Field-programmable gate array for real-time calculation of margin height

Abstract

Field-programmable gate array for real-time calculation of a margin amount, comprising: a transmitting and receiving unit (11) configured to receive at least one data packet (20), wherein the data packet (20) comprises a plurality of data fields (21), and wherein the data fields (21) may contain market data (211) and / or trading order data (212); and a computing unit (12) connected to the transmitting and receiving unit (11), wherein the computing unit (12) stores user position data (121) and a margin calculation program (122), wherein the computing unit (12) extracts bit data directly from at least one of the data fields (21) on the basis of the data packet (20) received via the transmitting and receiving unit (11) and performs a calculation; wherein the computing unit (12) executes the margin calculation program (122) using the user position data (121) or the trading order data (212) in combination with the market data (211) to calculate a margin amount (123), and wherein an unrealized profit determined on the basis of the market data (211) and the user position data (121) is taken into account in the calculation of the margin amount (123).

Description

[0001] The present invention relates to a field programmable gate array (FPGA) and, more particularly, to a field programmable gate array for real-time calculation of a margin height.

[0002] In current futures trading, exchanges require a minimum margin. If a trading account balance falls below this prescribed minimum, the futures brokerage firm triggers a margin call. Since the prices of financial instruments change continuously and sometimes abruptly, the required margin is also subject to ongoing fluctuations. Currently, the margin calculation is software-based. Incoming data packets must be interpreted step by step, which leads to delays. If the system cannot capture and process the trading data transmitted by the investor in real time, there is a risk that the investor will miss optimal trading opportunities.

[0003] The invention is based on the object of creating a field-programmable gate array for real-time calculation of a margin amount, which avoids the problem of the lack of real-time capability in margin calculation in conventional systems.

[0004] This object is achieved according to the invention by a field-programmable gate array for real-time calculation of a margin level, which has the features specified in claim 1. Further advantageous developments of the invention emerge from the features of the subclaims.

[0005] According to the invention, a field-programmable gate array (FPGA) for real-time calculation of a margin amount is provided, comprising a transmitting and receiving unit and a computing unit connected thereto. The transmitting and receiving unit is configured to receive at least one data packet comprising a plurality of data fields. These data fields may contain market data and / or trading order data.

[0006] The processing unit stores user position data and a margin calculation program. Based on the data packet transmitted by the transmitting and receiving unit, the processing unit directly extracts bit data from at least one of the data fields and then performs a calculation. The margin calculation program is executed based on the user position data or the trading order data in combination with the market data to determine the margin amount. The calculation also includes an unrealized profit derived from the market data and the user position data.

[0007] In one embodiment, the user position data consists of a combination of multiple futures products of the same investor, with each data packet being associated with one of these products.

[0008] In another embodiment, the margin calculation program defines multiple calculation weighting factors, each of which is assigned to one of the futures products. When executing the program, the calculation unit takes these calculation weighting factors into account when calculating the margin amount.

[0009] In a further embodiment, the computing unit executes the margin calculation program in a multi-thread architecture, with several computing subroutines each being assigned to a specific futures product.

[0010] In a further embodiment, the received data packet serves as a data source for updating the margin calculation program.

[0011] In another embodiment, if the data fields contain market data, the computing unit executes a comparison program that compares the market data with the user position data. If there is a match, the margin calculation program continues immediately; otherwise, the market data is saved.

[0012] Compared to the state of the art, this innovation offers the particular advantage that the processing unit can extract information directly from the data fields of a received packet and use it to calculate the margin amount in real time. By taking unrealized profits into account, a realistic representation of the funds actually available in the investor's trading account is obtained. The hardware-based implementation significantly improves processing speed compared to software-based solutions, which require incoming packets to be interpreted step by step, which previously led to delays and missed optimal trading times.

[0013] The invention and its embodiments are explained in more detail below with reference to the drawing. The drawing shows: Fig. 1 is a schematic block diagram of the unit structure of the field-programmable gate array for real-time calculation of a margin level according to the invention; Fig. 2 a schematic block diagram of the composition of the calculated margin amount; Fig. 3 a schematic representation of the relationship between futures products and received data packets; Fig. 4 shows a first embodiment of the present invention in schematic form; and Fig. 5 a first embodiment of the present invention in schematic form.

[0014] The following detailed description of the present invention and its technical contents is made with reference to the drawings.

