US 20050182709 A1
A financial scenario modeling and analysis tool, including a graphical user interface which enables a user of the tool to create a graphical model of a financial scenario, generally including at least one financial transaction, on a display screen, and an engine operable, in response to creation of the graphical model, to automatically generate information, such as financial or mathematical information, which at least partially models at least a part of the financial scenario using information collected by the engine during creation of the graphical model. The graphical user interface enables the user to create party graphics respectively representing parties to the financial deal, and to generate financial instrument graphics representing financial instruments, wherein each financial instrument graphic connects two of the party graphics. The engine generates, in response to the creation of a graphical model, an instrument information, such as an object or template, for each of the instruments in the graphical model. The tool includes a natural date language and a formula language for use in modeling a scenario. The tool enables optimization of optimizable parameters defined in the scenario, and includes a user-friendly, book-like and CAD-like user interface.
41. An electronic tool for performing mathematical calculations, comprising:
an information processing system; and
a graphical user interface for said information processing system;
wherein said graphical user interface is a non-cell based calculation interface which performs calculations using references which are based on a hierarchical outline and not a positional reference.
66. A financial scenario modeling and analysis tool, comprising:
means for creating a graphical model of a financial scenario on a display screen; and
means for generating, in response to creation of said graphical model, information which at least partially models at least a part of said financial scenario using data collected by said engine during creation of said graphical model.
67. The financial scenario modeling and analysis tool of
68. The financial scenario modeling and analysis tool of
69. The financial scenario modeling and analysis tool of
70. The financial scenario modeling and analysis tool of
71. The financial scenario modeling and analysis tool of
72. The financial scenario modeling and analysis tool of
73. The financial scenario modeling and analysis tool of
74. The financial scenario modeling and analysis tool of
75. The financial scenario modeling and analysis tool of
76. The financial scenario modeling and analysis tool of
77. The financial scenario modeling and analysis tool of
78. The financial scenario modeling and analysis tool of
79. The financial scenario modeling and analysis tool of
80. The financial scenario modeling and analysis tool of
81. The financial scenario modeling and analysis tool of
82. The financial scenario modeling and analysis tool of
83. The financial scenario modeling and analysis tool of
84. The financial scenario modeling and analysis tool of
85. The financial scenario modeling and analysis tool of
86. The financial scenario modeling and analysis tool of
87. The financial scenario modeling and analysis tool of
88. The financial scenario modeling and analysis tool of
89. The financial scenario modeling and analysis tool of
This application hereby claims priority on U.S. Provisional Patent Application No. 06/118,743 filed on Feb. 5, 1999, the disclosure of which is hereby incorporated by reference herein in its entirety.
The present invention relates to an automated tool for modeling the cash flows of financial scenarios, which typically involve use of at least one financial instrument, between various parties to a financial transaction by providing analysts with the ability to graphically represent the parties to the transaction, and their complex interrelationships in a manner that simplifies analysis of various options for completing the deal. In particular, the instant invention is directed to a modeling tool that analyzes various aspects of a proposed financial arrangement between parties to the transaction on the basis of information provided through a high level graphical user interface, and prepares competitive financial proposals, structures the proposals in an optimal manner, and which may further be used to manage and administer the final transaction to ensure compliance and delivery of the financial benefits determined by the tool.
The computer has become a critical tool for financial analysts whose job it is to analyze extremely complex financial transactions such as leveraged leases. The computer allows the numerous variables in such transactions to be manipulated and analyzed in a fraction of the time required for these calculations to be performed by hand. Of course, in order to allow a computer to perform useful functions, whether the area is financial analysis or virtually any other subject, software designed for the particular application is needed. Such software is often referred to as a “tool.”
Certain software tools for financial analysis of complex transactions have been developed; however, they have inherent limitations and are very difficult to use for a number of reasons, including, for example, their inflexibility in altering existing models, their requirement of complex commands and codes for building and modifying a proposed model, and their inability to manipulate a financial structure at the higher level of an overview. The invention described herein was designed to overcome the problems with these earlier tools and represents a major advance in the field of financial engineering and analysis. The invention incorporates extremely sophisticated aspects of computer aided design (CAD) resulting in a graphical user interface unique to financial analysis. As a result, an analyst using the invention is able to quickly and easily analyze many different potential scenarios and to determine optimal terms for the particular transaction under consideration. For example, this novel approach gives the analyst the ability to see partial results when building a model, provides the financial analyst with dynamic overviews (pictures) of the financial structure that can be directly manipulated to alter the financial structure, and provides an object-oriented distinction between high level structure and financial details which allow the user to defer details until they become available or relevant.
As described in greater detail below, an important part of the invention is a computer software engine which has been designed to automatically obtain and generate information relating to a particular financial transaction or scenario in response to inputs from the user. The software engine and the CAD-like graphical user interface have been designed to work cooperatively together in order to create a graphical representation of the particular transaction or scenario on the screen of the analyst's computer. The system is designed to allow the analyst to cause this graphical representation to be manipulated, modified or revised so that information useful to many different aspects of the transaction or scenario can be quickly and easily obtained. The end result is a system that is easy to use, extremely flexible and far more efficient than prior financial analysis tools.
There are many automated financial engineering and analysis tools currently available for use by analysts to determine various components of a financial transaction and to optimize the transaction based on the particular data associated with the parties to the transaction. One such well-known tool is provided by Warren and Selbert, Inc., of Santa Barbara, Calif. and is referred to as “ABC”. ABC has been used by analysts to generate various alternatives within the constraints of a particular financial instrument and optimize the results so that analysts can generate a deal that is acceptable to all parties to the transaction. One such commonly used financial instrument is referred to as a multiparty leveraged lease. There are various other proprietary systems that provide such automated financial analysis.
However, all of these known systems suffer from numerous disadvantages. For example, the ABC program, and others like it, require the use of complex commands and codes for building and altering a proposed model. Moreover, the models typically must be built prior to having the ability to view any intermediate results. This, in combination with the complex programming-like language that is required, results in a very long learning curve for analysts who use the tool. Furthermore, the model, once built and run, does not typically enable the analyst to easily change variables or to easily view the resulting change in the transaction.
A primary source of these problems is the complex and inflexible user interface typically associated with these known tools. Another problem with such prior art tools is that they do not enable a user to model the financial deal visually and mathematically and in a manner which enables interfunctionality and dependency between the visual model and the mathematical model. As a result, the tools currently in use provide limited ability to deal with higher levels of complexity and the ever expanding universe of evolving financial products in use today, and which will be used in the future. Additionally, the inflexible interface makes it very difficult for different analysts to be able to discern the exact relationships and variables of a model that another analyst may have been manipulating when the model was being built and later modified.
To overcome the above and other shortcomings with prior art financial modeling tools, the present invention provides a much more user friendly, flexible tool incorporating easy to understand graphics and interfaces to enable more efficient and practical application of the tool. To that end, the invention provides a financial modeling tool that addresses model complexity with a graphical CAD-like approach to financial and/or mathematical modeling, which facilitates, among other things: the ability to see partial results while building a model; a short learning curve; the ability to make changes when the user views the results of the analysis; flexible “point and click” interfacing; easy handling of indexed data; integrated and automatic handling of certain variables, e.g., taxes and accrual; menu of building blocks, e.g., loans, rents, fees, purchases, etc.; menu of built in reports; and an interactive and intelligent graphical representation of the model.
