EP3152731A1 - Börsenhandelsplattform - Google Patents
BörsenhandelsplattformInfo
- Publication number
- EP3152731A1 EP3152731A1 EP15727866.4A EP15727866A EP3152731A1 EP 3152731 A1 EP3152731 A1 EP 3152731A1 EP 15727866 A EP15727866 A EP 15727866A EP 3152731 A1 EP3152731 A1 EP 3152731A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- data processing
- processing device
- trading platform
- trading
- platform
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Definitions
- the present invention relates to an exchange trading platform, an exchange trading system and a network interface for an exchange trading system.
- matcher a commercial order processing and execution unit
- order processing system back office
- a trader If a trader is interested in a stock exchange transaction, he places an order on the trading platform. This contains the title concerned, information on whether the title should be sold or bought (buy / sell), the desired number (Quantity) and the price.
- the trading platform keeps the order book (order book).
- order book When an order arrives it is checked if there is an appropriate stored order. If not, the order will be placed in the order book stored and acknowledged to the client. If so, the trade is executed, entered in the sales book and both parties are informed.
- the information about the settled trade then goes anonymously (ie without information about who the two traders involved were) to all traders ('broadcast'). Of course, it is also possible that an order can only be partially implemented, then remains after the execution of the trade remaining remainder order in. Order book.
- any other order variants are also supported (inter alia: 'Market Order', 'Accept Order', 'Fill or ill Order', Jceberg Order ').
- a fast electronic commerce network should support these communication patterns.
- Processing speed in the trading system and processing speed in the dealer system is crucial. While electronic trading systems initially traded in seconds (and thus one order of magnitude faster than human trading systems), today they are trading in the millisecond range and aiming for times of a few microseconds.
- One of the objects of this invention is to eliminate disadvantages of a supercomputer network without losing the advantage of high speed.
- the trading platform has only a limited cache for incoming orders. A trader can overload the trading system and orders can be lost.
- Another challenge is the intelligent inclusion of information on the dealer side.
- the merchant data processing scheme will no longer be able to handle and sort out the flood of information entering it.
- most of these systems are not designed as high-availability systems. This property can be realized later only with great effort.
- an electronic platform for stock market trading is provided for a plurality of merchant data processing devices in communication with the trading platform, the trading platform having a
- the controlling unit is adapted to process orders generated by the merchant data processing equipment, thereby matching the received orders with stored orders stored in the storage means and forming an order book and, depending on a result of the reconciliation, to store an order received in the storage means or at least partially execute it on a matching order already stored in the storage medium and to delete the appropriate order at least partially in the storage means or marked as executed, such received messages to the sending merchant data processing device to acknowledge and also in a broadcast mode broadcast messages with information about executed orders and / or stored in the memory means Sending orders to a plurality of connected merchant data processing devices, as well as transaction data for the executed orders Submit order processing system.
- the trading platform is characterized in particular by the fact that the controlling unit of the trading platform data processing device is designed as a dedicated logic circuit.
- the controlling unit of the matcher which selectively controls execution or storage of orders, is the central core of the trading platform. This unit reconciles incoming orders with the order book, carries out trades (including confirmation, entry in the trading book and / or communication with the back office, possibly message to many or all participants) and / or enters orders into the order book, removes Orders processed or modified entries for partially executed orders.
- the "back office” is the order processing system in which the order execution is completed and, for example, the corresponding financial transaction is processed.
- the transaction data submitted to the back office after execution of the order includes the quantity and price of the securities traded and allows the transaction to be reconstructed financially and physically (delivery of the securities).
- a dedicated logic circuit is a hardware device or a group of hardware devices in which the logic functions by which signals or data are processed are implemented in a predetermined manner by predetermined and predetermined circuits in the hardware itself.
- the logic circuit thus forms a functional structure with interconnected blocks. This is in contrast to main processors of conventional computers (including server computers) which are formed as 'generic', universally applicable microprocessors, which in a memory (presented as software) sequence of commands then only this software and not the circuit itself is earmarked.
- Such a dedicated logic circuit can be designed, for example, as a Field Programmable Gate Array (FPGA).
- FPGA Field Programmable Gate Array
- An FPGA is dedicated by the configuration, i. custom designed.
- Such an FPGA may include memory cells (e.g., EEPROM, EPROM, SRAM, Flash) in which the configuration is stored. Each time it is powered up, the configuration is transferred to the actual circuitry.