[0015] As in the Fig. 1 to Fig. As shown in Figure 5, the present invention relates to an FPGA 10 (Field Programmable Gate Array) used for margin calculation according to the SPAN system. This FPGA 10 can be integrated into a server device to enable hardware acceleration. This allows information processing to be performed directly on the data link layer, enabling real-time calculation based on current market movements and investor trading orders.

[0016] The FPGA 10 comprises a transmitting and receiving unit 11 and a computing unit 12. The transmitting and receiving unit 11 is designed to receive at least one data packet 20. This data packet 20 can originate either from a user terminal 50 of one or more investors or from a server 60 of the centralized marketplace of a futures exchange. The data packet 20 has several data fields 21 containing information about the source, destination, and payload of the packet. The data fields 21 can include market data 211 and / or trading order data 212, with this information being stored in the payload area of ​​the data fields 21. The market data 211 contains, among other things, price information for at least one futures product 30, while the trading order data 212 contains the order data issued by the investor—e.g., purchase / sale, quantity, price, order type.

[0017] The computing unit 12 is connected to the transmitting and receiving unit 11 and stores both user position data 121 and a margin calculation program 122. The user position data 121 contains information about the positions held by the investor and their size, while the margin calculation program 122 is pre-stored in the computing unit 12 in the form of program code. When executing the margin calculation program 122, the computing unit 12 can calculate a required margin amount 123 based on the received data packet 20. The calculation is not performed by step-by-step interpretation of the individual data fields 21, but by directly reading bit data from at least one of the data fields 21. The margin amount 123 is calculated using the user position data 121 or the trading order data 212 in combination with the market data 211.The determined margin amount 123 is composed of at least one of the following components: an initial margin 124, a maintenance margin 125, i.e., the minimum collateral required to maintain the position, and an unrealized profit 126. The initial margin 124 is determined based on the trading order data 212 and the market data 211 and indicates the margin amount required when entering a position. The maintenance margin 125 is calculated based on the user position data 121 and the market data 211 and indicates the minimum collateral amount required to maintain an existing position. The unrealized profit 126 is also derived from the market data 211 and the user position data 121 and represents the book profit achieved on the held position based on the current market situation.When calculating the margin amount 123, the computing unit 12 does not rely exclusively on the current market price of the respective futures product 30. Rather, the calculation can also be based on the user position data 121 and the market data 211, taking into account combinations and sequences of product-related settlements resulting from the number and type of positions held. It should be noted that the unrealized profit 126 represents the profit from open positions at the current time. In contrast to unrealized profits or losses on other financial instruments, such as shares, which can be negative, the unrealized profit 126 within the meaning of the present invention is always greater than or equal to zero. Losses do not lead to a negative value of the unrealized profit 126, but instead cause an increase in the maintenance margin 125.

[0018] As can be seen from the above, this innovation, thanks to the processing architecture described above, enables a rapid response to market movements and eliminates the previous problem of delayed data processing, which was particularly disadvantageous with regard to high-frequency trading and the rapid calculation of margins. Furthermore, the unrealized profit 126 is taken into account when determining the margin amount 123, thereby achieving a more realistic representation of the funds actually available in the investor's trading account. This eliminates the previous problem of forced liquidation of positions due to insufficient and untimely margin payments.

[0019] Referring to the Fig. 1 to Fig. 5, one embodiment provides that the margin calculation program 122 defines a calculation weighting factor that is used by the computing unit 12 to carry out the previously described combinations of product-related settlements. This calculation weighting factor can be stored both in the market data 211 of the data packet 20 and in advance in the computing unit 12. When executing the margin calculation program 122, the computing unit 12 can additionally take the calculation weighting factor into account. If either the market price of the futures product 30 in question described in the market data 211 or the user position data 121 changes, the computing unit 12 executes the margin calculation program 122 again incorporating the calculation weighting factor. In a further embodiment, a transaction fee 127 is also taken into account when calculating the margin amount 123.

[0020] One and the same investor can select different futures products 30 as investment instruments. The trading order data 212 of the present invention is therefore composed of multiple futures products 30 of the same investor, with each received data packet 20 being assigned to a single futures product 30. In other words, the trading order data 212 each relate to a single futures product 30. If an investor places orders for different investment instruments, the transmitting and receiving unit 11 receives multiple corresponding data packets 20. Accordingly, each set of market data 211 contains exclusively information on the market situation of a single futures product 30. If the futures exchange intends to provide the FPGA 10 with market data on several different futures products 30, the transmitting and receiving unit 11 also receives multiple data packets 20.