In accordance with an important aspect of the instant tool, a software engine, hereinafter referred to as “engine”, is provided in the tool and is programmed to automatically obtain and generate information on a financial scenario in response to the user creating a graphical representation of the scenario with the CAD-like user interface. In other words, the manipulation of the graphical user interface to generate a visual representation of the scenario automatically results in the generation of information, such as formulas, objects, templates, timelines, calculations, constraints, parameters, optimizable parameters, cash flows, reports, or any other suitable information that is helpful in modeling the scenario represented by the visual representation created by the user using the CAD-like interface. The information generated preferably at least partially model at least a portion of the scenario. After drawing a scenario, such as a proposed financial deal, using the interface, the interface enables the user to enter data and formulas, edit the information automatically generated by the engine in response thereto, and to further define the scenario in a manner which enables the engine to fully model and analyze the scenario. Once the scenario is fully modeled, the tool gives the user the ability to instruct the engine to attempt to optimize the scenario, either directly or by creating formulations to be optimized and passing the formulations to a separate optimizing program. Once the deal is optimized, the results can be viewed by the user using the interface. The scenario can also be modified by the user and new results based on the modification can be viewed. When the visual representation of the scenario is modified, the engine automatically modifies the information previously generated in a manner which corresponds to the modification to the visual representation.
In accordance with a main aspect of the instant invention, a financial transaction modeling and analysis tool is provided which includes: a graphical user interface which enables a user of the tool to create a graphical model of a financial scenario, generally including at least one financial transaction, on a display screen; and an engine operable, in response to creation of the graphical model, to generate information which at least partially models at least a part of the financial scenario using information collected by the engine during creation of the graphical model.
The graphical user interface preferably enables the user to create party graphics respectively representing parties to the financial scenario, and to generate financial instrument graphics representing financial instruments, wherein each financial instrument graphic connects two of the party graphics. The party graphics and the financial instrument graphics define the graphical model of the financial scenario. Preferably, the financial instrument graphics indicate a direction of flow, relative to the financial instrument represented thereby, between the parties connected by the financial instrument graphic.
In accordance with an important aspect of the instant invention, the engine generates, in response to the creation of a graphical model, an instrument information, such as an instrument object or template, for each of the instruments in the graphical model. Once an instrument is defined, the graphical user interface enables the user to interact with the instrument information, such as adding scenario specific instrument data to each of the instrument objects generated by the engine. The instrument data entered in connection with the instrument object constitutes either a fixed part of the financial scenario or a variable part of the financial scenario.
The graphical user interface also enables the user to enter and define date information relating to the financial transaction for use by the engine. Preferably, the graphical user interface is operable to display the date information in graphical form on the display screen. The tool preferably enables the date information to be entered using a natural date language, wherein the engine is operable to process the date information from the natural date language.
In accordance with another aspect of the invention, the graphical user interface enables the user to modify the graphical model of the financial scenario, and the engine is operable, in response to the modification of the graphical model, to modify the information previously generated in accordance with the modification of the graphical model.
In accordance with another aspect of the invention, the engine is operable, in response to the creation of the financial instrument graphic, to define roles for parties represented by the party graphics connected by the financial instrument graphic, wherein the roles are used by said engine to define the parties interaction with the financial instrument represented by the financial instrument graphic when modeling the financial scenario.
The engine is preferably operable to determine and display an optimal solution or result for the financial scenario relative to at least one of the parties thereto, and to calculate optimal values for each of the variables defined by the instrument data based on the optimal solution.
The financial transaction modeling and analysis tool of the instant invention preferably includes an extensible library of predefined financial instruments, and the graphical user interface enables the user to select and use one or more of the predefined instruments during creation of the graphical model of the financial scenario. In other words, numerous common and canned financial instruments are provided to the user to facilitate easy modeling of common transactions that may be used in financial scenarios.
In accordance with another aspect of the invention, the engine is operable, in response to creation of each of the party graphics to generate a party-specific information on the party, and the graphical user interface enables the user to retrieve and modify the information in the party-specific information.
In accordance with another aspect of the invention, the graphical user interface includes a worksheet section, also referred to herein as “smart paper,” which enables the user to input scenario information which is independent of or supplementary to the date and instrument information relating to the financial scenario, and the engine is operable to use the scenario information when modeling the financial scenario. Preferably, the instant tool includes a formula language for use in creating the scenario information, wherein the formula language includes a library of predefined functions and keywords which can be used by the user when creating the scenario information.
The worksheet section is preferably a non-cell based, outline based interface for inputting data and formulas in an outline format. More particularly, the worksheet, also called “smart paper” herein, is a non-cell based calculation interface wherein references are based on a hierarchical outline rather than a position reference. In a preferred embodiment of smart paper, the interface enables one formula per row to be defined in an outline-type format.
The engine is preferably operable upon entry of scenario information, such as deal formulas, in the worksheet section to establish links between related scenario information and between scenario information and date information, thereby establishing a dependence therebetween, and further wherein the engine is operable to use the links when modeling the financial scenario. Preferably, the tool includes a library of predefined worksheets for use in the worksheet section, and the graphical user interface enables the user to select predefined worksheets from the library for use in the worksheet section.
The instant tool also enables a plurality of possible outcomes to be modeled based on different information provided by the user.
In accordance with yet another aspect of the invention, the graphical user interface is presented on the display screen in a book-like configuration in which a plurality of different sections of the graphical user interface are represented by different chapters in the book-like configuration, each of the chapters having a tab graphic associated therewith, wherein upon selection of the tab graphic by the user, the user interface is operable to display the chapter associated therewith. Preferably, the tab graphics are located along a side of the display screen, and each chapter may include a plurality of pages, the pages having page tab graphics which are also displayed to the user when a chapter having the pages is selected by the user.
Preferably, the graphical user interface enables the user to view two of the chapters simultaneously in a split-screen format on the display. The engine is preferably operable to update information in each chapter in response to changes made by the user in a chapter.
In a preferred embodiment of the graphical user interface, the chapters include a diagram chapter for creating the graphical model, a parties chapter for providing data relating to the parties, a time chapter for viewing and editing dates associated with the financial deal, an instruments chapter for viewing and editing instrument data, a worksheet chapter for enabling the user to define scenario information or formulas relating to the financial scenario, an optimization chapter for use in optimizing the financial scenario, a payment chapter for viewing payment flows in the financial scenario, and a reports chapter for enabling reports to be generated relating to the financial scenario.
These and other objects, features, aspects and advantages of the instant invention will become apparent to one skilled in the art upon review of the detailed description of the invention provided herein when read in conjunction with the appended drawings, in which:
Referring now to the drawings,
In order to enable a better understanding of the instant invention, the following glossary of definitions are provided for terms commonly used herein to describe the instant invention:
Referring now to
If the model as defined satisfies the requirements of all parties involved, the tool provides an interface which enables the user to create various reports relating to the model generated for the scenario. These reports include, for example, the flow of instruments and assets over various time periods. On the other hand, if variables exist in the scenario based on requirements of the parties, the tool enables the scenario to be optimized. For example, the deal may require that a party obtain a return on investment of at least 5%. The party may desire an even greater return provided all other aspects of the model of the scenario are satisfied. Such requirements are known as constraints of the model. The process of optimization involves creating the best model which satisfies all such constraints and determines the best possible model based on the requirements and goals of the parties. The instant tool enables optimization against a number of constraints that may exist in the scenario.