- the FPGA can also be permanently programmed by the connections between the switching units are permanently established, for example. With the so-called 'antifuse' technology.
- an FPGA In an FPGA or other configurable logic circuit, circuit structures are created using hardware description languages or schematics and subsequently these structures for configuration transfer the block. As a result, certain switch positions are activated and / or deactivated in the configurable logic circuit. This results in a concretely implemented digital circuit, which in general operates in a highly parallel manner, because every influence of the switch position works in parallel. In contrast, even the fastest microprocessors perform at most few and usually no operations in parallel.
- CPLDs Complex Prograinmable Logic Devices
- ASICs application-specific integrated circuits
- Dedicated logic circuits are common in the art as standard building blocks for performing given, always identical tasks, for example in control, for the conversion (encoding / decoding) of data, digital filtering etc.
- the advantages that can be achieved by the procedure according to the first aspect are striking.
- the data processing speed is much due to the parallelism of data processing resulting in dedicated logic circuits higher than with the use of microprocessors of the common type, even in comparison with highly refined and powerful Mehrkernreaoreri.
- the latencies are significantly reduced in comparison to the prior art.
- the matcher can be much smaller than the data centers used for this purpose in the prior art, and power consumption is only a fraction of the power consumption of these.
- the procedure according to the first aspect of the invention also allows the bypassing of PCIe transitions, as well as a network stack processing in the operating system is bypassed:
- the latency-critical processing can be performed completely independently by the dedicated logic in the matcher.
- the data does not have to be communicated via PCIe to the central processor (CPU) and possibly processed by the operating system through network stack processing. As a result, significant time is gained. It has been shown that the latency of 30-50 microseconds usual for very fast Platt forms according to the prior art can be reduced to about 10 microseconds.
- the requirements of the tasks to be performed by a matcher can be fulfilled almost ideally.
- This is highly complex in a sense.
- the complexity results from the sheer number of messages that can be processed virtually simultaneously, whereby for reasons of fairness the actual core task of the matcher can only be divided into parallel subtasks - the actual "matching" must take place according to strict stock exchange rules
- This is reflected in the sheer size, maintenance and energy consumption of the invention, and it is an underlying realization of the invention that, despite all this complexity, the It is possible to reduce tasks to a manageable core, and that these can be ideally implemented by a hardware solution - as such, a dedicated logic circuit - resulting in a massive reduction in complexity.
- a stock exchange trading system which comprises:
- a plurality of merchant data processing systems A plurality of merchant data processing systems
- a trading platform - in particular according to the first aspect - having a trading platform data processing device with storage means and a controlling unit and at least one switch via which the
- each merchant data processing device is capable of delivering messages with orders contained therein to the trading platform and the controlling unit is adapted to process these messages, matching the received orders with stored orders stored in the storage means and forming an order book, and dependent store a received order in the storage means from a result of the matching or at least partially execute it at a matching order already stored in the storage means, delete the matching order at least partially in the storage means or marked as executed and save or at least partially execute to acknowledge the merchant data processing device and to transmit transaction data for the executed orders to a job postprocessing system; - And wherein a switching logic of the at least one switch only messages from a dealer data processing device to the trading platform data processing device, from the
- Trading platform data processing device to a specific merchant data processing device and from the trading platform data processing device to a group of
- the second aspect of the invention also relates to an electronic platform for such a system.
- a "merchant data processing facility” may be an integrated computer, but it does not have to be a "personal computer” entity. It can also be divided into several, possibly spatially separated, affiliations; the same applies to other data processing devices described in this application.
- switch generally refers to a network node that allows selective forwarding to certain (single or multiple, possibly also to all) connected devices - in the present context, the switch allows at least one order received from a merchant data processing device
- a switch in the sense used here is a device guaranteeing this functionality, regardless of whether all the associated elements are physically adjacent to one another (and, for example, in one common housing) or not.
- the switch may also be referred to as an "aggregation switch".
- a switch according to the second aspect of the invention also has, in addition to this generally applicable characteristic, the switching logic specified above.
- the switch is designed in particular as a dedicated logic circuit.
- a group of retail data processing devices may be a predetermined group interested in a particular segment of traded securities.
- the system may be arranged to treat all subscribers who are interested in a particular title in the same way as regards information relating to that title. Often the system will even be set up that in broadcast mode always all participants will receive the corresponding data.
