Methods and systems for detecting causes of margin calls using machine learning

Abstract

A method for identifying and capturing all relevant features involved in a margin call process so that a root cause can be automatically identified using machine learning is provided. The method includes: obtaining data related to a securities portfolio; automatically determining data patterns related to triggering a margin call and corresponding characteristics that contributed to triggering the margin call; automatically generating a margin call model based on the data patterns and characteristics; and evaluating a first securities portfolio using the margin call model to determine information relevant to whether a first margin call should be triggered for the first securities portfolio.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of U.S. Provisional Patent Application Serial No. 62 / 946,096, filed December 10, 2019, which is hereby incorporated by reference in its entirety. Technical Field

[0003] The present technology generally relates to methods and systems for detecting the causes of margin calls, and more particularly, to methods and systems for identifying and capturing all relevant features involved in the margin call process to enable automatic identification of the root cause through the use of machine learning. Background Art

[0004] Today, many financial institutions use margin calls to ensure that credit investors remain within maintenance margin limits. Historically, the decision to issue margin calls to investors has been made after the fact, with varying degrees of success in reducing investment risk.

[0005] One drawback of using traditional analytical methods to determine whether an investor requires a margin call is that, in many cases, the analysis is inefficiently completed using multiple systems and data elements, after the portfolio has already exceeded an arbitrary threshold. Furthermore, the process is manual and subjective due to the need for manual interpretation of data from different regions. Consequently, financial institutions issuing margin calls expend significant resources on post-event information, failing to effectively mitigate risk.

[0006] Therefore, transaction modeling is needed to provide investors with information about the various factors that lead to margin calls so that they can effectively correct any deficiencies. Furthermore, transaction modeling is needed to provide financial institutions with information about expected margin calls so that they can correct deficiencies without taking on unnecessary risk exposure. Finally, the margin call process and how margin call features are interpreted across regions and locations need to be standardized so that there is a single interpretation that eliminates human bias. Summary of the Invention

[0007] The present disclosure, through one or more aspects, embodiments, and / or specific features or subcomponents thereof, provides various systems, servers, devices, methods, media, programs, and platforms for identifying and capturing all relevant features involved in the margin call process so that the root cause can be automatically identified through the use of machine learning.

[0008] According to one aspect of the present disclosure, a method for identifying and capturing all relevant features involved in a margin call process so that a root cause can be automatically identified using machine learning is disclosed, the method being implemented by at least one processor. The method may include: obtaining, by the at least one processor, a plurality of data related to a plurality of securities portfolios; automatically determining, by the at least one processor, using the plurality of data, at least one data pattern related to triggering a margin call and at least one corresponding characteristic that causes the triggering of the margin call; automatically generating, by the at least one processor, at least one margin call model based on the at least one data pattern and the at least one characteristic; and evaluating, by the at least one processor, the first securities portfolio using the at least one margin call model to determine first information related to whether to trigger the first margin call for the first securities portfolio.

[0009] According to an exemplary embodiment, the evaluating may further include: selecting, by the at least one processor, a first data pattern among the at least one data pattern and a first characteristic among the at least one corresponding characteristic, the first data pattern and the first characteristic being related to triggering the first margin call for the first securities portfolio; and displaying, by the at least one processor, the first information and the selected first data pattern among the at least one data pattern and the selected first characteristic among the at least one corresponding characteristic on a graphical user interface via a display.

[0010] According to an exemplary embodiment, the at least one margin call model may be automatically generated using a machine learning analysis of the plurality of data.

[0011] According to an exemplary embodiment, the at least one margin call model may include at least one of a data model, a mathematical model, a process model, and a machine learning model.

[0012] According to an exemplary embodiment, the method may further include: evaluating a second securities portfolio using the at least one margin call model to determine second information related to whether a future second margin call for the second securities portfolio is triggered; and displaying the second information on a graphical user interface via the display.

[0013] According to an exemplary embodiment, the graphical user interface may include at least one panel displaying monitoring information related to the first securities portfolio and / or classification information related to the at least one characteristic.

[0014] According to an exemplary embodiment, the at least one characteristic may include at least one of the following: a securities appreciation characteristic, a securities depreciation characteristic, an asset market value difference characteristic, a securities market value difference characteristic, a securities market value purchase characteristic, an exposure difference characteristic, a cash market value difference characteristic, a collateral difference characteristic, a cash market value sale characteristic, a discount difference characteristic, a securities market value sale characteristic, a securities liquidation value purchase characteristic, a cash market value appreciation characteristic, a cash market value depreciation characteristic, a pledge market value outflow characteristic, a pledge market value difference characteristic, a cash market value purchase characteristic, a derivative exposure increment characteristic, a derivative initial margin exposure increment characteristic, a pledge liquidation value inflow characteristic, an accrued interest accumulation characteristic, a pledge market value inflow characteristic, a pledge market value appreciation characteristic, a pledge market value depreciation characteristic, a derivative negative market value increment characteristic, and an unlimited pledge difference characteristic.

[0015] According to an exemplary embodiment, the method may further include: receiving feedback information from at least one user via a graphical user interface; and updating the at least one margin call model based on the feedback information.

[0016] According to an exemplary embodiment, the feedback information includes at least one of a first information accuracy ranking, a second information accuracy ranking, and a characteristic accuracy ranking.

[0017] According to an exemplary embodiment, the method may further include: parsing the plurality of data; identifying personally identifiable information in the parsed plurality of data; and anonymizing the plurality of data by removing the identified personally identifiable information.

[0018] According to one aspect of the present disclosure, a computing device is disclosed that is configured to implement a method for providing information related to triggering a margin call for a first portfolio of securities. The computing device includes a processor, a memory, and a communication interface coupled to each of the processor and the memory. The processor is configured to: obtain a plurality of data related to a plurality of portfolios of securities; automatically determine, using the plurality of data, at least one data pattern related to triggering a margin call and at least one corresponding characteristic that causes the triggering of the margin call; automatically generate at least one margin call model based on the at least one data pattern and the at least one characteristic; and evaluate the first portfolio of securities using the at least one margin call model to determine first information related to whether to trigger the first margin call for the first portfolio of securities.

[0019] According to an exemplary embodiment, for the evaluation, the processor may be further configured to: select a first data pattern among the at least one data pattern and a first characteristic among the at least one corresponding characteristic, the first data pattern and the first characteristic being related to triggering the first margin call for the first securities portfolio; and display the first information and the selected first data pattern among the at least one data pattern and the selected first characteristic among the at least one corresponding characteristic on a graphical user interface via a display.

[0020] According to an exemplary embodiment, the processor may be further configured to automatically generate the at least one margin call model by using a machine learning analysis of the plurality of data.

[0021] According to an exemplary embodiment, the at least one margin call model may include at least one of a data model, a mathematical model, a process model, and a machine learning model.

[0022] According to an exemplary embodiment, the processor may be further configured to: evaluate a second securities portfolio using the at least one margin call model to determine second information related to whether a future second margin call for the second securities portfolio is triggered; and display the second information on a graphical user interface via the display.

[0023] According to an exemplary embodiment, the graphical user interface may include at least one panel displaying monitoring information related to the first securities portfolio and / or classification information related to the at least one characteristic.

[0024] According to an exemplary embodiment, the at least one characteristic may include at least one of the following: a securities appreciation characteristic, a securities depreciation characteristic, an asset market value difference characteristic, a securities market value difference characteristic, a securities market value purchase characteristic, an exposure difference characteristic, a cash market value difference characteristic, a collateral difference characteristic, a cash market value sale characteristic, a haircut difference characteristic, a securities market value sale characteristic, a securities liquidation value purchase characteristic, a cash market value appreciation characteristic, a cash market value depreciation characteristic, a pledge market value outflow characteristic, a pledge market value difference characteristic, a cash market value purchase characteristic, a derivative exposure increment characteristic, a derivative initial margin exposure increment characteristic, a pledge liquidation value inflow characteristic, an accrued interest accumulation characteristic, a pledge market value inflow characteristic, a pledge market value appreciation characteristic, a pledge market value depreciation characteristic, a derivative negative market value increment characteristic, a loan value fluctuation due to a foreign exchange fluctuation characteristic, a private equity subscription or accrued interest debit with a future value date characteristic, and an unlimited pledge difference characteristic.