The tool operates in two basic modes: build mode and end user mode. In build mode, the user creates the definition of certain aspects of the tool, such as creating instruments which model real world financial instruments. These instruments are then stored in the tool for use by end users when modeling a scenario using the tool. In other words, the builder user provides a library of “canned” instruments which can be used by the user to more easily and efficiently model the scenario with the tool. The instruments involve a set of inputs and calculations based on those inputs. The end user incorporates the built instruments into a model and supplies the real inputs corresponding to the actual deal that is being modeled. The build user mode also enables the builder user to create calculation templates to be used by the end user in conjunction with various instruments.
As indicated above, one step involved in modeling a deal using the tool of the instant invention involves creating a party diagram (step 30) or graphical model of the deal.
The Graphical User Interface (GUI)
The following is a description of the functionality of the present invention in terms of the graphical user interface (also referred to herein as “GUI”). It is noted that, in accordance with the instant invention, the GUI may be implemented using Windows 98, Windows NT, MAC, or it may be Web-based. In other words the instant system may be implemented on any suitable standalone, networked or web-based platform. In a preferred embodiment, the tool can be implemented using the following hardware, however any suitable hardware may be used in accordance with the invention:
The instant invention preferably uses a book-like display as a viewing device. More particularly, the invention displays its data as pages in a familiar-looking tabbed notebook configuration on the display screen. Each tab of the notebook represents a “Chapter”, corresponding to one part of system's functionality. Users can work with one copy of the book visible (see
The first chapter in the GUI is the Payment Diagram (see
The next chapter is the Parties chapter. The program simulates “parties,” which are entities that participate in a financial transaction. It provides automatic creation and deletion of party-specific information. This party-specific information includes items, such as tax rates, fiscal-year-ending months, or yield requirements. The GUI provides support to the Payment Diagram for party detail information. The GUI also provides the user with a standardized and extensible location for party data. Each party can have individual tax attributes, paying different kinds of taxes to different governments (U.S. states or foreign countries).
The next chapter is the Time and Decision Organizer, which is also called “Time Organizer” in the program for short (see
The next chapter is the Instruments Chapter (see
The system allows paying and receiving parties to have distinct interpretations of the instrument payments, tax effects, and classification. This is an important feature, because tax law often makes these distinctions. Instruments contain pre-built and pre-tested Smart Paper computation sections, making the system more reliable and freeing the user from having to do repetitive programming. The system automatically activates and deactivates instrument Smart Paper computation sections according to the user's selection of role information for the parties. Thus, sections not in use remain visible and available and do not distract the user. Glyphs are used to highlight the relationship between role specifications and the calculations. Specific items representing payment or receipt of funds, or of taxable income or deductions, are identified with symbols that make those items easy to find. The system also allows the user to change the name of the instance of the instrument. These name changes show up on the party diagram and the payment organizer, making both of these a lot easier to read and understand. In addition, model pieces can be named with the names that other transaction negotiators are using, making communication a lot easier. The system also automatically generates the parallel payment streams for different outcomes using handlers for each possible decision. Each instrument can thus generate different specific payments depending on the state of various contingencies. This is important because exercise or non-exercise of certain options can mandate different behavior on the part of the same instrument. The system allows the user to modify the handlers' termination behavior. Thus, special cases do not require modifications to the system. In addition, the system uses Smart Paper “protection” modes to preclude user corruption of instrument functionality, but otherwise allows users the ability to modify instrument Smart Paper. Thus, users can be confident that canned (and therefore tested and reliable) model parts are being used, instead of model parts which may not be correct given an unforeseen peculiarity of a particular model.
Users can supply their own formulas. Such formulas are clearly marked, so that other users know they have to validate the formulas before using the model. The system also provides canned calculations for specific types of financial elements (e.g. rent, loans, etc.). These canned calculations cover a very large fraction of the payments users would run into when modeling a financial scenario. Thus, users will spend little time having to invent new payment mechanisms. In addition, this set of canned calculations is contained in an expandable library, so as the industry changes, additions can be added to the library to keep it up to date. The system provides the user the ability to customize the calculations, making use of the invention a lot easier. Pre-defined reports of the instrument results are also provided. As a result, explanations of what an instrument is doing are only a click away. The system also allows the user to specify that an instrument only exists when a certain decision is assumed. Contingent instruments can be put into the model and will thus automatically be properly handled. The system also supports automatic decision and outcome creation for termination values and other make-whole payments. These contingency dates are too numerous to include as separate instruments or outcomes, and so including them here provides a compact way of computing them as a class. In addition to general financial instruments, the system may also include instruments for advanced corporate finance operations, such as mergers, acquisitions and the like.
The next chapter is the Smart Paper chapter (see FIGS. 20 & 25-29). Smart paper is a powerful non-cell based calculation interface wherein references are based on a hierarchical outline as opposed to a positional reference. Unlike spreadsheet programs, smart paper is non-cell based and does not rely on a positional reference for use in calculations. Smart paper is the bridge between the ease-of-use that spreadsheet users depend on, and the power of financial and optimization packages. It makes computations visible, understandable, and accessible. Users do not need to be computer programmers or learn to work as programmers in order use the system effectively and efficiently. The system has enhanced data capabilities, which automatically perform a lot of rote date-related manipulation that makes spreadsheets hard to create and even harder to modify. Thus, the system provides capabilities normally found in relational database packages. The smart paper chapter is the component where the user specifies the computations he wants the system to perform. Users can define values directly, or they can provide a formula which will tell the system how to compute the desired values. Smart paper provides user control over outline-like (i.e. tree-like) format of parameters and several nested layers of headings. This makes Smart Paper work much more readable, and provides a mechanism for the system to resolve (or ask the user to resolve) formula ambiguities. Smart paper allows the user to create one or more sheets of smart paper. Related computations can be kept together, and unrelated ones can be segregated. Tabs are provided for moving among sheets of smart paper. As a result, users do not feel lost in the program and are able to find quickly what they are looking for using the GUI. The GUI provides controls for viewing “formulas” versus “results” or both. Thus, users can get immediate feedback as to whether they have properly specified a formula. The GUI provides editing capabilities for headings and parameter names, as well as provides access to the template library and instantiation of templates. Smart paper defines parameter name scope and parameter index scope automatically via outline format. This resolves many “name clashes,” which would be otherwise inevitable in a model of any size. It also provides a view of dependency relationships among parameters. Users can thus identify information relationships among their parameters. Also provided is general support goal-directed “search” for setting parameter values. The system automates some of the trial-and-error involved in changing parameters' values in order to produce the desired answer. It also allows the user to specify formulas that define (dynamically) activation/deactivation of sub-trees. This gives models the ability to be “context-sensitive,” responding sensibly to particular values of input data. It also provides capability for import/export of data and formulas from/to Microsoft EXCEL or the like. Models can be created using the system and the system will automatically reconstruct the model in Microsoft EXCEL.
The GUI also provides a date stream bar that always displays the index of the uppermost indexed stream. The important pairing between dates and date-indexed data is therefore always visible on the screen. This eliminates a lot of meaningless clicking back and forth to keep the index visible, and eliminates the need for a lot of “split-screen” display. The GUI supports multiple data-entry modes: simple-edit, full-edit and rapid-entry. These modes make constructing models easier and faster. Also, smart paper items (e.g. headings, templates, formulas, indexes) can be changed into each other. Tool bar and menus provide editing to morph, promote, demote, insert, and delete operations. This eliminates the need for learning a lot of jargon. Instead, all the alternatives are displayed, and the user can choose the most logical one. The system also provides graphic feedback as to the type of indexed data represented by parameters. Moreover, in accordance with the invention formulas for dates resemble “English-language” instruction. Formulas can be printed, providing human-readable documentation for a model. This differs significantly from spreadsheets, wherein formulas consist largely of a list of data locations instead of identities, resulting in a nearly useless documentation tool for a human. The system creates templates as white-box functions which allow the user internal access. Thus, there is no need to refer to external, written documentation to figure out what a template/function is doing, because all the code is right there, visible.