- the switch-for example, designed as a dedicated logic circuit- provides physically different paths for data streams from the dealers in the direction of the matcher on the one hand and the matcher in the direction of the dealer on the other hand.
- the first path (from the dealer to the matcher) provides a wide line of messages coming from one of the inputs towards the matcher, while the second path (from the matcher to the dealers) optionally has a purp-to-pimple shadowing (message goes to the output, which is assigned to a dealer) or a broadcast connection (message is sent to a plurality of outputs, in particular to all outputs).
- a first arbiter (responsible for the transmission of messages in the direction of the dealer to the matcher, "uplink”) an uplink switch (“multiplexer”) so sets that the corresponding connection from the transmitting data processing device or the corresponding network node up to the matcher out.
- a second arbiter (responsible for sending messages towards the matcher to the merchant, "downlink”) places a downlink switch ("demultiplexer”) such that either a connection from the matcher or network node to or to a particular merchant / network node Group / to all arises.
- the communication pattern resulting in electronic stock exchange trading is implemented immediately in the switching logic of the network. This is so that message streams are not possible if they do not correspond to the GayLinikationsmuster.
- the procedure according to the second aspect of the invention has, for example, the advantage of simplicity - also in terms of security and review of procedures.
- the advantage of the improved network speed also results from this procedure, since the communication paths are predetermined by the switching logic and thus eliminating the processing of corresponding predetermined in a software steps.
- the procedure is also advantageous for reasons of fairness. In a simple manner, it is ensured that, for example, merchant computer to merchant computer messages are not stopping certain operations or accelerated - since such message paths do not occur in the trading system.
- the procedure according to the second aspect of the invention does not rule out that the components of the trading system are in communication with each other by other means, for example via a conventional, slower network.
- a merchant data processing apparatus of a perch system in particular according to aspect 1 or 2, which, as known per se, has a central unit with a central processor (CPU), central storage slots with an operating system and generally also having a user interface and a communicating with the central unit network interface as an interface to the trading platform down.
- a central unit with a central processor (CPU), central storage slots with an operating system and generally also having a user interface and a communicating with the central unit network interface as an interface to the trading platform down.
- Such an interface may, for example, be implemented by on a network card.
- the network interface now has storage means for storing status data of a current state of the stock exchange trading system, for example for tracking the order book at least for the title relevant to the corresponding trader, and a computing device which is designed as a dedicated logic circuit. such that the computing device is capable of issuing orders in the form of messages to the trading platform according to specifications of the central unit as a function of the status data.
- Dedicated logic circuits for example ASICs or FPGAs on network cards are known per se, for example for protocol conversions or other established tasks, possibly also in connection with alogtrading.
- the fast programmable arithmetic unit on the network interface which, without conversion from a standard protocol, directly picks up the incoming data and delivers jobs directly, and which also has fast local memory, enables the merchant to implement the third aspect of the invention to react even faster to events in the stock market.
- the network interface has at least one network interface bus via which incoming messages are forwarded from the network to the central inlay and, on the other hand, forward messages (orders or the like) from the central infeed to the network (specifically: the network interface) become.
- the computation device designed as dedicated logic circuit is then, for example, likewise arranged directly on this bus, ie it is capable of both reading data on the bus and of writing data to it.
- the network interface bus may have a 'read bus' (over which data coming from the network is running) and a 'write bus' (over which data is to be sent to the network), in which case the computing device will communicate with both the read bus 'and the' write bus' is in communication.
- the network interface bus may optionally be different than the bus of the central unit (WCB Workstation Communications Bus), in which case an interface (WI Workstations Interface) connects the two bus systems.
- the dealer data processing device can be informed very promptly when a specific event arrives (eg if a predetermined market value has been reached, the number of entries in the order book (or in the part of the order book kept in copy by the merchant data processing device) is undershot, Etc.).
- a specific event eg if a predetermined market value has been reached, the number of entries in the order book (or in the part of the order book kept in copy by the merchant data processing device) is undershot, Etc.
- the merchant data processing device can be additionally relieved. Often it is worthwhile for the trader to know the time for the individual trade in a very high time resolution. To achieve this goal, there is an optional timer on the communication card.
- the third aspect also relates to a network interface (in particular network card) for such a merchant data processing device.