[0021] The margin calculation program 122 defines a plurality of calculation weighting factors, each of which is assigned to a respective futures product 30. The calculation weighting factors are based on a risk-related classification of the various futures products 30. When executing the margin calculation program 122, the computing unit 12 includes these multiple calculation weighting factors in the calculation.

[0022] In another embodiment, see Fig. 1 and Fig. 3, the computing unit 12 executes the margin calculation program 122 in a multi-threaded architecture, with several computing subroutines 128 being processed in parallel. Each of these computing subroutines 128 is intended for processing a specific futures product 30 from a group of several products.

[0023] On the other hand, in the present invention, the data packet 20 also serves as a data source for updating the margin calculation program 122. For example, based on the information contained in the market data 211, the computing unit 12 can detect a price change of the futures product 30 in question and subsequently adjust the corresponding calculation weighting factor within the margin calculation program 122. Furthermore, the data packet 20 can also function as an update source for the user position data 121. For example, after a transaction has been successfully completed, the computing unit 12 stores the contents of the trading order data 212 in order to update the calculation factors for the margin amount 123 on this basis.

[0024] In one embodiment, the computing unit 12 stores a comparison program 129. If the received data packet 20 contains market data 211 in its data fields 21, the computing unit 12 compares the market data 211 with the user position data 121 using the comparison program 129. If the comparison result is positive - i.e., the market data 211 matches an existing position - the computing unit 12 immediately continues executing the margin calculation program 122. Otherwise, the market data 211 is saved and serves as an updated data basis for the margin calculation program 122. For example, the computing unit 12 can archive the market data 211 as the price history of a specific futures product 30 and use this historical market data as a reference for the later calculation of the margin amount.

[0025] A field-programmable gate array for real-time calculation of a margin amount, comprising a transmitting and receiving unit (11) and a computing unit (12). The transmitting and receiving unit (11) is configured to receive at least one data packet (20) comprising a plurality of data fields (21). These data fields (21) can contain market data (211) and / or trading order data (212). The computing unit (12) stores user position data (121) and a margin calculation program (122) and extracts bit data from at least one of the data fields (21) for performing calculations. Based on the user position data (121) or the trading order data (212) in combination with the market data (211), the computing unit (12) executes the margin calculation program (122) and calculates a margin amount (123).

Claims

[1] Field-programmable gate array for real-time calculation of a margin amount, comprising: a transmitting and receiving unit (11) configured to receive at least one data packet (20), wherein the data packet (20) comprises a plurality of data fields (21), and wherein the data fields (21) may contain market data (211) and / or trading order data (212); and a computing unit (12) connected to the transmitting and receiving unit (11), wherein the computing unit (12) stores user position data (121) and a margin calculation program (122), wherein the computing unit (12) extracts bit data directly from at least one of the data fields (21) on the basis of the data packet (20) received via the transmitting and receiving unit (11) and performs a calculation; wherein the computing unit (12) executes the margin calculation program (122) using the user position data (121) or the trading order data (212) in combination with the market data (211) to calculate a margin amount (123), and wherein an unrealized profit determined on the basis of the market data (211) and the user position data (121) is taken into account in the calculation of the margin amount (123). [2] Field-programmable gate array for real-time calculation of a margin amount according to claim 1, characterized bythat the user position data (121) consist of a combination of several futures products (30) of the same investor and the received data packet (20) is assigned to one of the futures products (30). [3] Field-programmable gate array for real-time calculation of a margin level according to claim 2, characterized by , - that the margin calculation program (122) defines a plurality of calculation weighting factors, each calculation weighting factor being assigned to one of the futures products (30), and - that the computing unit (12) takes these calculation weighting factors into account when calculating the margin amount (123) when executing the margin calculation program (122). [4] Field-programmable gate array for real-time calculation of a margin level according to claim 3, characterized bythat the computing unit (12) executes the margin calculation program (122) in a multi-thread architecture, wherein a plurality of computing subroutines (128) are provided, each of which is assigned to one of the futures products (30). [5] Field-programmable gate array for real-time calculation of a margin level according to one of claims 1 to 4, characterized by that the data package (20) serves as a data source for updating the margin calculation program (122). [6] Field-programmable gate array for real-time calculation of a margin amount according to claim 5, characterized by , - that, if the data fields (21) of the data packet (20) contain market data (211), the computing unit (12) executes a comparison program (129) to compare the market data (211) with the user position data (121); and - that if the comparison result is positive, the margin calculation program (122) is continued and if the comparison result is negative, the market data (211) is saved.

Images