In addition, smart paper has an extremely powerful formula language (with input/edit wizards). This formula language automates many tasks which are routine in finance but which are now cumbersome for spreadsheet users. The formula language, which is described in greater detail below, has the following exemplary features:
The next chapter is the Optimization chapter (see
In addition, the system gathers all model constraints for viewing on one page. This is important because constraints are often of definitive concern in tax-motivated transactions. The system also sorts constraints by failing, binding, non-binding, and inactive. This helps negotiators identify the critical points in their deals, or helps a user figure out why his model might fail to provide an answer he would expect. It also provides detail results of the values, slack, shadow prices or failure margin of constraints. This information helps the user explain anomalous results, or suggest ways that financial objectives may be attained at less cost. Also provided is a simple algebraic list of all constraints, making it easier to make sure that no constraints have been mistakenly taken out or left in. This list can be printed, and becomes an important “output” for deal participants to examine and approve. The system also gathers all optimizable parameters, search parameters, and non-linear parameters for display separately on their own pages, helping make sure that the linear program model has been set up properly. It also automatically determines whether there are any parameters causing non-linearity and thus precluding being solved by the built-in linear optimizer The user can then more easily decide how to change the model to be solvable by the available optimizer. The system automatically traces the non-linear parameters to corresponding optimizable parameters and displays them with the non-linear parameter. This helps modelers find linear models that are approximate solutions to otherwise unsolvable nonlinear problems. The system also allows users to specify facets of optimization in a way that is intuitive and easy to understand. Constraints are expressed in terms of comparisons and not just formulas. Prior art systems treat formulas with constraints as one and the same making it difficult to discern between a formula that is incorrect and a formula which is correct but not satisfied in the model. Constraints are entered and evaluated separately from the associated formulas making it easy to see where the true problem lies. Results from the optimization software (e.g. CPLEX) are likewise translated back into these terms producing a result that users can easily interpret. It also provides button access to the optimizer. The system automatically converts the model to “optimizable” form for the optimizer, and then re-converts the optimizer's solution to code; thus, optimization is a transparent process to the user.
The GUI presents user with detailed progress screen and saves the results back to the user's file. The system also re-evaluates constraints in the context of the user's model to provide more useful feedback from optimization. This is important because the specific analytical assignment is often to reverse-engineer the set of constraints that produced a particular answer. The system also automatically invokes Successive Linear Programming (SLP) when needed to solve for a search parameter. This saves the user the labor of having to determine that the problem at hand is SLP and then perform the SLP by successively invoking the optimizer. It also automatically combines point-by-point searching and mathematical optimization, saving the user the labor of searching over various valid values of nonlinear data.
The next chapter is the Payment Organizer (see
The next chapter is Reports (see
The next chapter is Template Builder. A “template” is a white-box piece of pre-built and pre-tested Smart Paper. A library of such templates provides a large part of the built-in financial knowledge that users can draw on. This makes it worthwhile for an organization to invest in well-constructed Smart Paper pieces that can be reliably shared. The GUI provides the ability to create Template definitions to be used for instantiation. Generally, only specially trained “builder/users” would invoke the Template Builder. It provides user access to protection controls over read/write specifications of parameters. It also allows the builder of templates to work in the same manner as in regular Smart Paper. Thus, they are familiar in appearance and do not require users to learn how separate components look and work. The system provides linkage to the currently active model so as to provide useful feedback to the builder. It also allows the builder to create, edit and delete template definitions, and to make these definitions available to the larger user community.
The next chapter is Instrument Builder. This chapter provides a special “build” mode for instrument builders to create the definitions of instruments. Instruments can be designed to, for example, calculate tax effects for numerous governments (U.S. states, foreign countries) simultaneously. Tax instruments can be designed to, for example, handle complex multinational tax interactions (foreign tax credits and the implications of various international tax treaties).
The next chapter or feature contains Main Menu Items. These items include an extensive, full-featured on-line “help” system which provides documentation for system features and behavior, and also provides financial examples which can serve as a tutorial; options for adjusting the program's appearance (horiz/vert tabs); and a program that follows the Windows NT “Control Panel” settings for cosmetics (dates, mouse clicks).
In addition to the above, there are some general features found in several of the chapters or components of the system described above. For example, print previews are provided for enabling the user see output before committing it to paper, thereby saving time, paper, and aggravation. In addition, general cosmetic formatting capability is provided which is similar to that found in commercially available word processing or spreadsheet programs.
Another chapter or feature of the invention provides the ability to manage multiple cases. A financial analyst may spend as much time managing and summarizing the cases he produces as he would spend producing the cases in the first place. The GUI also provides tree, containment, or list views of the files. It also provides an overlapping grouping capability, so files can be treated as a group in addition to individually. The system also provides a mechanism for extracting key results for several files and presenting them in a tabular report. It also provides file search capabilities based on the contents of the file in addition to the file's name, as well as filtering and sorting capability (e.g. show only my files). A “recent files” section and “all-files” section is provided. This is important because a great majority of the time users work on a single project continuously, and, with this feature, they don't have to look through hundreds of files to find one they were recently working on. The system intelligently identifies differences between two or more selected files, making it easier to explain their differences. It also provides capability to reorganize files (i.e. the tree), and the capability to read and add notes to files. Thus, a file's owner can protect the file from changes, while letting colleagues look at, and then attach notes to the file. In addition, the system provides detail records of file (e.g. ancestry). This is important because the great majority of files will probably not be created anew, but rather will be modified versions of existing files. The ability to track the changes that produced a file is a very advantageous and time-saving feature.
It is noted that not all of the “chapters” discussed above necessarily have a “Tab” always visible on the GUI. In other words, some of the chapters or features, i.e. Template Builder, are accessed through menu items or other suitable means for enabling the selection thereof.
Referring now more particularly to
In accordance with an important aspect of the invention, the Engine is operable, in response to creating of a graphical model, to automatically create useful information in certain of the other chapters. More particularly, in response to drawing an instrument in the Payment Diagram chapter, such as “loan” as shown in
Of course, all of the information shown in
The following description provides an example of case or financial scenario modeled using the tool of the instant invention.
This exemplary case is called a QTE or Qualified Telecommunications Equipment case. A graphical representation of this exemplary financial scenario is shown in
The following steps present a high level description of what is happening in this deal:
The first step in creating a case involves drawing the diagram shown in
This image also shows the general interface for the product. It is noted that not all figures herein show the full interface of the full chapter. In other words, some of the figures only show an enlarged partial view of the interface or chapter. The tabs 76 on the left side of the screen are used to navigate around the program to the different interfaces or “chapters” of the user interface. Each major step in creating a deal is represented by a chapter. The icons under the menu bar are general purpose tools used, for example, to save a case to a hard disk or load a new case.
To do this first step, the user clicks the party tool (the round box with a plus sign) and then clicks in the drawing area where he wishes to locate the party. He can then select from a list of predefined parties or enter his own.