- a network interface in particular network card
- the trading system can be constructed tree-hierarchically with multiple levels, the network nodes of the levels each by
- Switches are formed with at least one "uplink” connection and several "downlink” connections.
- Each merchant data processing device is connected to the trading platform data processing device by the same number of levels (with the same number of switches).
- Each merchant data processing device has an identical network interface (an identical network card TB).
- trading platform data processing devices may also be used, each one of them Trading platform data processing device is responsible for one or more titles.
- the forwarding to the responsible / correct trading platform data processing device can take place in the "uppermost" network node (switch), for example, based on the destination address.
- the "top” network node may have another output storing all the relevant messages relayed by the network node in an archive A.
- the archiving unit that logs the jobs and transactions will be located directly at the network node (ie at the switch ) and not necessarily on the matcher itself.
- Such an archive can, for example, be available to the "back-office" for further processing.
- the advantage is that the trading platform data processing device is not burdened with the archiving: the data are picked up there by the quasi-independent archive, where they pass by anyway.
- the matcher frees himself from a very important task, which complements very well with the other measures to increase the speed.
- each trading platform data processing device may have its own archive (A *).
- the merchant data processing devices keep copies of the parts of the order book of interest to the respective merchants - these can comprise the entire order book. -
- the time between order and response of the trading system can be accurately measured by said timer, for example, for the purpose of debugging.
- the system can be set up to immediately delete all open orders from a specific trading participant in a single order from the system (Mass Cancel MC).
- the orders to the trading platform can be seen with a time limit and automatically deleted after expiry of the time limit (time-of-life TOL).
- Dealer be present a unit that processes the first valid message (eg forwarding) and suppresses identical copies.
- FIG. 1 shows a representation of an exchange trading system
- FIG. 5 shows elements of a trading platform data processing device
- - Fig. 6 shows schematically elements of a matcher.
- FIG. It is a hierarchical network that is optimally adapted to the communication needs of stock market applications.
- there are several trading platform data processing arrangements (matcher M [... M m ), ie the function of the matcher is split over several matchers. Each matcher is then responsible for one or more titles. The forwarding to the correct matcher can be done by the central switch S 0, for example, based on the destination address.
- the central switch over which the merchant data processing devices (referred to herein as clients (C t , ... C k )) communicate with the matcher.
- an optional further switch level with decentralized switches S is also shown, wherein at least one merchant data processing device (generally several merchant data processing areas) are attached to each of these decentralized switches. It is also an extension to even more levels and accordingly even more tree-like branched architecture possible.
- the distributed switch approach allows for good scalability and can be viewed when the central switch does not physically have enough interfaces to all of its vendors. It may be advantageous if each merchant is connected to the central switch via the same number of levels, since then the forwarding times are approximately identical, and in the approach discussed above with sorting according to a time, the waiting time before a message (eg order) is up forwarded can be kept shorter.
- TB designates the network interfaces (Trader Boards) of the merchant data processing systems.
- Di stands for the so-called downlinks of the switches, U or U, for the respective uplink.
- the top-level switch (ie, the topmost network node) may also optionally have another output that stores all relevant messages forwarded through that switch in an archive A, and thus this information, for example, the back-office processing system for further processing provide.
- This option exists regardless of whether only one matcher or as in Figure 1 multiple matchers are used. When using multiple matchers, however, it is particularly favorable because it enables central archiving. Alternatively or additionally, everyone can Matcher has its own archive A, (in the top level in Fig. 1).
- the internal circuit of a switch S is shown symbolically in FIG.
- the structure shown in FIG. 2 is used in particular in the top-level switch, i. the top node of the network.
- switches of the underlying levels may be designed accordingly, whereby at least the broadcast function should also be implemented on the possible underlying levels.
- Fig. 2 generally includes a "U” in the name for so-called “uplinks” (in the direction of matcher), and “D” for “downlinks” (in the direction of traders).
- Fig. 2 only one matcher M is drawn; If the matcher function is split up among multiple matchers, a split is still made on the uplink side of the switch based on the addresses.
- a message (data packet, in particular with a job) is transmitted by the merchant data processing device (referred to here as client Ci ... C ”) via a transmitter Tx (a transmitter module).
- a conversion to a serial data format takes place. This is received in the switch S. and, for example, again converted in parallel (KDRxo ... KDRx ,,) and written into the buffer memory (UBo ... UB n ).
