status, enabling in-place data processing. This avoids complex data transmission and object construction between modules and improves data access efficiency.

[0089] In the specific implementation, a lightweight and high-performance MQ (Message Queue) middleware is built to achieve efficient decoupling and asynchronous communication between modules. Its main functions include:

[0090] When an external system sends a market order or external simulated order request through the matching client (API), it first enters the message queue;

[0091] The message queue pushes messages to the matching engine in a strict order, and the engine reconstructs the order book and performs matching.

[0092] For external simulated entrustment, the system generates an entrustment confirmation message immediately after receiving the request and pushes it to the client asynchronously via MQ;

[0093] When the external simulated order is completed, the system will generate a transaction confirmation message, which will also be pushed back through MQ to ensure that the order and transaction feedback are asynchronous and highly real-time.

[0094] The MQ module has the characteristics of high throughput, low latency and sequential guarantee, meeting the dual requirements of the trading system for reliability and performance.

[0095] The system implements a concurrent scheduling mechanism based on Goroutines, employing a "one-per-stock goroutine" approach for order book reconstruction and simulated matching. This design fully leverages multi-core CPU resources, significantly improving the system's parallel processing capabilities and making it particularly suitable for high-frequency, multi-stock concurrent matching scenarios. Compared to traditional threading models, this approach offers lower context switching overhead and higher resource utilization, significantly improving overall system throughput and responsiveness.

[0096] In summary, the order simulation matching method provided by the embodiment of the present invention includes: utilizing order and transaction data of each transaction to reconstruct the order book from forward matching transactions according to the exchange trading rules, and restore each transaction in the real transaction; thereby constructing a high-precision matching environment;

[0097] Furthermore, by faithfully restoring every detail of all trading stages, including the opening call auction, suspension and resumption call auction, and closing call auction, it is possible to output snapshots of each trading stage, frequency, and gear.

[0098] Furthermore, by providing three matching modes: over-price transaction mode, independent mode, and interactive mode, a variety of high-precision simulated matching is achieved;

[0099] Furthermore, by building the context MatchContext and a lock-free asynchronous communication mechanism, the data processing process is optimized and the processing speed is improved, so that the system can remain stable and efficient when processing large amounts of data.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An order book reconstruction method, characterized in that: The method comprises: Utilize the order and transaction data of each transaction to reconstruct the order book from forward matching transactions according to the exchange's trading rules, and restore every transaction in the real transaction; The exchange trading rules include call auction and continuous auction trading rules, and the call auction includes opening call auction, suspension and resumption call auction and closing call auction.

2. The order book reconstruction method according to claim 1, characterized in that: The call auction period includes: All orders are original orders. After obtaining all original orders, they are matched and executed according to the call auction transaction rules of the exchange to obtain the transaction results of each order. During this period, each original order is added to the order book to obtain a corresponding new order book. After the call auction is completed, the final order book is obtained. The call auction transaction rules of the exchange include: the price at which the maximum transaction volume is achieved; the price at which all buy orders above the price and sell orders below the price are executed; and the price at which at least one buyer or seller at the same price has all transactions completed.

3. The order book reconstruction method according to claim 1, characterized in that: The call auction period for suspension and resumption of trading includes: When the stock price meets the following conditions, a temporary trading halt will be implemented: The intraday trading price of a stock with no price limit increases or decreases by 30% or more compared to the opening price of the day for the first time; and The intraday trading price of a stock without price fluctuation limits rises or falls by 60% or more compared to the opening price of the day for the first time; Among them, the temporary suspension time during a single trading session is 10 minutes.

4. The order book reconstruction method according to claim 1, characterized in that: The continuous bidding period includes: During the continuous auction period of the Shanghai Stock Exchange, the transaction data sent first is converted into order data, which is then added to the order book for matching transactions. The remaining new order data is then added to the order book for subsequent transactions. During the continuous auction period of the Shenzhen Stock Exchange, the amount corresponding to the market order is removed from the subsequent transaction data. Then, the subsequent transaction data is processed accordingly. During the continuous auction period, the order book corresponding to each transaction and each order is obtained. The transaction rules during the continuous bidding period include: If the highest bid price is the same as the lowest ask price, the transaction price will be that price. If the bid price is higher than the lowest bid price immediately disclosed, the lowest bid price immediately disclosed shall be the transaction price; If the selling price is lower than the highest buying price immediately disclosed, the highest buying price immediately disclosed shall be the transaction price.

5. An order simulation matching method using the order book reconstruction method according to any one of claims 1 to 4, characterized in that: Based on the reconstructed order book, three external simulated entrustment order simulation matching modes are implemented, and the three external simulated entrustment order simulation matching modes include over-price transaction mode, independent mode and interactive mode.

6. The order simulation matching method according to claim 5, characterized in that: When in over-price execution mode, all external simulated orders for the same stock symbol are matched and executed based on the serial transactions in the same order book. When an external simulated order passes the legitimacy check and comes in, it first returns an order confirmation response and is added to the external simulated order pool. Then, based on the latest serial transaction data, it is determined whether the external simulated order can be executed immediately. If the current external simulated order is a buy order, the order price is greater than or equal to the latest transaction price and the order time is less than the latest transaction time, or if the current external simulated order is a sell order, the order price is less than or equal to the latest transaction price and the order time is less than the latest transaction time, then all orders will be executed and a complete transaction confirmation return will be returned; Otherwise, no transaction report is returned, and the system waits for the next transaction to arrive before determining whether the external simulated order in the external simulated order pool can be executed.

7. The order simulation matching method according to claim 5, characterized in that: When in independent mode, different parent orders of the same stock are independent of each other, each consuming its own copy of the order book from the same base order book. Different external simulated orders of the same price and direction on the same parent order consume the same amount of pending orders at the same price level in the same order book independently. After one external simulated order consumes all the pending orders at a certain level, another external simulated order at the same price and direction can continue to consume the same amount of pending orders at the same price level.

8. The order simulation matching method according to claim 5, characterized in that: When in interactive mode, different parent orders for the same stock are independent of each other and each consumes its own copy of the order book from the same base order book. Different external simulated orders of the same price and direction on the same parent order affect each other when consuming the pending orders at the same price level in the same order book. After an external simulated order consumes the pending orders at a certain level, another external simulated order with the same price and direction cannot consume the same amount of pending orders at that price level and can only wait for new pending orders to appear at that price level.

9. The order simulation matching method according to claim 5, characterized in that: Construct a unified context object MatchContext, which is used to maintain and update the buy and sell order books, external entrustment pools, and transaction status in real time.

10. The order simulation matching method according to claim 5, characterized in that: Build a lightweight, high-performance MQ middleware to achieve efficient decoupling and asynchronous communication between modules.

Images