[0025] According to an exemplary embodiment, the processor may be further configured to: receive feedback information from at least one user via a graphical user interface; and update the at least one margin call model based on the feedback information.

[0026] According to an exemplary embodiment, the feedback information may include at least one of a first information accuracy ranking, a second information accuracy ranking, and a characteristic accuracy ranking.

[0027] According to an exemplary embodiment, the processor may be further configured to: parse the plurality of data; identify personally identifiable information in the parsed plurality of data; and anonymize the plurality of data by removing the identified personally identifiable information. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In the following detailed description, the present disclosure is further described by way of non-limiting examples of preferred embodiments thereof, with reference to the several accompanying drawings in which like reference numerals represent like elements throughout the several views thereof.

[0029] Figure 1 An exemplary computer system is shown.

[0030] Figure 2 An exemplary diagram of a network environment is shown.

[0031] Figure 3 An exemplary system is shown for implementing a method for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified through the use of machine learning.

[0032] Figure 4 is a flow chart of an exemplary process for implementing a method for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified through the use of machine learning.

[0033] Figure 5 is an illustration of a machine learning process that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0034] Figure 6 is a graphical user interface illustrating variable importance of features identified in a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0035] Figure 7is a screenshot illustrating a graphical user interface that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0036] Figure 8 is a flow chart of an exemplary process for implementing a method for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified through the use of machine learning.

[0037] Figure 9 is a flow chart of an exemplary process for implementing a method for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified through the use of machine learning.

[0038] Figure 10 is a screenshot of a graphical user interface illustrating an unclassified ticket that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0039] Figure 11 is a screenshot of a graphical user interface illustrating an unclassified note dashboard that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0040] Figure 12 is a screenshot of a graphical user interface illustrating a classification confirmation that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0041] Figure 13 is a screenshot of a graphical user interface illustrating an unclassified ticket that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0042] Figure 14A is a screenshot of a graphical user interface illustrating artificial intelligence features that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0043] Figure 14B is a screenshot of a graphical user interface showing ticket details that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0044] Figure 15 is a screenshot of a graphical user interface illustrating an unclassified ticket panel that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0045] Figure 16 is a screenshot of a graphical user interface illustrating model accuracy that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0046] Figure 17 is a screenshot of a graphical user interface illustrating accuracy monitoring that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning, according to an exemplary embodiment.

[0047] Figure 18 is an exposition model illustrating feature interactions that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic root cause identification through the use of machine learning, according to an exemplary embodiment.

[0048] Figure 19A is a screenshot illustrating a user feedback form according to an exemplary embodiment, the graphical user interface may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of root causes through the use of machine learning.

[0049] Figure 19B is a screenshot illustrating an expanded graphical user interface displayed based on user input in a user feedback form that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process to enable automatic identification of a root cause through the use of machine learning, according to an exemplary embodiment. DETAILED DESCRIPTION

[0050] The embodiments and / or specific features or subcomponents of the present disclosure are intended to provide, through one or more of its various aspects, one or more advantages as specifically described above and pointed out below.

[0051] Examples may also be embodied as one or more non-transitory computer-readable media having stored thereon instructions for performing one or more aspects of the present technology as described and illustrated by the examples herein. The instructions in some examples include executable code that, when executed by one or more processors, causes the processors to perform the steps required to implement the methods of the examples of the present technology described and illustrated herein.

[0052] Figure 1 is an exemplary system for use in accordance with embodiments described herein. System 100 is generally shown and may include computer system 102, generally indicated.

[0053] The computer system 102 may include a set of instructions that can be executed to cause the computer system 102 to perform any one or more of the methods or computer-based functions disclosed herein, either alone or in combination with other described devices. The computer system 102 can operate as a standalone device or can be connected to other systems or peripherals. For example, the computer system 102 can include or be included in any one or more computers, servers, systems, communication networks, or cloud environments. Furthermore, the instructions can operate in such a cloud-based computing environment.

[0054] In a networked deployment, the computer system 102 can operate as a server or as a client user computer in a server-client user network environment, a client user computer in a cloud computing environment, or as a peer computer system in a peer (or distributed) network environment. The computer system 102 or parts thereof can be implemented as or incorporated into various devices, such as a personal computer, a tablet computer, a set-top box, a personal digital assistant, a mobile device, a handheld computer, a laptop computer, a desktop computer, a communication device, a wireless smart phone, a personal trusted device, a wearable device, a global positioning satellite (GPS) device, a web device, or any other machine capable of executing a set of instructions (sequentially or otherwise) that specify the actions to be taken by the machine. In addition, although a single computer system 102 is shown, additional embodiments may include any collection of systems or subsystems that execute instructions or perform functions individually or in combination. Throughout this disclosure, the term "system" should be deemed to include any collection of systems or subsystems that execute one or more sets of instructions to perform one or more computer functions individually or in combination.

[0055] like Figure 1As shown, the computer system 102 may include at least one processor 104. The processor 104 is tangible and non-transitory. As used herein, the term "non-transitory" should not be interpreted as a permanent characteristic of a state, but rather a characteristic of a state that will persist for a period of time. The term "non-transitory" specifically negates transient characteristics, such as characteristics of a specific carrier or signal or other forms that only exist temporarily at any place at any time. The processor 104 is an article of manufacture and / or machine component. The processor 104 is configured to execute software instructions to perform the functions described in various embodiments herein. The processor 104 may be a general-purpose processor or may be part of an application-specific integrated circuit (ASIC). The processor 104 may also be a microprocessor, a microcomputer, a processor chip, a controller, a microcontroller, a digital signal processor (DSP), a state machine, or a programmable logic device. The processor 104 may also be a logic circuit, including a programmable gate array (PGA) such as a field programmable gate array (FPGA), or another type of circuit including discrete gates and / or transistor logic. The processor 104 may be a central processing unit (CPU), a graphics processing unit (GPU), or both. Additionally, any processor described herein may include multiple processors, parallel processors, or both. Multiple processors may be included in or coupled to a single device or multiple devices.

[0056] Computer system 102 may also include computer memory 106. Computer memory 106 may include static memory, dynamic memory, or both. Memory, as described herein, is a tangible storage medium that can store data and executable instructions and is non-transitory during the time the instructions are stored therein. Again, as used herein, the term "non-transitory" is not to be construed as a permanent characteristic of a state, but rather a characteristic of a state that persists for a period of time. The term "non-transitory" specifically disclaims transient characteristics, such as those of a particular carrier wave or signal, or other forms that exist only temporarily at any one place at any one time. Memory is an article of manufacture and / or machine component. Memory, as described herein, is a computer-readable medium from which a computer can read data and executable instructions. Memory, as described herein, may be random access memory (RAM), read-only memory (ROM), flash memory, electrically programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disk, cache memory, removable disk, magnetic tape, compact disc read-only memory (CD-ROM), digital versatile disk (DVD), floppy disk, Blu-ray disc, or any other form of storage medium known in the art. The memory may be volatile or non-volatile, secure and / or encrypted, unsecure and / or unencrypted.Of course, the computer memory 106 may include any combination of memories or a single storage device.

[0057] Computer system 102 may also include a display 108, such as a liquid crystal display (LCD), an organic light emitting diode (OLED), a flat panel display, a solid state display, a cathode ray tube (CRT), a plasma display, or any other type of display, examples of which are known to those skilled in the art.

[0058] The computer system 102 may also include at least one input device 110, such as a keyboard, a touch-sensitive input screen or tablet, voice input, a mouse, a remote control device with a wireless keypad, a microphone coupled to a voice recognition engine, a camera such as a video camera or a still camera, a cursor control device, a global positioning system (GPS) device, an altimeter, a gyroscope, an accelerometer, a proximity sensor, or any combination thereof. Those skilled in the art will appreciate that various embodiments of the computer system 102 may include multiple input devices 110. Furthermore, those skilled in the art will appreciate that the exemplary input devices 110 listed above are not intended to be exhaustive, and that the computer system 102 may include any additional or alternative input devices 110.

[0059] The computer system 102 may also include a media reader 112 configured to read any one or more sets of instructions, such as software, from any memory described herein. When executed by a processor, the instructions may be used to perform one or more of the methods and processes described herein. In certain embodiments, the instructions may reside completely or at least partially within the memory 106, the media reader 112, and / or the processor 110 during execution by the computer system 102.