- a dedicated buffer memory is assigned to each first receiver (KDRx 0 ... KDRx n , ie each downlink). In alternative embodiments, it would also be possible to provide central buffering.
- the first arbiter UA decides which packet is forwarded and accordingly controls the multiplexer UM.
- the data is now converted back into a serial protocol by the first switch transmitter KUTx, transmitted and converted back into a parallel format by a matcher receiver MRx. Via an optional matcher buffer MB, the message is sent to the matcher M.
- matcher M now sends a message (for example the acknowledgment of a job) to a specific dealer Co... Cn, then it is converted back into a serial format via a matcher transmitter MTx, transmitted and received by the second switch receiver KURx and again converted into parallel format.
- the second arbiter DA now decides, based on the address, which dealer CO ... Cn should receive the message and closes the corresponding switch DDo. DD1, ... or DDn. Subsequently, the message is transmitted via a second switch transmitter KDTxo, ..KDTxn to the corresponding merchant receiver RxO ... Rxn and is now available to the merchant Cn for further processing.
- the drawn in Figure 2 circuit according to the second aspect of the invention corresponds exactly to the erKindlelinten communication pattern.
- the switches operated by the arbitems UA, DA allow for communications of the three types a) to c), but not, for example, direct point-to-point communication between the dealers.
- each dealer data processing device with a timer here in the form of a precisely synchronized clock CLKo ... CLKn equipped.
- the clock reference for this clock can be specified by the matcher or the highest network node (the topmost switch) by periodically sending reference time messages, which then readjust the local clocks CLKo .. CLKn.
- the actual clock cycle can be generated both by a local oscillator and by the global network clock, which in turn is specified by the matcher M.
- each message is now provided with a time stamp by the client C0... C11, and the network system ensures that older messages always arrive first at the matcher; the network "sorts" the messages based on time.
- This function is done here by the first arbiter UA first forwarding the older message. For this selection to be fair, it must at least wait for the time (waiting time) which arises due to unequal construction methods (eg cable lengths, module tolerances). Only then can he make his decision. On each uplink or at least on the uplink of the topmost network node, the messages are therefore ordered strictly in time and thus arrive in the timely order (also over multiple switch levels) in the trading system M. This optional time sorting principle can be combined with any of aspects 1 through 3.
- an optional credit system can be introduced: The trading system assigns C0 to all traders. , , C11 loans. A loan allows the merchant to send an order (or, more generally, a message). The number of credits is so great that when the system is unloaded, a single merchant can still work at full speed and will not overflow the individual buffers even when the system is heavily loaded. For each message, a loan is debited and credited again when the message from the matcher to the. Dealer was acknowledged.
- arrows at the bottom of Fig. 2 indicate the use of a local time base for providing the data packets with a pre-shipment timing as described above. While in FIG. 2 receivers and transmitters of the merchant data processing devices are drawn separately, they can be realized in reality in one and the same hardware component, for example as separately configured units in one and the same FPGA.
- FIG. 3 A redundant system is shown in FIG. 3.
- the principle of redundancy is shown here by means of an example of embodiment in which the matcher is responsible for all titles.
- the transfer to a system as shown in Fig. 1 with matchers Mi, .... M m with different responsibilities is readily possible, in which case at least one matcher, preferably all matchers duplicated and each with two top network node SQ.O, SQ, O are connected. If an archive A connected to the top network node, this can also be redundant, ie executed twice.
- both the matcher M on M * is doubled, as well as the network with the switches Sij with Sij *.
- One of the two matchers M or M * can (in corresponding exemplary embodiments) specify the time (time master, i.e. both the network clock and the reference time); the other matcher then takes over. If a matcher fails, he will make other matchers the time master, if he has not already been the time master before. In actual order processing, both matchers have a completely identical structure, and in both, order bookkeeping and order execution take place independently of each other.
- the received message streams are forwarded to the other trading system (PS (t) and PS * (t)). So that the own message streams are not preferred by the transmission, they are delayed by a delay time TD or TD *.
- the prioritizers P and P * now ensure that messages with earlier times are taken first. In addition, in the prioritizer P identical messages (coming through the 2 independent networks) are suppressed. Thus it can be guaranteed that identical, temporally ordered message streams are transmitted to the actual trading systems M and M *.
- WI workstation interface
- NI the interface to the network nodes of the trading system.