When the user adds a new box to the party diagram, such as party 78, there is no specific action in the engine, except for the possible generation of a party object in the Parties Chapter. If this is a party that does not appear anywhere else in the payment diagram, then the engine adds a new section to the Parties chapter with two parameters which the user can then edit and augment. The two parameters represent, for example, the first month of that party's fiscal year, and the name of his tax counsel. This data is reflected in engine parameters to provide persistence to the data, and to make the data available to instruments which will be connected to this party.
Referring now to
When the user draws an instrument 80 on the Payment Diagram, the application creates a representative object in the engine to calculate and generate the calculations and cash flows for that instrument. More specifically, the client application tells the engine General Registry to create a new instrument object, which initially is an empty shell. The client application reads in from a data file the representation of the instrument that has been saved by the instrument designer, and splits this up into graphic and engine data and passes the engine part to the new Instrument object. The engine then reconstructs the instrument from this data; this includes a hierarchy of parameters and parameter lists, a couple of instrument party objects, and a default instrument handler. Each instrument party contains the role information needed to create the cash and tax implications for the party at one end of the instrument arrow. The default instrument handler contains the information needed on whether and how to truncate the flows in non-base outcomes. Each parameter's formula is parsed into its expression objects, and named references are registered with the general registry's name reference manager, which, if possible, resolves them, so that the parameter's values can be calculated when needed. As these references are resolved, links are made between the parameters' dependency managers so that changes are correctly propagated through the entire system. For each decision that the user has created in the time organizer, the engine creates an instrument handler containing the information on whether and how to truncate the flows for outcomes containing that decision. Each handler starts as a simple copy of the default handler, but can be changed by the user. The client tells the engine instrument object the names of the parties at each end of the arrow. The engine then looks in the parties data to find a section of data for that party, which it then uses to complete the instrument party role information (e.g. date of fiscal year end). The engine instrument synchronizer object then springs into action, generating the actual flow parameters for each party and outcome. The instrument party information is used to determine which parameters contain the cash and tax flows for that party. For non-base outcomes, the synchronizer searches for the first decision that terminates the flow, according to the instrument handlers. It then generates a truncating parameter (if necessary), and a final parameter which it identifies to the internal database by attaching badges indicating the party, cash or income classification, outcome, instrument name, other party and tax authority (for income flows). The engine internal database recognizes these new badged parameters and changes any collected data in (for example) yield calculation templates to update their values. This process also enables the payment organizer to update its display to show the new instrument.
The user continues to add parties and instruments until the deal is modeled fully as shown in
Step 2—Define the Dates
Dates and timelines are defined in the Time Organizer as depicted in
In the bottom section 84 of the interface, the user defines the overall deal dates called the EventDates. This sets the basic start and end of a deal and also the periodicity (annual, quarterly). The user changes the EventDates by right clicking on the line and selecting “Edit timeline” from the menu. The other lines are for visual purposes only. All key dates defined in the deal are shown. Each primary cash flow for an instrument is represented by a line in the time organizer. Badging information stored within the engine for each instrument defines which is the primary cash flow for that instrument.
For this exemplary deal, the following changes are defined in the Time Organizer:
Once the EBO date is defined, it is turned into a “decision” by the engine, which means the deal splits in to two possible courses or “outcomes”. The normal case, or BaseOutcome, means the deal comes to term and the assets are just sold to the market place. The EBO outcome occurs when the lessee purchases the assets prior to the end of the deal. For the purpose of financial analysis, all outcomes can be fully modeled to get the deal approved by the client. The outcome tool, the triangle with a plus sign (currently disabled in
When the user adds an outcome to the time organizer chapter, the engine performs numerous functions. More particularly, the engine creates a new outcome object to house data specific to this outcome, with links to the decision objects which comprise the outcome. For each tax-like instrument (for which the instrument designer has specified a “fresh copy for each outcome”), a complete copy of its parameter, party and classification data is generated. For each instrument, the instrument synchronizer generates new terminating parameters as necessary. These parameters terminate the flows generated by this instrument at the first decision contained in the new outcome that has been designated by the user as a terminating decision in the instrument's decision handlers. For each instrument, the instrument synchronizer generates new badged parameters to identify the flows to the internal database. These parameters take their value from the terminating parameters defined in the previous section (or the original parameters if not terminated), with possibly a sign change. They are given badges based on the data in the party and classification sections of the instrument; the categories are Party, Outcome, Cash or Income Classification, Other Party, Instrument Name and (for income classifications) Tax Authority. When the internal database receives the information about the new badged parameters, it signals all effected collector parameters that a change has occurred. These are parameters which have been designated as extraction parameters, or as formula parameters using one of the special “collect” functions. The next time that their values are requested, they will re-establish all links with badged parameters so that the new ones are included. The collector parameters signal a value change via their dependency managers to tell all other parameters whose value depends on theirs that a change has occurred. In this way the flows generated by the new outcome spread their effect throughout the model.
Step 3—Entering Instrument Data
Once the diagram is created and the dates are defined, the user next fills in all the data and calculations necessary to complete the deal. The first part of this is filling in instrument data. Instrument data is entered in the Instruments chapter as seen in
The first two tabs near the top of the figure and next to the “Calculations” tab are used to enter role information for each party associated with an instrument (as defined by the payment diagram explained earlier). In
Starting with the loan instrument, the only item which needs to change is the loan rate. The other default or initial values for the instrument are sufficient for how the loan should be setup in this deal. Money is borrowed at a fixed rate of 5.0625%. The steps are as follows: 1) Double click on the current value for the loan rate (this is the Rate parameter in the Calculations section of the Loan instrument); and 2) Change the value from the default value to 5.0625%. (See
When the user changes the Rate formula in the user interface, the engine will record the new formula for that parameter, parse it into its expression objects, and transmit a value-changed message through the network of dependency managers for parameters whose values depend on this parameter. Each of these parameters will then know to recalculate its value when it is next requested.
To complete the instrument changes, data will need to be entered for several other instruments. Instrument data generally falls into two categories. It is either a fixed part of a deal such as the cost of an asset or the interest rate of a loan. Or it is an optimized value that gets calculated when the optimal solution is found. The instruments in this case are set up in the following way:
No changes are needed in the Taxes instrument since this is based on the standard federal tax rate.
Step 4—Building Smart Paper
Any data or calculation not specified as date or in the instruments is entered in Smart Paper. Smart Paper is a calculation based feature very similar to a enhanced spreadsheet (more details on Smart Paper are provided below). However, while a spreadsheet is based on individual cells linked together strictly by formulas, Smart Paper formulas know about each other and about links to dates. More particularly, as explained above, Smart Paper is a non-cell based calculation interface where references are based on a hierarchical outline as opposed to a positional reference. The linking information is stored in the engine. For example, one formula may contain a set of values linked to the first date of every year for 20 years. A second formula may only need the value from a specific date, such as the fifth date, within the first formula. The second formula need only specify the specific date and the engine will search out the most appropriate value.
The Smart Paper in this deal is built up in two ways. First, the user has a variety of templates he can add that perform pre-defined calculations. Second, the user can create custom calculations or enter custom data into Smart Paper.
Each tab is a different sheet of Smart Paper where the user can create his outline and enter his data and calculations. When the user adds a piece of Smart Paper, the engine creates a worksheet in the General Registry. As the user creates Headings and Parameters in the piece of Smart Paper, the engine creates mirroring Parameter Lists and Parameters. When a Parameter is created and named, the engine registers it with the name reference manager. This will attempt to resolve any outstanding references to this name by formulas in other parameters. It will also see if references to other parameters of the same name need to be more fully qualified to prevent ambiguities. When a reference to this parameter is resolved, the referencing parameter sets up a link between its dependency manager and that of the new parameters, so that any changes in value of this parameter will be signaled to the referencing parameter.