[0060] Furthermore, the computer system 102 may include any additional devices, components, parts, peripherals, hardware, software, or any combination thereof that are generally referred to as being included within or included in a computer system, such as, but not limited to, a network interface 114 and an output device 116. The output device 116 may be, but is not limited to, a speaker, an audio output, a video output, a remote control output, a printer, or any combination thereof.

[0061] Each component of the computer system 102 can be interconnected and communicate via a bus 118 or other communication link. Figure 1 As shown, the components can be interconnected and communicate via an internal bus. However, those skilled in the art will appreciate that any component can also be connected via an expansion bus. In addition, bus 118 can communicate via any standard or other specification that is generally known and understood, such as, but not limited to, Peripheral Component Interconnect (PCI), PCI Express (PCI Express), Parallel Advanced Technology Attachment (PATA), Serial Advanced Technology Attachment (STAT), etc.

[0062] The computer system 102 can communicate with one or more additional computer devices 120 via a network 122. The network 122 can be, but is not limited to, a local area network, a wide area network, the Internet, a telephone network, a short-range network, or any other network known and understood in the art. Short-range networks can include, for example, Bluetooth, Zigbee, infrared, near field communication, ultra-wideband, or any combination thereof. It will be appreciated by those skilled in the art that additional networks 122 known and understood can be used in addition or in lieu thereof, and that the exemplary networks 122 are not limiting or exhaustive. Furthermore, while the network 122 may be Figure 1 Although shown as a wireless network in FIG, those skilled in the art will appreciate that the network 122 may also be a wired network.

[0063] Additional computer equipment 120 in Figure 1 102 is shown as a personal computer. However, those skilled in the art will appreciate that in alternative embodiments of the present application, the computer device 120 may be a laptop computer, a tablet PC, a personal digital assistant, a mobile device, a handheld computer, a desktop computer, a communication device, a wireless phone, a personal trusted device, a web device, a server, or any other device capable of sequentially or otherwise executing a set of instructions specifying an action to be taken by the device. Of course, those skilled in the art will appreciate that the devices listed above are merely exemplary devices, and the device 120 may be any additional device or apparatus known and understood in the art without departing from the scope of the present application. For example, the computer device 120 may be the same as or similar to the computer system 102. Furthermore, those skilled in the art will similarly appreciate that the apparatus may be any combination of an apparatus and a device.

[0064] Of course, those skilled in the art will appreciate that the components of the computer system 102 listed above are merely exemplary, not exhaustive and / or inclusive. In addition, the examples of the components listed above are also meant to be exemplary, and similarly are not meant to be exhaustive and / or inclusive.

[0065] According to various embodiments of the present disclosure, the methods described herein may be implemented using a hardware computer system that executes a software program. Furthermore, in exemplary, non-limiting embodiments, implementations may include distributed processing, component / object distributed processing, and parallel processing. Virtual computer system processing may be constructed to implement one or more of the methods or functions described herein, and the processors described herein may be used to support virtual processing environments.

[0066] As described herein, various embodiments provide optimized methods and systems for identifying and capturing all relevant features involved in the margin call process so that the root cause can be automatically identified through the use of machine learning.

[0067] refer to Figure 2, a schematic diagram of an exemplary network environment 200 is shown for implementing a method for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified through the use of machine learning. In an exemplary embodiment, the method can be executed on any networked computer platform, such as a personal computer (PC).

[0068] A method for identifying and capturing all relevant features involved in the margin call process so that the root cause can be automatically identified by using machine learning can be implemented by the margin call data collection and analysis (MCDCA) device 202. The MCDCA device 202 can be used with respect to Figure 1 The MCDCA device 202 may be the same as or similar to the computer system 102 described above. The MCDCA device 202 may store one or more applications that may include executable instructions that, when executed by the MCDCA device 202, cause the MCDCA device 202 to perform actions such as sending, receiving, or otherwise processing network messages, and to perform other actions described and illustrated below with reference to the figures. The applications may be implemented as modules or components of other applications. Furthermore, the applications may be implemented as operating system extensions, modules, plug-ins, and the like.

[0069] Furthermore, the application can operate in a cloud-based computing environment. The application can be executed within or as a virtual machine or virtual server that can be managed in the cloud-based computing environment. Moreover, the application and even the MCDCA device 202 itself can be located in a virtual server running in the cloud-based computing environment, rather than being tied to one or more specific physical network computing devices. In addition, the application can run in one or more virtual machines (VMs) executing on the MCDCA device 202. In addition, in one or more embodiments of the present technology, the virtual machines running on the MCDCA device 202 can be managed or supervised by a hypervisor.

[0070] exist Figure 2 In a network environment 200, an MCDCA device 202 is coupled to a plurality of server devices 204(1)-204(n) hosting a plurality of databases 206(1)-206(n) via a communication network 210, and is also coupled to a plurality of client devices 208(1)-208(n). The communication interface (e.g., Figure 1The network interface 114 of the computer system 102 is operatively coupled and communicates between the MCDCA device 202, the server devices 204(1)-204(n), and / or the client devices 208(1)-208(n), all of which are coupled together via the communication network 210, although other types and / or numbers of communication networks or systems having other types and / or numbers of connections and / or configurations to other devices and / or elements may also be used.

[0071] The communication network 210 can communicate with Figure 1 The network 122 is the same or similar as that described, but the MCDCA device 202, the server devices 204(1)-204(n), and / or the client devices 208(1)-208(n) can be coupled together via other topologies. Additionally, the network environment 200 can include other network devices, such as one or more routers and / or switches, for example, which are well known in the art and therefore will not be described herein. The present technology provides numerous advantages, including methods, non-transitory computer-readable media, and MCDCA devices that effectively implement a method for identifying and capturing all relevant features involved in a margin call process so that a root cause can be automatically identified using machine learning.

[0072] By way of example only, the communication network 210 may comprise a local area network (LAN) or a wide area network (WAN) and may utilize TCP / IP and industry standard protocols over Ethernet, although other types and / or numbers of protocols and / or communication networks may also be used. The communication network 210 in this example may utilize any suitable interface mechanism and network communication technology, including, for example, any suitable form of remote service (e.g., voice, modem, etc.), a public switched telephone network (PSTN), an Ethernet-based packet data network (PDN), combinations thereof, and the like.

[0073] For example, the MCDCA device 202 can be a standalone device or integrated with one or more other devices or apparatuses (e.g., one or more of the server devices 204(1)-204(n)). In one specific example, the MCDCA device 202 can include or be hosted by one of the server devices 204(1)-204(n), and other arrangements are possible. Furthermore, for example, one or more of the MCDCA devices 202 can be in the same or different communication networks, including one or more public, private, or cloud networks.

[0074] The plurality of server devices 204(1)-204(n) may be associated with Figure 1The computer systems 102 or computer devices 120 described herein are the same as or similar to each other, including any features or combinations of features described therewith. For example, any of the server devices 204(1)-204(n) may include, among other features, one or more processors, memory, and communication interfaces coupled together via a bus or other communication link, although other numbers and / or types of network devices may also be used. For example, in this example, the server devices 204(1)-204(n) may process requests received from the MCDCA device 202 via the communication network 210 according to a protocol based on HTTP and / or JavaScript Object Notation (JSON), although other protocols may also be used.

[0075] The server devices 204(1)-204(n) may be hardware or software, or may represent a system having multiple servers in a pool, which may include an internal or external network. The server devices 204(1)-204(n) host databases 206(1)-206(n) configured to store data related to margin calls, data patterns, characteristics, and margin call models.

[0076] Although server devices 204(1)-204(n) are shown as a single device, one or more actions of each of server devices 204(1)-204(n) may be distributed across one or more different network computing devices that together comprise one or more of server devices 204(1)-204(n). Furthermore, server devices 204(1)-204(n) are not limited to a particular configuration. Thus, server devices 204(1)-204(n) may comprise multiple network computing devices that operate using a master / slave approach, whereby one of the network computing devices of server devices 204(1)-204(n) operates to manage and / or otherwise coordinate the operations of the other network computing devices.

[0077] For example, server devices 204(1)-204(n) may operate as multiple network computing devices in a cluster architecture, a peer-to-peer architecture, within a virtual machine, or within a cloud architecture. Thus, the technology disclosed herein should not be construed as limited to a single environment, and other configurations and architectures are also contemplated.