- the network interface is set up for a redundantly designed trading system and accordingly has two interfaces NI and ⁇ , which are provided for connection to a respective network node (switch) of the redundant system.
- an input filter (F First) will ensure that identical messages do not passed twice, ie the first valid Nachiicht forwarded and suppressed all other copies.
- Incoming messages can be read and processed via the read bus RB as known per se by the central unit.
- the central storage of titles takes place in the OB (Order Book) and is permanently tracked by the dedicated logic based on the incoming messages (Market Updates).
- the network interface (network card) of each merchant data processing device also has a memory UM for the titles relevant to the corresponding merchant and a fast programmable arithmetic unit UF (user FPGA) designed as a dedicated logic circuit. This enables the trader to react even faster to events on the stock market.
- a memory UM for the titles relevant to the corresponding merchant
- a fast programmable arithmetic unit UF user FPGA
- the flow control unit FC implements a flow control to realize the optional overload protection based on credits discussed above. Only a certain number of unacknowledged jobs are allowed before the merchant data processing device denies further jobs.
- the network interface has at least one network interface bus, which is functionally divided into a write bus ("write bus”) and a read bus (“read bus”).
- a write bus write bus
- read bus read bus
- the computing device which is designed as a dedicated logic circuit, is arranged directly on this bus, i. It is capable of both reading data on the bus and writing data about it to other units.
- the network interface bus is different from the central unit (WCB Workstation Communications Bus), with one interface (WI Workstations Interface) connecting the two bus systems. This means that the internal data exchange of the central unit - for example between its CPU and peripherals or data memories - does not burden the network interface bus.
- the dealer data processing direction can also be informed very promptly when a particular event arrives (eg a predetermined market value has been reached). With this additional function of the dealer card, the dealer data processing device can be additionally relieved. Often it is desirable for the trader to know the time for the individual trade in a very high time resolution.
- a timer CL is optionally located on the network interface TB - this may be, for example, a timer CLK of the type and function described above, or else a separate timer. If the trader wants to react very quickly to a change in the market, then he first of all transmits the data on the network interface to the merchant data processing device, processes it there and then writes it back to the network interface.
- Figure 5 shows an embodiment of a trading platform (network nodes (switches) are not shown).
- OB designates the data store with the order book
- MA the actual matcher whose controlling unit according to the first aspect of the invention is designed as a dedicated logic circuit, in particular as an FPGA.
- Connected to and integrated with the matcher for example, are a synchronization unit SY Sync which, together with the corresponding unit of the other matcher in a system with two redundant matchers (FIG. 3), coordinates the mutual monitoring and synchronization of the two matchers Stop detector SD and an avalanche detector AD (detector of avalanche news).
- the messages are first optionally channeled through an input filter (F First) and / or a verification processor VP, which checks the authorization of the customer to execute a specific order and / or a time-of-life processor (TOL) Automatically deletes orders and upon expiration of a time limit are provided and / or a Mass Canccl entity can intervene to immediately delete all open orders of a particular trading participant in a single order from the system.
- F First input filter
- VP verification processor
- TOL time-of-life processor
- WL Write Log write event log file
- an order response unit OR order response; generates 'private' messages (intended only for a merchant data processing device) for one
- MU Market Update generates messages to all for a specific order
- the trading platform data processing scheme is drawn with two network interfaces NI Port, ⁇ Port), i. it is designed for a redundant system as shown in FIG.
- FIG. 6 very schematically shows the matcher M.
- the controlling unit SE which is embodied as a dedicated logic circuit, in particular as an FPGA
- this has a data memory.
- This may be designed as RAM and / or optionally distributed to various physical units.
- RAM technologies can be used and optionally combined with one another, for example SDRAM, QDRI I, SRAM, where, for example, SDRAM can be used as the actual data memory and QDRII and / or SRAM for accelerator structures.
- the use of other memory such as SSD is not excluded.
- the data memory or units thereof can be physically integrated with the dedicated logic circuit and / or be present separately therefrom, for example on a circuit board.
- the data store contains the order book that is managed by the controlling unit. If an order arrives from the switch, the controlling unit matches it with the data available in the order book. If there is an at least partially matching order, the appropriate orders are executed; This is acknowledged to the two clients and transmitted to the order processing system BO and anonymized by broadcast to all participants. If the incoming order can only be partially or not executed due to the lack of suitable stored jobs, it is stored in the data memory. Even then his input - but of course not the execution - acknowledged. The process of matching incoming orders with stored orders according to given rules is the core task of the matcher.