When the user specifies the formula for the new parameter, the engine parses it into its basic expression nodes. Any references to other parameters are registered with the name reference manager which will attempt to resolve it immediately. If it cannot be resolved immediately, then the name reference manager keeps the request as pending, in case it can be later resolved. Any resolved reference causes links to be set up between the dependency managers as described above.
As the user enters parameters into Smart Paper, the values of these parameters are immediately displayed. It does this by asking the engine to calculate the parameter's value, which triggers an evaluation of the parsed expression nodes. These nodes do the basic arithmetic operations, references to other parameters and evaluation of functions. The value returned can be either a scalar or some sort of array. Scalars are single quantities like numbers, dates, frequencies, elapsed times, character strings. Arrays are lists of these scalars.
Step 5—Optimizing the Deal
Optimization is the process of imposing constraints or requirements on a case and the varying values and other parts of the case until the best result is found. By a constraint, it is meant, for example, that some cases fall under certain restrictions from, for example, tax laws relating to leasing and rents which must be satisfied if the case involves a lease or rents. The elements of a case that can be varied are called optimizable parameters.
In this deal, we are maximizing the present value benefit to the lessee. The following constraints exist on this deal and must be satisfied when optimizing to the best result:
The following parameters are then varied to reach the optimal deal:
The optimization screen is divided into several pages by the tab across the top of the screen, as shown in
When the user hits the optimize button, the engine analyzes all the parameter definitions and constraints that the user has entered, and tries to set up a linear (or mixed integer) programming representation of these suitable to be sent to the CPLEX linear optimizer. Assuming that this can be done, it sends the model to CPLEX, gets the results back and puts the resulting values for optimizable parameters back into their formulas.
Step 6—Viewing Output
The final step is viewing the data either in the reports chapter or in the payment organizer. A report from the reports chapter is displayed in
The data for a report is collected directly from Smart Paper and instruments. The only function the reports chapter performs is formatting the data for professional output. Likewise, the Payment Organizer chapter, allows the user to view the data and cash flows according to a specific party, outcome and time frame within the deal. This again is only a viewing interface which collects data directly from the data and calculations entered in other parts of the model. The Payment Organizer interface is displayed in
As can be seen from the example case above, the user can model a financial scenario easily and quickly using the tool of the present invention.
The following are examples demonstrating the functionality of the worksheet section or Smart Paper feature of the instant invention.
Smart Paper is a non-cell based calculation interface where references are based on a hierarchical outline as opposed to a positional reference.
The smart paper example of
From this example, we see some of the benefits of the non-cell based worksheet of the instant invention. For example, if another plane is added under the heading Aircraft, and the rent stream is called Rents, then the TotalRents parameter will always show the total of all rents, because the Subtotal function finds all parameters named Rents under the Aircraft heading. Likewise, if a rent payment is added to any of the existing Rents parameters, TotalRents is automatically updated. In a conventional spreadsheet, solving these two problems would ultimately involve inserting cells or rows or columns and updating formulas that sum the data. The hierarchical nature of the outline, made possible by Smart Paper, lets the same name be used more than once in the manner indicated above. As a result, a very convenient, flexible and powerful calculation interface is provided by the Smart Paper chapter of the instant invention.
This example also demonstrates the advantage of the dynamic non-cell based formulas used in Smart Paper. For example, the AnnualTotals collects all the rents paid for each year. In a spreadsheet, the user would have to examine each rent stream and individually select which rent payments fall in each year. If the Annual table then needed to be changed to quarterly, the user would have to go back and re-do the entire process from scratch. However, with the non-cell based worksheet of the instant invention, formulas know how to link values to dates so that the final formula can interpret the input values based on the actual date rather than the position the date falls in a spreadsheet, which relies on positional references rather than the hierarchical references of the instant invention.
Similar to the AnnualTotals, the TotalDates parameter benefits from non-cell based references if a new date is entered for any rent stream. The TotalDates will always collect all rent dates regardless of how many or few there are.
A second smart paper example is shown in
A third smart paper example is shown in
As can be seen from the examples above, the Smart Paper feature of the instant invention provides a very useful calculation interface and tool. It is noted that the Smart Paper tool can, in accordance with the instant invention, be used independently from the modeling and analysis tool of the instant invention as an improvement to spreadsheet applications.
As explained above, the graphical user interface and the engine provide an intelligent interface which enables data to be generated which models the deal in response to graphical modeling of the deal by the user. Thus, the graphical model not only provides a visual representation of the deal, but it also causes the engine to generate useful information which at least partially model the deal based on the information the engine is able to obtain from actions performed by the user during creation of the graphical model of the deal.
In the preferred embodiment of the instant invention, the engine operates in accordance with the description below. More particularly, the engine is a computational server designed to support client applications wanting spreadsheet-like formula evaluation, manipulation of indexed streams of quantities and linear and mixed integer programming optimization. The engine has the following main features:
The engine is designed as a COM server which can be initiated either in-process or out-of-process. In the latter case, it can be either local or remote, and can handle multiple clients.
The engine has a hierarchical organization of data; at the topmost level the predefined general registry can contain multiple worksheets and instruments; these can contain an arbitrarily deeply nested hierarchy of parameter lists, each containing parameters and other parameter lists.
The engine has interfaces which stream in or out all of the data that has been specified by the client in the form of an indexed bit stream. This enables the client to save and restore cases in files. The index can be used by the client to view the structure of the data and compare files; it is used by the engine to recover as best it can from a corrupted bit stream. Copies of each index entry are included in the bit stream so that a client may attempt to recover from a corrupted index.
Parameter lists may be independently streamed in and out. This enables the client to maintain a library of templates, which are sets of parameters which can be instantiated into any case. Instruments may also be independently streamed in and out. This enables the client to maintain a library of instruments which can be instantiated into any case.
Each parameter is a fundamental calculation unit. It has a name by which it can be referred to by other parameters, a value, and possibly a means of calculating that value. Parameters can have badges identifying them to the internal database. Each badge is a list of “category=member” specifications, where the list of possible categories is defined by the client, though there are specific categories assumed by instruments. The tool uses categories like “cash classification”, “party” and “outcome” to model the flows of a financial model. Parameters can be designated as defining results. Each result is attached to a specific name, party and outcome. The result parameters can then be collated by the client into capsule summary reports, for example. The values manipulated by the parameters can have many types. Some are single quantities (scalars) representing numbers, logical values (true and false), dates, time intervals, frequencies (annual, monthly, etc.), character strings and enumerated types. Other values are arrays, plain, sorted or indexed. There are also some values which are neither scalars nor arrays, but become indexed arrays when referred to from a keyed parameter. These are used to represent income that will automatically be accrued, for example. There are special values to represent null, which is like an empty cell in a spreadsheet, and the results of calculation errors (e.g. divide-by-zero).
There are some built-in enumerated types (e.g. “calendar”), which enumerates the different ways of calculating the length of time between two dates, and the client can create its own using special list-definition parameters.