[0078] Multiple client devices 208(1)-208(n) may also communicate with Figure 1The computer systems 102 or computer devices 120 described herein are the same or similar, including any features or combinations of features described therewith. For example, the client devices 208(1)-208(n) in this example may include any type of computing device that can interact with the MCDCA device 202 via the communication network 210. Thus, the client devices 208(1)-208(n) may be, for example, mobile computing devices, desktop computing devices, laptop computing devices, tablet computing devices, virtual machines (including cloud-based computers), etc. that host chat, email, or voice-to-text applications. In an exemplary embodiment, at least one client device 208 is a wireless mobile communication device, i.e., a smartphone.

[0079] The client devices 208(1)-208(n) can run an interface application, such as a standard web browser or a stand-alone client application, which can provide an interface to communicate with the MCDCA device 202 via the communication network 210 to transmit user requests and information. Among other features, the client devices 208(1)-208(n) can also include a display device (e.g., a display screen or touch screen) and / or an input device (e.g., a keyboard).

[0080] Although an exemplary network environment 200 having an MCDCA device 202, server devices 204(1)-204(n), client devices 208(1)-208(n), and a communication network 210 is described and illustrated herein, other types and / or numbers of systems, devices, components, and / or elements in other topologies may be used. It should be understood that the example systems described herein are for exemplary purposes, as many variations in the specific hardware and software used to implement the examples are possible, as will be understood by those skilled in the relevant art.

[0081] One or more of the devices depicted in the network environment 200 (e.g., the MCDCA device 202, the server devices 204(1)-204(n), or the client devices 208(1)-208(n)) may be configured to operate as virtual instances on the same physical machine. In other words, one or more of the MCDCA device 202, the server devices 204(1)-204(n), or the client devices 208(1)-208(n) may operate on the same physical device rather than as separate devices communicating over the communication network 210. Additionally, there may be more than one physical instance of the MCDCA device 202. Figure 2 More or fewer MCDCA devices 202, server devices 204(1)-204(n), or client devices 208(1)-208(n) are shown.

[0082] In addition, in any example, two or more computing systems or devices can replace any one of the systems or devices. Therefore, the principles and advantages of distributed processing, such as redundancy and replication, can also be implemented as needed to increase the robustness and performance of the devices and systems of the examples. The examples can also be implemented on computer systems that extend across any suitable network using any suitable interface mechanism and business technology, including, by way of example only, any suitable form of remote business (e.g., voice and modem), wireless business networks, cellular business networks, packet data networks (PDNs), the Internet, intranets, and combinations thereof.

[0083] MCDCA device 202 Figure 3 1 and 2 as including a margin call data collection and analysis (MCDCA) module 302, although it may also include, for example, other rules, policies, modules, databases, or applications. As described below, the MCDCA module 302 is configured to implement a method for identifying and capturing all relevant features involved in the margin call process so that the root cause can be automatically identified through the use of machine learning.

[0084] Implements all relevant features for identifying and capturing the margin call process so that it can be utilized Figure 2 An exemplary process 300 of a mechanism for automatically identifying root causes in a network environment is shown as Figure 3 . Specifically, a first client device 208(1) and a second client device 208(2) are shown as communicating with the MCDCA device 202. In this regard, the first client device 208(1) and the second client device 208(2) may be "clients" of the MCDCA device 202 and described as such herein. However, it should be known and understood that the first client device 208(1) and / or the second client device 208(2) are not necessarily "clients" of the MCDCA device 202 or any entity described herein in association therewith. Any additional or alternative relationship may exist between either or both of the first client device 208(1) and the second client device 208(2) and the MCDCA device 202, or no relationship may exist.

[0085] In addition, the MCDCA device 202 is shown as being able to access a margin call data warehouse 206(1) and a data pattern, characteristic, and margin call model database 206(2). The MCDCA module 302 can be configured to access these databases to implement a method for identifying and capturing all relevant features involved in the margin call process so that the root cause can be automatically identified through the use of machine learning.

[0086] The first client device 208(1) can be, for example, a smartphone. Of course, the first client device 208(1) can be any of the additional devices described herein. The second client device 208(2) can be, for example, a personal computer (PC). Of course, the second client device 208(2) can also be any of the additional devices described herein.

[0087] The process may be performed via the communication network 210, which may include multiple networks as described above. For example, in an exemplary embodiment, either or both of the first client device 208(1) and the second client device 208(2) may communicate with the MCDCA device 202 via broadband or cellular communications. Of course, these embodiments are merely exemplary and are not intended to be limiting or exhaustive.

[0088] At the outset, the MCDCA module 302 performs a process for identifying and capturing all relevant features involved in the margin call process so that the root cause can be automatically identified by using machine learning. Figure 4 An exemplary process for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified through the use of machine learning is generally indicated at flowchart 400 in FIG.

[0089] exist Figure 4 In process 400, at step S402, data related to multiple securities portfolios may be obtained. In an exemplary embodiment, the data may include information related to the securities portfolios of multiple investors. In another exemplary embodiment, the multiple investors may include unrelated entities and / or related entities (i.e., affiliated entities), and the multiple investors may also include individual investors and / or institutional investors (i.e., collateral funds).

[0090] In another exemplary embodiment, the data may be parsed and anonymized to remove personally identifying information corresponding to multiple investors. In another exemplary embodiment, the securities portfolio may include a group of any financial assets, such as stocks, bonds, commodities, currencies, cash equivalents, mutual funds, exchange-traded funds, and closed-end funds. Other types of non-tradable instruments (e.g., time deposits, pledges received, and guarantees) may also be part of the portfolio. In addition, the securities portfolio may also include non-publicly tradable securities, such as real estate, art, and private investments.

[0091] In another exemplary embodiment, the data may be obtained from any publicly accessible source (i.e., stock market data from an exchange) and privately held source (i.e., account data from a bank). In another exemplary embodiment, the data may include any information such as transaction type data, stock price data, asset quantity data, and bond yield data associated with each financial asset within a securities portfolio.

[0092] In another exemplary embodiment, the transaction type data may reference any trade order, such as a market order, a limit order, a stop order (i.e., a stop-loss order), and a buy stop order. In another exemplary embodiment, the transaction type data may reference any demand placed on an investor by a financial institution, such as a federal margin call, a maintenance margin call, and an exchange margin call associated with the investor's securities portfolio. As will be understood by one of ordinary skill in the art, the transaction type data may reference any trade and corresponding transaction information associated with each security within the securities portfolio.

[0093] At step S404, the data may be used to automatically determine data patterns associated with triggering a margin call and corresponding characteristics that contribute to triggering the margin call. In an exemplary embodiment, the margin call may include generalized margin call scenarios generated by analysis of the obtained data. In another exemplary embodiment, the margin call may include predetermined scenarios established by the organization as standard triggers for a margin call.

[0094] In another exemplary embodiment, analyzing the data to automatically determine data patterns and characteristics may include comparing a pre-developed analytical model to the data. In another exemplary embodiment, analyzing the data to automatically determine data patterns and characteristics may include applying machine learning analysis techniques to the data. In another exemplary embodiment, the machine learning analysis techniques may include supervised learning algorithms, such as regression analysis, decision tree analysis, random forest analysis, k-nearest neighbor analysis, logistic regression analysis, etc. In another exemplary embodiment, the machine learning analysis techniques may include unsupervised learning algorithms, such as prior analysis, K-means cluster analysis, etc. In another exemplary embodiment, the machine learning analysis techniques may include reinforcement learning algorithms, such as Markov decision processes, etc.

[0095] In another exemplary embodiment, data patterns can be identified using historical data pattern recognition techniques (i.e., comparison to known data patterns in historical data) and statistical pattern recognition techniques (i.e., machine learning and artificial intelligence, involving the identification of patterns and regularities in data). For example, in the historical data pattern recognition technique, first data can be received and compared to pre-developed data patterns for securities portfolios known to have triggered margin calls in the past. Alternatively, for example, in the statistical pattern recognition technique, data from past margin calls can be analyzed by a computing system using an algorithm that enables the computing system to identify data patterns. The identified data patterns can then be applied to the first securities portfolio to predict the probability of a margin call for the first securities portfolio.