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
Description
Claims
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CH00866/14A CH709741A1 (de) | 2014-06-05 | 2014-06-05 | Börsenhandelsplattform. |
PCT/CH2015/000083 WO2015184557A1 (de) | 2014-06-05 | 2015-06-03 | Börsenhandelsplattform |
Publications (1)
Publication Number | Publication Date |
---|---|
EP3152731A1 true EP3152731A1 (de) | 2017-04-12 |
Family
ID=51260533
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP15727866.4A Ceased EP3152731A1 (de) | 2014-06-05 | 2015-06-03 | Börsenhandelsplattform |
Country Status (5)
Country | Link |
---|---|
US (3) | US20170109825A1 (de) |
EP (1) | EP3152731A1 (de) |
CH (1) | CH709741A1 (de) |
SG (1) | SG11201610187RA (de) |
WO (1) | WO2015184557A1 (de) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110137785A1 (en) * | 2009-12-04 | 2011-06-09 | Lutnick Howard W | Multicomputer distributed processing of trading information |
CA2960047C (en) | 2014-10-08 | 2024-04-30 | Tsx Inc. | Selective delayed and undelayed database updating |
US20170004575A1 (en) * | 2015-07-01 | 2017-01-05 | Chicago Mercantile Exchange Inc. | Dissemination of order status information present on an electronic exchange |
Family Cites Families (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0411748A3 (en) * | 1989-06-02 | 1991-11-21 | Reuters Limited | System for matching of buyers and sellers with risk minimization |
US20050152406A2 (en) * | 2003-10-03 | 2005-07-14 | Chauveau Claude J. | Method and apparatus for measuring network timing and latency |
US7921046B2 (en) * | 2006-06-19 | 2011-04-05 | Exegy Incorporated | High speed processing of financial information using FPGA devices |
US8566218B2 (en) * | 2008-10-07 | 2013-10-22 | Chicago Mercantile Exchange Inc. | Systems and methods for matching one or more incoming order to a standing order as a function of an inner market parameter |
EP2370946A4 (de) * | 2008-12-15 | 2012-05-30 | Exegy Inc | Verfahren und vorrichtung zur hochgeschwindigkeitsverarbeitung von finanzmarkttiefendaten |
US9501795B1 (en) * | 2010-08-23 | 2016-11-22 | Seth Gregory Friedman | Validating an electronic order transmitted over a network between a client server and an exchange server with a hardware device |
US9038189B1 (en) * | 2011-02-24 | 2015-05-19 | Rexante, LLC | System and method for programming a trading system |
US8515858B2 (en) * | 2011-05-19 | 2013-08-20 | Omx Technology Ab | Method and arrangement for pre-match risk validation of orders in an automated exchange |
US20140143121A1 (en) * | 2012-02-23 | 2014-05-22 | Rexante Systems, Inc. | System and Method for Programming a Trading System |
US8655769B2 (en) * | 2012-03-16 | 2014-02-18 | Cape City Command, Llc | Method and system for improving equity trade order acknowledgement times |
US11436672B2 (en) * | 2012-03-27 | 2022-09-06 | Exegy Incorporated | Intelligent switch for processing financial market data |
US10121196B2 (en) * | 2012-03-27 | 2018-11-06 | Ip Reservoir, Llc | Offload processing of data packets containing financial market data |
US20140249979A1 (en) * | 2013-03-01 | 2014-09-04 | Secodix Corporation | Enhancing the handling speed of electronic financial services messages |
US20150127509A1 (en) * | 2013-11-07 | 2015-05-07 | Chicago Mercantile Exchange Inc. | Transactionally Deterministic High Speed Financial Exchange Having Improved, Efficiency, Communication, Customization, Performance, Access, Trading Opportunities, Credit Controls, and Fault Tolerance |
-
2014
- 2014-06-05 CH CH00866/14A patent/CH709741A1/de not_active Application Discontinuation
-
2015
- 2015-06-03 US US15/315,408 patent/US20170109825A1/en not_active Abandoned
- 2015-06-03 WO PCT/CH2015/000083 patent/WO2015184557A1/de active Application Filing
- 2015-06-03 SG SG11201610187RA patent/SG11201610187RA/en unknown