A plain array is a set of values indexed by the natural numbers (1, 2, 3, . . . ). Normally there is only a finite number of elements in the array, but limited support is provided for infinite arrays which are regular beyond a certain point. A sorted array is always in ascending order of its elements, with no duplicates; it is used for streams of dates representing events in a financial model. An indexed array, or stream, is an array which is attached to a sorted array for indexing purposes. This is used to represent streams of cash flows in a financial deal, where each flow is attached to a date. The elements of arrays can be other arrays, thus providing support for multi-dimensional arrays.
A keyed parameter is a parameter that has been connected to another parameter for the purposes of providing a key (or index) for its array value. Normally, the key parameter has a date stream (i.e. a sorted array of dates) as its value, and these represent the dates of the flows defined in the keyed parameter. When a keyed parameter refers to another parameter in a formula, it triggers special calculations to convert the keys. This is normally an “accumulate to date” algorithm, but can be changed to effect accrual, table lookup, interpolation and extrapolation by using formula prefixes.
The normal rules of arithmetic have been extended to handle all the different types of values, wherever possible. Thus a date and a time interval can be added to produce another date. A scalar can be added to an array (it is added to each element of the array). Two arrays can be added by adding corresponding elements. If these are indexed arrays, then the corresponding elements are found by matching the indexes (i.e. two streams of payments are added by adding the payments on the same dates).
Parameter values can be specified in several ways. The client could simply specify a value, or it could specify a way of calculating the value. A formula parameter has a formula specified, which is an algebraic combination of constants, references to other parameters, and functions. A copied parameter simply duplicates the value of another parameter. An extracted parameter is designed to extract data from the internal database. In that case, the client specifies a list of “category=member” specifications and the value is calculated by matching these requirements with the badges of all parameters, and adding up those which match. This enables the client to request, for example, all of a certain party's rental income, without knowing the details of the instruments in the model.
The formula language includes an extensive set of functions to provide spreadsheet capabilities for manipulating data. Also included are functions for manipulating dates and arrays. The formula language also includes a set of prefixes specified at the beginning of the formula. These prefixes affect the way that the parameter is handled in references by other parameters, and can trigger automatic accrual, table lookup, interpolation and extrapolation. The prefixes are also used to specify variability during optimization and to trigger search and repetitive calculations.
In addition to the built-in functions, the client can define custom functions. These are named objects created in a worksheet which can then be referred to from any worksheet in the same way that a parameter could be, except that the reference is followed by a list of arguments. The definition of the custom function specifies how the result of the function is to be calculated from its arguments. Arguments can be specified as mandatory or optional, with a default value.
There is a special formula syntax used to specify date streams and arrays of values. This uses keywords like “starting”, “ending”, “then”, “also” to make the specification of these types of values easier and more understandable. The engine can provide the client with detailed parsing information about formulas. This can be used to write formula wizards. The engine parses each formula into a tree of basic expression nodes. Each expression node handles a specific job like addition, multiplication, references to other parameters, function evaluation, etc. Parameters can refer to themselves in expressions with array values, provided that they use either the “previous” or “next” function to avoid a logical circularity. In this case, the engine duplicates the expression tree for each element in the array (normally for each date), so that the expression nodes can be evaluated without encountering circular reasoning. More generally, a set of parameters can form a self-referential group, triggering duplication for each parameter in the group.
Each expression node keeps its calculated value until it is invalidated by a client change. This “intelligent recalculation” minimizes unnecessary repetition of calculations. When a formula refers to another parameter, it does so by specifying the other parameter's name, possible qualified by the names of its worksheet or instrument and levels of the parameter list hierarchy. Qualifiers are other names preceding the parameter name, separated by periods. Many levels of qualification are possible, e.g. “Loan1.Calculations.Amortization.Interest”. Internally, the general registry has a name reference manager which maintains all these references. As parameters or parameter lists are created, destroyed and renamed by the client, the references get automatically updated. Unnecessary qualifiers are removed. Qualifiers are added if the original reference becomes ambiguous, thus maintaining the intended parameter linkages.
Each parameter has a dependency manager which handles the invalidation of expression nodes when the client changes a parameter. When the name reference manager resolves a reference, the target parameter sets up a link between the source's and target's dependency managers. If the source value changes, an event is triggered in its dependency manager which is transmitted via the link to the target; in turn this is passed on to the target's dependents. The dependency managers can also provide the client with lists of dependents and precedents for any parameter.
Dependency managers are also created for parameter lists, worksheets, instruments, and indeed the general registry. Changes are propagated up the parent chain so that each level knows when there has been some change inside them. This information can be tapped by clients to provide an intelligent refreshing mechanism; i.e. don't bother to redraw interfaces for objects which have not changed. The client requests a modification server for any level from parameter up to general registry. This modification server can be polled to determine whether there has been any change since the last poll.
Search parameters are formula parameters with a search prefix. There are three kinds: the optimization search, the targeting search and the maximization search. The optimization search only uses the first three arguments (low, high and accuracy) to the search prefix; it has no effect on calculations outside of the optimizer. The other types of search specify a target formula as the fourth argument. The targeting search specifies a value to target in the fifth argument, while the maximization search uses one of the keywords “maximum” or “minimum” as the fifth argument. Whenever a non-optimizing search parameter's value is requested, it iterates guesses for its value until the target formula is equal to the target value, or maximized, or minimized, depending on the fifth argument. Its value is saved until the dependency managers invalidate it, to avoid pointless recalculation.
The targeting search preferably uses third-party software which is designed to find the zeroes of functions. This software does a good job if the function is monotonic. However, it can get confused by non-monotonic functions which may have several solutions. The maximization search uses a custom algorithm which uses quadratic interpolation to refine the guesses. It depends on the function being fairly well-behaved as well. However, any suitable application can be used to perform this function.
The engine has a facility for collating multiple parameters into a single date-indexed table. It is called a parameter date table, and there is one in every worksheet (more are available on request). The client specifies which parameters it wants in the table, and the engine collates their dates and outputs a combined list of dates and a matrix of values. If the client wants to collate the data in regular intervals (e.g. annually), it can specify any number of date buckets; these override the table's normal “daily” rule.
The engine maintains a set of client-specified numeric formatting rules. These can be specified at any level of the data tree, from parameter up to general registry. The client can then request the effective formatting for any parameter, and use it to format numeric values using special engine calls. The facilities include comma insertion, fixed number of decimal places, prefixes and suffixes, percentages and scaling.
The internal database uses a bill-of-materials structure to enable parameters to collate data which has been identified via badges on parameters. Inside each category there are members which can be connected in a directed a cyclic graph structure, with numeric coefficients applied to each link. Thus a category representing the parties in a deal can establish ownership links between the parties, e.g. A owns 50% of B. When a parameter which has a badge specifying party B is extracted by a parameter requesting flows for party A, its values will automatically be multiplied by the 50% factor. The links are also used to establish rules on the cash classifications, like “after-tax cash” equals “pre-tax cash” plus “taxes”. The categories and members are completely arbitrary, and maintained by the client. An interface is provided to stream the entire table in or out, so that the client can save a default table in a file.
Parameters can extract data from the internal database either by designating them as extraction parameters, or by designating them as formula parameters and using one of the “collect” functions. Either way, the parameter specifies a list of category-member pairs and receives back from the database a list of parameters matching those specifications, with corresponding coefficients, and it then combines their values to get a value.
Parameter lists can be given an activation formula by the client. This is a formula that should evaluate to true or false. When the formula evaluates to false, the parameter list and all of its contents are labeled as inactive. Inactive parameters do not have values. If an active parameter tries to reference an inactive one, it will get an error value. Clients can use this facility to de-emphasize blocks of the model that are not currently being used, and prevent useless calculations from slowing down the program. There are several restrictions on the formula that can be used—for example, it must not refer to a parameter inside its own parameter list. Invalid formulas will always make the parameter list inactive. The engine can supply the reason why a parameter list is inactive.