[0096] In one exemplary embodiment, the aforementioned machine learning analysis techniques can be used to automatically determine characteristics based on data patterns. In another exemplary embodiment, the characteristics can include a contribution factor to a margin call and / or a root cause of a margin call based on the magnitude of the impact. For example, a decline in the value of Bond A due to a decline in the price of Stock B may have led to an already active margin call. In another exemplary embodiment, the magnitude of each characteristic's impact can be expressed as a percentage of the actual contribution to the margin call and a percentage of the probability of contributing to the margin call.

[0097] In another exemplary embodiment, the characteristics may include at least one of the following: a securities appreciation characteristic, a securities depreciation characteristic, an asset market value difference characteristic, a securities market value difference characteristic, a securities market value buy characteristic, an exposure difference characteristic, a cash market value difference characteristic, a collateral difference characteristic, a cash market value sell characteristic, a haircut difference characteristic, a securities market value sell characteristic, a securities liquidation value buy characteristic, a cash market value appreciation characteristic, a cash market value depreciation characteristic, a pledge market value outflow characteristic, a pledge market value difference characteristic, a cash market value buy characteristic, a derivative exposure increment characteristic, a derivative initial margin exposure increment characteristic, a pledge liquidation value inflow characteristic, an accrued interest accumulation characteristic, a pledge market value inflow characteristic, a pledge market value appreciation characteristic, a pledge market value depreciation characteristic, a derivative negative market value increment characteristic, a loan value fluctuation due to a foreign exchange fluctuation characteristic, a private equity subscription or accrued interest debit with a future value date characteristic, and an unlimited pledge difference characteristic.

[0098] At step S406, a margin call model based on the data patterns and characteristics can be automatically generated. In an exemplary embodiment, the margin call model can include a mathematical model, such as a financial model designed to represent real-world financial situations. In another exemplary embodiment, the margin call model can include a mathematical model representing the likelihood of a margin call for any given portfolio of securities. In another exemplary embodiment, a margin call model can be generated for any of the aforementioned transaction types. In another exemplary embodiment, a margin call model can be automatically generated to represent specific analyses, such as projected risk analysis and projected outcome analysis. In another exemplary embodiment, any combination of data patterns and characteristics can be used to generate a margin call model.

[0099] In another exemplary embodiment, the margin call model may include a trained model, such as a machine learning model that is generated for further training on additional data. Once the trained model has been fully trained, it can be deployed to various connected systems to be utilized. In another exemplary embodiment, the trained model can be fully trained when a model evaluation method, such as a holdout method, a K-fold cross-validation method, and a bootstrap method, determines that the model's true positive rate, true negative rate, false positive rate, and false negative rate correspond to predetermined ranges. In another exemplary embodiment, the trained model can be operational (i.e., actively utilized by the organization) while continuing to be trained using new data. In another exemplary embodiment, the margin call model can be generated using at least one of artificial neural network technology, decision tree technology, support vector machine technology, Bayesian network technology, and genetic algorithm technology.

[0100] In another exemplary embodiment, a margin call model may be trained based on at least one attribute. The attribute may correspond to data that is both internally available and externally retrieved. The attribute may include at least one of: ticket value, platform value, date value, today's available collateral value, yesterday's available collateral value, today's total loan value, yesterday's total loan value, today's total haircut value, yesterday's total haircut value, today's accrued interest value, yesterday's accrued interest value, today's derivative initial margin (IM) exposure value, today's derivative market-to-market (MTM) exposure value, yesterday's derivative IM exposure value, yesterday's derivative MTM exposure value, today's total exposure value, yesterday's total exposure value, today's security liquidity value (LV) value, yesterday's security LV value, today's unlimited cross collateral value, yesterday's unlimited cross collateral value. Cross pledge value, today's limited cross pledge LV value, yesterday's limited cross pledge LV value, mortgage property value LV value, mortgage risk exposure value, market value (MV) purchase securities value, loan value purchase securities value, loan value sale securities value, loan value purchase asset value, margin call threshold, cash debit future value, loan cash debit future value, collateral increment generated by LV value, collateral increment generated by price value, collateral increment generated by foreign exchange (FX) value, cash increment generated by FX value, cash loan value change value, AI reasons for margin call value and reasons for margin call value.

[0101] At step S408, the margin call model may be used to evaluate the individual portfolio of interest to determine a first set of information related to whether a margin call should be triggered for the portfolio of interest. In an exemplary embodiment, a user may identify a first portfolio of interest to the MCDCA module 302 via input on a graphical user interface. In another exemplary embodiment, the first set of information related to whether a margin call should be triggered may include information predicting a margin call as a probability metric and information on the actual requirements for triggering the first margin call for the first portfolio of interest. In another exemplary embodiment, the evaluation may further include determining a first data pattern, a classification of relevant parameters, and a first characteristic of the first portfolio of interest. The first data pattern and the first characteristic may correspond to the first portfolio of interest and may be related to triggering the first margin call.

[0102] Then, in step S410, the MCDCA module 302 may display the first set of information on a graphical user interface via a display. In another exemplary embodiment, the first information may include the first data pattern, the first characteristic, and any other data related to the first securities portfolio. In another exemplary embodiment, the first information may be displayed on any connected computing device, such as a computer, laptop, smartphone, or smartwatch. The displayed information may include, for example, information that an analyst may use to decide whether to execute a margin call and / or information providing a specific recommendation regarding whether to execute a margin call. In another exemplary embodiment, the displayed information may also include an alert, such as an alert notification of an impending margin call.

[0103] In another exemplary embodiment, the first portfolio can be received in a batch of portfolios (i.e., a file containing many portfolios) or as individual entries for each portfolio. In another exemplary embodiment, the first portfolio can be stored on a connected system and retrieved upon user request (i.e., a user can enter the first customer number into a graphical user interface to analyze the first portfolio).

[0104] In another exemplary embodiment, the information related to triggering a margin call may include a probability of a margin call for a given portfolio of securities. In another exemplary embodiment, the information may include at least one characteristic related to and contributing to a likely margin call for a given portfolio of securities.

[0105] In another exemplary embodiment, a user may select a specific data pattern and corresponding characteristic from a graphical user interface. The selected data pattern and corresponding characteristic may be associated with triggering a margin call for the currently analyzed securities portfolio. After the user selects the data pattern and corresponding characteristic, the graphical user interface may display targeted margin call information based on the selection.

[0106] In another exemplary embodiment, the graphical user interface may display margin call information in various formats (e.g., in a panel). The panel may display additional information related to the securities portfolio, such as, for example, monitoring information related to the securities portfolio and classification information related to the determined characteristics, as well as the margin call information.

[0107] In another exemplary embodiment, the graphical user interface may include graphical elements, such as text boxes, selection planes, and marker features, to receive feedback from the user. The feedback information may include data such as a first information accuracy ranking, a second information accuracy ranking, and a feature accuracy ranking. The feedback information may also include the accuracy of the predicted information corresponding to a specific data pattern. In another exemplary embodiment, the feedback information may be used to update the margin call model to increase the margin call model's prediction accuracy and analysis precision. As will be appreciated by those skilled in the art, the margin call model may be updated based on machine learning techniques, for example, by training the margin call model based on the feedback information.

[0108] Figure 5 is an illustration 500 of a machine learning process that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified through the use of machine learning, according to an exemplary embodiment. Figure 5 The interconnectedness of the various steps in the machine learning process is shown. In an exemplary embodiment, the steps may include "collect data" 502, "clean / transform" 504, "explore / visualize" 506, "model" 508, "evaluate" 510, and "deploy" 512, where the different steps may be performed periodically to improve accuracy as more data from previous margin calls becomes available. New data is retrofitted to the system to learn from previous margin calls.

[0109] like Figure 5 As shown, the machine learning process can begin at the collect data step 502. The collected data can then be cleaned to remove unnecessary information and converted into a processable data format at the clean / transform step 504. Next, the cleaned / transformed data can move to the explore / visualize step 506, where available information, such as data patterns and characteristics, can be determined. At this point in the process, if the collected data may not be sufficient to determine available information, the process can move back to the collect data step 502 to obtain additional data. However, if available information has been determined, a model can be generated at the model step 508.