- 2015-06-03 EP EP15727866.4A patent/EP3152731A1/de not_active Ceased
-
2019
- 2019-12-28 US US16/729,376 patent/US20200349645A1/en not_active Abandoned
-
2022
- 2022-07-12 US US17/812,110 patent/US20220343424A1/en active Pending
Non-Patent Citations (2)
Title |
---|
None * |
See also references of WO2015184557A1 * |
Also Published As
Publication number | Publication date |
---|---|
US20170109825A1 (en) | 2017-04-20 |
US20220343424A1 (en) | 2022-10-27 |
US20200349645A1 (en) | 2020-11-05 |
SG11201610187RA (en) | 2017-01-27 |
CH709741A1 (de) | 2015-12-15 |
WO2015184557A1 (de) | 2015-12-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
DE112009002716B4 (de) | Verbessertes automatisches Handelssystem | |
DE69907776T2 (de) | Verfahren und Vorrichtung zur Identifizierung gefährdeter Bauteile in einem System mit redundanten Bauteilen | |
DE102019122363A1 (de) | Programmierbare doppelreihige arbeitsspeichermodul-beschleunigerkarte (dimm-beschleunigerkarte) | |
DE69725687T2 (de) | Transaktionsübertragung zwischen Datenbussen in einem Rechnersystem | |
DE68909426T2 (de) | Datenverarbeitung und -übertragung. | |
DE3134428T1 (de) | Cached multiprocessor system with pipeline timing | |
DE3586486T2 (de) | Interface-vorrichtung zwischen mindestens einem kanal und mindestens einem bus. | |
DE3888353T2 (de) | Unterbrechungsknoten zum vorsehen von unterbrechungsanforderungen auf einem anstehenden bus. | |
EP3152732A1 (de) | Börsenhandelssystem | |
DE19924922A1 (de) | System und Verfahren für Nachrichtenübermittlung zwisfchen Netzwerkknoten, die durch parallele Verbindungen verbunden sind | |
US20220343424A1 (en) | Stock exchange trading platform | |
DE112012004551T5 (de) | Mehrkernverknüpfung in einem Netzprozessor | |
DE112017004907T5 (de) | System, Vorrichtung und Verfahren zur Durchführung einer verteilten Arbitration | |
DE112019003854T5 (de) | Flusssteuerungssichtbarkeit | |
DE102022200763A1 (de) | System und verfahren für kundenspezifische erinnerungen | |
EP3152884B1 (de) | Verfahren zur weiterleitung von daten zwischen computersystemen, computernetz-infrastruktur sowie computerprogramm-produkt | |
DE112004002043T5 (de) | Verfahren, System und Programm zum Aufbau eines Pakets | |
DE102012017339B4 (de) | Rechnersystem | |
DE102004003102A1 (de) | System und Verfahren zum Bestimmen einer Transaktionszeitüberschreitung | |
DE112016004362T5 (de) | Niederlatenz-internode-kommunikation | |
EP1477882A2 (de) | Dezentrales, tokenbasiertes Accountingsystem für autonome, verteilte Systeme | |
DE102016207144B4 (de) | Auswertungssystem | |
DE69712587T2 (de) | Mikrorechner mit Zugriff auf einen externen Speicher | |
DE10310055A1 (de) | Systeme und Verfahren zum Erfassen und Kompensieren von Kleinblockdatenübertragungen | |
DE3247083A1 (de) | Mehrprozessorsystem |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
17P | Request for examination filed |
Effective date: 20161205 |
|
AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
AX | Request for extension of the european patent |
Extension state: BA ME |
|
RIN1 | Information on inventor provided before grant (corrected) |
Inventor name: GUNZINGER, ANTON Inventor name: BURKHARD, HANS Inventor name: BERNER, MARKUS Inventor name: BLUM, EDGAR Inventor name: PIANTONI, RICO Inventor name: ZAHND, SAMUEL Inventor name: MUELLER, DAVID |
|
DAV | Request for validation of the european patent (deleted) | ||
DAX | Request for extension of the european patent (deleted) | ||
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
17Q | First examination report despatched |
Effective date: 20171122 |
|
REG | Reference to a national code |
Ref country code: DE Ref legal event code: R003 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE APPLICATION HAS BEEN REFUSED |
|
18R | Application refused |
Effective date: 20190606 |