[0110] Once the model is generated, the process can return to the collect data step 502 to collect more data, return to the cleansing / transformation step 504 to further refine the data already collected, or move forward to the evaluation step 510. At the evaluation step 510, the developed model can be tested to determine the model's accuracy. When the model is sufficiently accurate based on predetermined criteria, the process can continue to the deployment step 512. However, when additional refinement is required, the process can move back to the collect data step 502 or the cleansing / transformation step 504. At the deployment step 512, the model can be used to analyze and predict outcomes of user input. After deployment, additional collected data, additional available information, and additional models can be used to continuously improve the machine learning process.

[0111] Figure 6 is a graphical user interface 600 illustrating variable importance of features identified in a method for identifying and capturing all relevant features involved in a margin call process such that root causes can be automatically identified through the use of machine learning, according to an exemplary embodiment. Figure 6 The graphical user interface shown in FIG has a title "Variable Importance". The graphical user interface may also include a bar graph with an x-axis representing the percentage of importance in numbers and a y-axis representing the contributing characteristics in text. In one exemplary embodiment, the contributing characteristics may be arranged in ascending or descending order based on relative importance, for example, the most important characteristics may be arranged at the top of the bar graph, or vice versa, the most important characteristics may be arranged at the bottom of the bar graph. In another exemplary embodiment, the relative importance may include the percentage of each characteristic's contribution to a transaction type (e.g., a margin call).

[0112] In another exemplary embodiment, the graph shown may include any visual representation of the data, such as a line graph, a pie chart, a mosaic chart, a spider chart, a stacked bar chart, etc. In another exemplary embodiment, the contribution characteristics may include at least one of the following: a security appreciation characteristic, a security depreciation characteristic, an asset market value difference characteristic, a security market value difference characteristic, a security market value buy characteristic, an exposure difference characteristic, a cash market value difference characteristic, a collateral difference characteristic, a cash market value sell characteristic, a haircut difference characteristic, a security market value sell characteristic, a security liquidation value buy characteristic, a cash market value appreciation characteristic, a cash market value depreciation characteristic, a pledge market value outflow characteristic, a pledge market value difference characteristic, a cash market value buy characteristic, a derivative exposure increment characteristic, a derivative initial margin exposure increment characteristic, a pledge liquidation value inflow characteristic, an accrued interest accumulation characteristic, a pledge market value inflow characteristic, a pledge market value appreciation characteristic, a pledge market value depreciation characteristic, a derivative negative market value increment characteristic, a loan value fluctuation due to a foreign exchange fluctuation characteristic, a private equity subscription or accrued interest debit with a future value date characteristic, and an unlimited pledge difference characteristic.

[0113] Figure 7 is a screenshot illustrating a graphical user interface 700, according to an exemplary embodiment, that can be used to implement a method for identifying and capturing all relevant features involved in the margin call process so that root causes can be automatically identified using machine learning. Graphical user interface 700 may include an exemplary graphical user interface that facilitates user interaction with various automatically generated margin call models. As shown, graphical user interface 700 may include a sortable table of multiple investors and their corresponding securities portfolios. In an exemplary embodiment, the table provides a summary view of the processed securities portfolios. In graphical user interface 700, the table can be sorted by categories titled "Note ID," "Client Name," "Aging," "Age Buckets," "Client ID," "Capital Advisor," "Credit Execution," "Margin Call Amount USD," "Margin Call Reason," and "Predicted Margin Call Reason."

[0114] In an exemplary embodiment, information in graphical user interface 700 can be displayed using any combination of letters, words, and numbers. In another exemplary embodiment, the Age Bucket category can display information as a range of days. In another exemplary embodiment, the Margin Call Amount category can display amounts in any user-selectable, legally recognized legal repayment. In another exemplary embodiment, the Margin Call Reason category can display characteristics known to have caused margin calls in the collected data. In another exemplary embodiment, the Predicted Margin Call Reason category can display characteristics predicted to cause future margin calls. In another exemplary embodiment, a comparison of the Predicted Margin Call Reason category and the Margin Call Reason category for a particular securities portfolio can be used to validate margin call models. In another exemplary embodiment, the predicted reasons for the data can be sent to graphical user interface 700 via an application programming interface (API) (e.g., a Representational State Transfer (REST) ​​API) by data management software (e.g., SPLUNK).

[0115] Figure 8 800 is a flowchart of an exemplary process for implementing a method for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified through the use of machine learning. As shown, flowchart 800 may include user interfaces operable on various user terminals labeled "System 1," "System 2," "System 3," and "System n."

[0116] The exemplary process in flowchart 800 begins at step S802, where features related to shared characteristics of margin calls are collected. The collected features may include, for example, accrued interest, FX rate fluctuations, and market price fluctuations. Then, at step S804, the collected features may be used to train a margin call model using artificial intelligence (AI) or machine learning (ML) techniques. At step S806, the trained margin call model may be used to predict the cause of a margin call. The predicted cause of a margin call may then be validated using known outcomes at step S808 to further refine the developed margin call model. Finally, at step S810, the model may be retrofitted by repeating the steps using additional collected features to further train the developed margin call model.

[0117] Figure 9 Flowchart 900 is a flowchart illustrating an exemplary process for implementing a method for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified using machine learning. As shown, flowchart 900 begins with a business problem that can be solved using machine learning techniques. A machine learning problem framing method can be used to frame the business problem to identify possible technical solutions related to machine learning techniques that can be applied to solve the business problem. Once the business problem has been formulated, implementation of the machine learning techniques can begin by collecting data related to the business problem. The collected data can be integrated with other relevant data, cleansed, and prepared for data visualization and analysis. Feature engineering of the analyzed and visualized data uses domain knowledge of the data to create features related to the attributes or properties of the business problem.

[0118] The developed model can then be trained and parameters adjusted using the engineered features. The trained model can then be evaluated using known results to determine whether the business objectives associated with the solution to the business problem are met. Once the business objectives are determined to have been met, the developed model can be deployed to predict and solve the business problem. After deployment, the model can be maintained using monitoring and debugging techniques that can continuously retrain the model using newly collected data.

[0119] Figure 10 is a screenshot 1000 of a graphical user interface illustrating an unclassified note according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified using machine learning. Figure 10As shown, unsorted notes are displayed in a margin call tool consistent with embodiments of the present application. The margin call tool may include a table containing a plurality of unsorted notes and a counter titled "Unreviewed" that counts the number of unsorted notes.

[0120] Figure 11 is a screenshot 1100 of a graphical user interface illustrating an unclassified note panel according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified using machine learning. Figure 11 As shown, the Unclassified Notes panel provides information for multiple notes. In an exemplary embodiment, the Unclassified Notes panel shows notes that have not been classified by artificial intelligence and may also include historical margin calls that have been previously classified.

[0121] Figure 12 is a screenshot 1200 of a graphical user interface illustrating a classification confirmation according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified using machine learning. Figure 12 As shown, a confirmation email can be received within a specified amount of time to confirm the successful classification of the selected data. The confirmation email can include information corresponding to the selected classified data. As will be understood by those skilled in the art, the confirmation email can be received by any email client, such as proprietary email clients and email clients developed by third parties.

[0122] Figure 13 is a screenshot 1300 of a graphical user interface illustrating a classified ticket according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified using machine learning. Figure 13 As shown, classified notes are displayed in a margin call tool consistent with an embodiment of the present application. Notes successfully classified by artificial intelligence are displayed on the margin call tool.

[0123] Figure 14A is a screenshot 1400A of a graphical user interface illustrating an artificial intelligence feature according to an exemplary embodiment that may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified using machine learning. Figure 14AAs shown, the classified instruments are displayed in a margin call tool consistent with an embodiment of the present application. The margin call tool can also provide available instrument details including artificial intelligence classification features and attributes.

[0124] Figure 14B is a screenshot 1400B of a graphical user interface illustrating ticket details according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified using machine learning. Figure 14B As shown, categorized tickets are displayed in a margin call tool consistent with an embodiment of the present application. The margin call tool may also provide available ticket details for a particular ticket, such as a ticket identifier, a reason for the margin call ticket identified, and relevant attribute values ​​used to identify the margin call reason.

[0125] Figure 15 is a screenshot 1500 of a graphical user interface illustrating a classified ticket panel according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified using machine learning. Figure 15 As shown, a panel consistent with embodiments herein displays classified tickets. The panel can display the AI-classified tickets along with data corresponding to the analysis. This data can include the number of valid margin calls, the number of invalid margin calls, and the total number of margin calls. The panel can also present data in graphical elements, such as a bar chart showing the reasons for margin calls in each region.

[0126] Figure 16 is a screenshot 1600 of a graphical user interface illustrating model accuracy according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified using machine learning. Figure 16 As shown, model analysis data is displayed in a panel consistent with embodiments herein. The panel can show the accuracy of the current model's artificial intelligence analysis for a given reason for a margin call (e.g., accrued interest reason). The model analysis data can include the number of valid margin calls, the number of invalid margin calls, and the total number of margin calls. The panel can also present the data in a graphical element, such as a bar chart of the reasons for margin calls in each region.

[0127] Figure 17is a screenshot 1700 of a graphical user interface illustrating accuracy monitoring according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause may be automatically identified using machine learning. Figure 17 As shown, historical model data is displayed in a panel consistent with embodiments herein. The panel can show a graphical representation of the monitoring of the accuracy improvement over time as the model is trained with more data. The graphical representation can include a graph, such as a line graph and a bar graph, that tracks the increase in accuracy of the model as a function of time.

[0128] Figure 18 FIG1 is an illustrative model 1800 illustrating feature interactions according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified using machine learning. Figure 18 As shown, the interactions between variables and characteristics can be highlighted in a graphical form. As will be understood by one of ordinary skill in the art, variables and characteristics can be presented on the X-axis or Y-axis of a graph such that intersecting squares indicate the interactions between a particular variable and a particular characteristic.

[0129] Figure 19A 1900A is a screenshot of a graphical user interface illustrating a user feedback form according to an exemplary embodiment, which may be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that a root cause can be automatically identified using machine learning. Screenshot 1900A provides graphical user elements, such as, for example, a menu or other utility window presented to a user.

[0130] like Figure 19A As shown, the graphical user element can include a user feedback form related to the AI ​​classification results. In an exemplary embodiment, data collected from the user feedback form can be used to further train the margin call model to improve classification accuracy using techniques consistent with embodiments herein. The user feedback form can include information corresponding to a specific ticket and a user-selectable graphical element for receiving user input.

[0131] Figure 19B 1900B is a screenshot illustrating an expanded graphical user interface displayed based on user input in the form of user feedback according to an exemplary embodiment, which can be used to implement a method for identifying and capturing all relevant features involved in a margin call process so that the root cause can be automatically identified through the use of machine learning. Screenshot 1900B provides expanded graphical user elements, such as, for example, menus or other utility windows presented to the user. Figure 19A The graphical user element in the example embodiment receives user input to present an extended graphical user element to the user.

[0132] like Figure 19B As shown, a graphical user element can include an expanded user feedback form related to the AI ​​classification results. In an exemplary embodiment, data collected from the expanded user feedback form can be used to further train the margin call model to improve classification accuracy using techniques consistent with embodiments herein. The expanded user feedback form can display information based on user selections made in previous graphical user elements. The expanded user feedback form can include additional information corresponding to a specific ticket and additional user-selectable graphical elements for receiving user input.

[0133] Thus, utilizing this technique, an optimized process is provided for identifying and capturing all relevant features involved in the margin call process so that the root cause can be automatically identified through the use of machine learning.

[0134] While the present invention has been described with reference to several exemplary embodiments, it is to be understood that the words which have been used are words of description and illustration, rather than of limitation. Changes may be made, as presently described and amended, within the scope of the appended claims without departing from the scope and spirit of the various aspects of the present disclosure. While the present invention has been described with reference to particular devices, materials, and embodiments, the invention is not intended to be limited to the details disclosed; rather, the invention extends to all functionally equivalent structures, methods, and uses, such as are within the scope of the appended claims.

[0135] For example, although a computer-readable medium may be described as a single medium, the term "computer-readable medium" includes a single medium or multiple media, such as a centralized or distributed database, and / or associated caches and servers that store one or more sets of instructions. The term "computer-readable medium" shall also include any medium that can store, encode, or carry a set of instructions to be executed by a processor or cause a computer system to perform any one or more embodiments disclosed herein.

[0136] The computer-readable medium may include non-transitory computer-readable media and / or include transient computer-readable media. In a specific non-limiting exemplary embodiment, the computer-readable medium may include a solid-state memory, such as a memory card or other package that houses one or more non-volatile read-only memories. In addition, the computer-readable medium may be a random access memory or other volatile rewritable memory. In addition, the computer-readable medium may include magneto-optical or optical media, such as a disk or tape or other storage device to capture carrier signals, such as signals transmitted over a transmission medium. Therefore, the present disclosure is considered to include any computer-readable medium or other equivalents and successor media in which data or instructions can be stored.

[0137] Although the present application describes specific embodiments that can be implemented as computer programs or code segments in a computer-readable medium, it should be understood that dedicated hardware implementations such as application-specific integrated circuits, programmable logic arrays, and other hardware devices can be constructed to implement one or more embodiments described herein. Applications that may include the various embodiments set forth herein may broadly include various electronic and computer systems. Therefore, the present application may encompass software, firmware, and hardware implementations or combinations thereof. Nothing in this application should be interpreted as being implemented or achievable solely with software rather than hardware.

[0138] Although this specification describes components and functions that can be implemented in specific embodiments with reference to specific standards and protocols, the present disclosure is not limited to such standards and protocols. Such standards are periodically superseded by faster or more efficient equivalents having substantially the same functions. Therefore, replacement standards and protocols having the same or similar functions are considered equivalents thereof.

[0139] The description of the embodiments described herein is intended to provide a general understanding of the various embodiments. These illustrations are not intended to be used as a complete description of all elements and features of the apparatus and systems utilizing the structures or methods described herein. After reading this disclosure, many other embodiments will be apparent to those skilled in the art. Other embodiments can be utilized and derived from this disclosure so that structural and logical replacements and changes can be made without departing from the scope of this disclosure. In addition, the illustrations are merely representative and may not be drawn to scale. Certain proportions within the illustrations may be exaggerated, while other proportions may be minimized. Therefore, this disclosure and the accompanying drawings should be considered illustrative and not restrictive.

[0140] One or more embodiments of the present disclosure may be referred to herein individually and / or collectively by the term "invention", which is merely for convenience and is not intended to voluntarily limit the scope of this application to any particular invention or inventive concept. In addition, although specific embodiments have been shown and described herein, it should be understood that any subsequent arrangements designed to achieve the same or similar purposes may replace the specific embodiments shown. The present disclosure is intended to cover any and all subsequent modifications or variations of the various embodiments. Combinations of the above-described embodiments and other embodiments not specifically described herein will be apparent to those skilled in the art upon reading this specification.

[0141] The Abstract of the Disclosure is submitted with the understanding that it will not be used to interpret or limit the scope or meaning of the claims. Additionally, in the foregoing Detailed Description, various features may be grouped together or described in a single embodiment for the purpose of simplifying the invention. This disclosure should not be interpreted as reflecting an intention that the claimed embodiments require more features than those expressly recited in each claim. Rather, as reflected in the appended claims, inventive subject matter may be directed to fewer than all features of any of the disclosed embodiments. Accordingly, the following claims are incorporated into the Detailed Description, with each claim independently defining separately claimed subject matter.

[0142] The above-disclosed subject matter is to be considered illustrative and not restrictive, and the appended claims are intended to cover all such modifications, enhancements, and other embodiments that fall within the true spirit and scope of the disclosure. Thus, to the maximum extent allowed by law, the scope of the present disclosure is to be determined by the broadest permissible interpretation of the appended claims and their equivalents, and shall not be restricted or limited by the foregoing detailed description.

Claims

1. A method for providing information related to triggering a margin call for a first portfolio of securities, the method being implemented by at least one processor, the method comprising: obtaining, by the at least one processor, a plurality of data related to a plurality of portfolios of securities from a public data source and a private data source; parsing the plurality of data by the at least one processor; identifying, by the at least one processor, personally identifiable information corresponding to at least one investor in the parsed plurality of data; removing, by the at least one processor, the identified personally identifiable information from the plurality of data and anonymizing the remaining data of the plurality of data; automatically determining, by the at least one processor using the anonymized data, at least one data pattern associated with triggering the margin call and at least one corresponding characteristic that caused the triggering of the margin call, identifying, by the at least one processor, at least one contributing factor to the margin call from the anonymized data; determining, by the at least one processor, an impact magnitude of each of the at least one contributing factor, the impact magnitude comprising a percentage of contribution to the margin call and a probability percentage of contribution to the margin call; as well as determining, by the at least one processor, the at least one corresponding feature from the at least one contribution factor based on the impact magnitude; automatically generating, by the at least one processor, in a memory, at least one model based on the at least one data pattern and the at least one characteristic; training, by the at least one processor, the at least one model by using the anonymized data; evaluating, by the at least one processor, the at least one trained model by determining, for the at least one trained model, whether a true positive rate, a true negative rate, a false positive rate, and a false negative rate correspond to predetermined ranges; determining, based on predetermined criteria, whether a level of accuracy of the at least one trained model developed using the anonymized data is sufficient; When it is determined that the at least one model is not sufficiently accurate, repeating the steps of collecting additional data and adjusting the trained at least one model by incorporating the additional data until the accuracy level of the at least one model is determined to be sufficient according to the predetermined criteria; When the at least one model is determined to be sufficiently accurate, deploying the at least one model; evaluating, by the at least one processor, the first securities portfolio using at least one deployed model based on a result of the assessment to automatically determine first information related to whether to trigger a first margin call for the first securities portfolio, wherein the first information includes a probability metric and corresponding triggering requirements; displaying, by the at least one processor, the first information via a graphical user interface; receiving, by the at least one processor, feedback information from at least one user via the graphical user interface; and The deployed at least one model is improved by the at least one processor based on the first information and the feedback information.

2. The method according to claim 1, wherein The assessment also includes: selecting, by the at least one processor, a first data pattern of the at least one data pattern and a first characteristic of the at least one corresponding characteristic, the first data pattern and the first characteristic being associated with triggering the first margin call for the first portfolio of securities; and The first information is displayed on a graphical user interface by the at least one processor via a display along with the selected first data pattern of the at least one data pattern and the selected first characteristic of the at least one corresponding characteristic.

3. The method according to claim 1, wherein The at least one model is automatically generated using a machine learning analysis of the anonymized data.

4. The method according to claim 1, wherein The at least one model includes at least one of a data model, a mathematical model, a process model, and a machine learning model.

5. The method according to claim 1, further comprising: evaluating, by the at least one processor, a second portfolio of securities using the at least one model to determine second information related to whether to trigger a future second margin call for the second portfolio of securities; as well as The second information is displayed on a graphical user interface by the at least one processor via a display.

6. The method according to claim 5, wherein: The graphical user interface includes at least one panel, and the at least one panel displays monitoring information related to the first securities portfolio and / or classification information related to the at least one characteristic.

7. The method according to claim 1, wherein The at least one characteristic includes at least one of the following: securities appreciation characteristic, securities depreciation characteristic, asset market value difference characteristic, securities market value difference characteristic, securities market value purchase characteristic, exposure difference characteristic, cash market value difference characteristic, collateral difference characteristic, cash market value sale characteristic, haircut difference characteristic, securities market value sale characteristic, securities liquidation value purchase characteristic, cash market value appreciation characteristic, cash market value depreciation characteristic, pledge market value outflow characteristic, pledge market value difference characteristic, cash market value purchase characteristic, derivative exposure increment characteristic, derivative initial margin exposure increment characteristic, pledge liquidation value inflow characteristic, accrued interest accumulation characteristic, pledge market value inflow characteristic, pledge market value appreciation characteristic, pledge market value depreciation characteristic, derivative negative market value increment characteristic, loan value fluctuation due to foreign exchange fluctuation characteristic, private equity subscription or accrued interest debit with future value date characteristic, and unlimited pledge difference characteristic.

8. The method according to claim 5, wherein The feedback information includes at least one of a first information accuracy ranking, a second information accuracy ranking, and a characteristic accuracy ranking.

9. A computing device configured to execute a method for providing information related to triggering a margin call on a first portfolio of securities, the computing device comprising: processor; Memory; as well as a communication interface coupled to each of the processor and the memory; Wherein, the processor is configured to: obtaining a plurality of data relating to a plurality of securities portfolios from public and private data sources; parsing the plurality of data; identifying personally identifiable information corresponding to at least one investor in the parsed plurality of data; removing the identified personally identifiable information from the plurality of data and anonymizing the remaining data of the plurality of data; By causing the processor to automatically determine, using the anonymized data, at least one data pattern associated with triggering the margin call and at least one corresponding characteristic that caused the triggering of the margin call in the following manner: identifying at least one contributing factor to the margin call from the anonymized data; determining an impact magnitude of each of the at least one contributing factor, the impact magnitude comprising a percentage of contribution to the margin call and a probability percentage of contribution to the margin call; and determining the at least one corresponding feature from the at least one contribution factor based on the impact magnitude; automatically generating at least one model in the memory based on the at least one data pattern and the at least one characteristic; training the at least one model by using the anonymized data; evaluating, by the at least one processor, the at least one trained model by determining, for the at least one trained model, whether a true positive rate, a true negative rate, a false positive rate, and a false negative rate correspond to predetermined ranges; determining, based on predetermined criteria, whether a level of accuracy of the at least one trained model developed using the anonymized data is sufficient; When it is determined that the at least one model is not sufficiently accurate, repeating the steps of collecting additional data and adjusting the trained at least one model by incorporating the additional data until the accuracy level of the at least one model is determined to be sufficient according to the predetermined criteria; When the at least one model is determined to be sufficiently accurate, deploying the at least one model; by causing the processor to evaluate the first portfolio of securities using at least one deployed model based on the results of the assessment to automatically determine first information related to whether to trigger a first margin call for the first portfolio of securities, wherein the first information includes a probability metric and a corresponding triggering requirement; displaying the first information via a graphical user interface; receiving feedback information from at least one user via the graphical user interface; and The deployed at least one model is improved based on the first information and the feedback information.

10. The computing device of claim 9, wherein: For the evaluation, the processor is further configured to: selecting a first data pattern of the at least one data pattern and a first characteristic of the at least one corresponding characteristic, the first data pattern and the first characteristic being associated with triggering the first margin call for the first portfolio of securities; as well as The first information is displayed on a graphical user interface via a display along with the selected first data pattern of the at least one data pattern and the selected first characteristic of the at least one corresponding characteristic.

11. The computing device of claim 9, wherein: The processor is further configured to: The at least one model is automatically generated by using a machine learning analysis of the anonymized data.

12. The computing device of claim 9, wherein: The at least one model includes at least one of a data model, a mathematical model, a process model, and a machine learning model.

13. The computing device of claim 9, wherein: The processor is further configured to: evaluating a second portfolio of securities using the at least one model to determine second information related to whether to trigger a future second margin call for the second portfolio of securities; as well as The second information is displayed on a graphical user interface via a display.

14. The computing device of claim 13, wherein: The graphical user interface includes at least one panel, and the at least one panel displays monitoring information related to the first securities portfolio and / or classification information related to the at least one characteristic.

15. The computing device of claim 9, wherein: The at least one characteristic includes at least one of the following: securities appreciation characteristic, securities depreciation characteristic, asset market value difference characteristic, securities market value difference characteristic, securities market value purchase characteristic, exposure difference characteristic, cash market value difference characteristic, collateral difference characteristic, cash market value sale characteristic, haircut difference characteristic, securities market value sale characteristic, securities liquidation value purchase characteristic, cash market value appreciation characteristic, cash market value depreciation characteristic, pledge market value outflow characteristic, pledge market value difference characteristic, cash market value purchase characteristic, derivative exposure increment characteristic, derivative initial margin exposure increment characteristic, pledge liquidation value inflow characteristic, accrued interest accumulation characteristic, pledge market value inflow characteristic, pledge market value appreciation characteristic, pledge market value depreciation characteristic, derivative negative market value increment characteristic, loan value fluctuation due to foreign exchange fluctuation characteristic, private equity subscription or accrued interest debit with future value date characteristic, and unlimited pledge difference characteristic.

16. The computing device of claim 9, wherein: The feedback information includes at least one of a first information accuracy ranking, a second information accuracy ranking, and a characteristic accuracy ranking.

Images