The general registry contains a predefined worksheet called the timeline. This is designed to hold basic date and date stream definitions for the rest of the model, and has special interfaces to enable the client to manage them. The timeline acts as a preferred source of parameters to the name reference manager. If a reference cannot be resolved within the parameter's own worksheet or instrument, then the timeline is searched before going to any other worksheet or instrument. This gives a “global” nature to the timeline parameters.
The engine maintains lists of decisions and outcomes. Decisions are named objects attached to date-valued parameters. They are designed to represent points in a financial model from which the deal could proceed in different directions; e.g. depending on whether an option is exercised. Outcomes are named objects that are the result of saying whether each decision has been taken or not. The outcome in which no decisions have been taken always exists and is called the base outcome. Other outcomes can be created by the client by adding a decision to an already existing outcome. For example, if the client has created three decisions. then there are seven possible outcomes in addition to the base outcome obtained from the seven different ways of saying which combination of the three decisions has been exercised. These extra outcomes are not created automatically because the decisions may not correspond to independent decisions in real life—only certain combinations of decisions may be realistic.
Instruments are objects created by the client in the general registry. They are like worksheets in that they have an arbitrarily complex parameter list, but they also have data geared toward modeling financial instruments like loans, rent agreements, etc. They simplify designing badges for parameters.
Each instrument can have one or more parties specified. (The tool always uses precisely two parties for each instrument, corresponding to the two ends of the arrows in the party diagram.) The client can specify a role for each party. For example in an instrument modeling a loan, the two parties could be designated “lender” and “borrower”. In the parameters, the client may refer to the roles in formulas; they are automatically defined to take the value of the name of the party filling that role. For example, if there is a party called “MyBank” with a role of “lender”, then a formula may use the identifier “lender” with the same effect as specifying the character string “MyBank”. This allows instrument builders to write parameters which automatically follow party substitutions.
The general registry has a special worksheet called the parties worksheet. It corresponds with the parties chapter in the tool. It is completely maintained by the client. However, if an instrument party finds a section in the parties worksheet that has the same name as itself, then it generates role parameters echoing the data inside that section. For example, if the parties worksheet has a section called “MyBank” with a parameter called “FirstFiscalMonth”, then an instrument parameter could refer to its value as “Lender.FirstFiscalMonth”. (Assuming that the instrument contains a party with name “MyBank” and role “Lender”.)
Each instrument party can have one or more cash classifications, and each cash classification can have one or more income classifications. Each cash classification generates cash flows for that party by specifying the name of a parameter defined in the instrument, the section of the topmost parameter list in which to find the parameter, whether there is a sign change (for paying parties), and the name of the category member to be used to generate badges. Each income classification functions similarly with the extra information of which tax authority to badge it for.
Each instrument has a default decision handler which specifies how the flow-generating parameters are treated in outcomes involving a decision. The client can override this behavior for any actual decision that has been created. There are only two possibilities: either the decision is ignored, i.e. has no effect on cash flows, or the flows are truncated at the decision date. If they are truncated, the client can specify a formula for an extra amount to be assessed on the decision date. To generate the flows for a particular outcome, the earliest truncating decision is found which is in that outcome, and that controls the flow for that outcome. If there is no truncating decision for that parameter in that outcome, the flows are the same as the base outcome flows.
The client may specify for each instrument that it only generates flows for outcomes containing a certain required decision. This is designed to model purchase options which are only present if the corresponding decision is taken.
The client may instruct an instrument to generate its own decision and outcome for the purposes of calculating parameters and generating flows. This is designed to model termination values where a set of cash values needs to be calculated to terminate the deal at any of a fixed set of dates. The instrument contains a date-valued parameter which becomes the decision date; the client can vary this date to get a table of termination values, or have the engine vary it automatically by specifying formula parameters with the TerminateByDate prefix.
The object responsible for generating flows from instruments is the instrument synchronizer. Whenever the client changes some data, the synchronizer regenerates (as necessary) parameters defining the flows and identifies them to the internal database with badges. Within each classification within each party within each instrument, the synchronizer identifies the instrument parameter representing that flow according to the client inputs. For each outcome, it generates an auxiliary parameter which may change the sign of the flow and/or truncate it at the appropriate decision date. It badges the auxiliary parameters using the categories “party”, “other party”, “cash or income classification”, “outcome”, “instrument name” and “tax authority” for income classifications.
The client may designate that certain instruments are to be cloned for each outcome. This is designed to model tax payment instruments where the basic parameters have to take different values for each outcome. In this case, the instrument synchronizer clones the parameter lists and party sections for each non-base outcome before generating the auxiliary parameters. It sets up default collection parameters for each clone to ensure that parameters extracting data from the internal database pick the right outcome in each clone.
The general registry may also contain specialists. These are named objects designed to act like building blocks. Each specialist contains its own version of the general registry which can have worksheets and instruments and other specialists. Since they are in their own registry, there is no contact between them and the world outside; no danger of references being erroneously resolved. They are safer versions of the “template” concept. Special parameters are designated as inputs to and outputs from the specialist, and they do have interactions with the world outside the specialist.
An action is a named object representing a calculation that is too complicated for simple formulas to accomplish. There are predefined actions for building the optimization model, and executing the optimizer. The client can create new actions inside worksheets to do things like targeting and repetitive (sensitivity) calculations. Also the client can create action sequences, which are sequential series of actions, to create a primitive macro language.
Actions can be executed synchronously or asynchronously. In the latter case, the client starts off the action, and sits in a loop requesting action progress, until the progress report indicates that the action is complete. This enables the client to provide visual feedback to the user during actions which could take some time. For example, the client could display messages sent back from Cplex during optimization. Internally, actions are executed in a sub-thread to free the main thread to respond to progress requests.
The engine preferably has an interface with the Cplex optimizer. This is a solver, provided by a third-party, of linear (LP) and mixed integer (MIP) programming problems. It is encapsulated as an out-of-process COM server which can be run locally or remotely, with multiple users. The server can be started by the client or by the engine with the client specifying on which machine to initiate it. The set of instructions for the Cplex optimizer is generated by an optimizer model. This is a named object containing the data needed for the optimization. There can be more than one optimizer model in each case, if desired. An optimizer needs three kinds of information to operate. It must know which parameters the optimizer can vary, what constraints have to be satisfied, and what the objective function is. For each parameter with some information pertinent to an optimization model, there is a model parameter which houses this information. The client can request a model parameter from an optimization model for any given parameter in the system.
To specify that a parameter is variable in a given optimizer model, it is made an input parameter and the corresponding model parameter is given certain properties defining the variability, count (for array parameters), continuity (continuous, binary, integer), and any special ordering instructions for arrays (increasing, decreasing, SOS1, SOS2). Alternatively, the parameter can be made a formula parameter and the formula given an “optimize” prefix; this only works for the main optimizer model.
To specify constraints, the client requests constraint parameters from a model parameter. Constraints are parameters with some extra properties. The value of a constraint parameter represents the right-hand-side of a constraint, and could be a formula involving variable parameters. The left-hand-side of the constraint is the owning parameter, and the relation is specified by the client: greater-than, equals or less-than. The client can specify that adjacent constraints be combined in an OR-relation rather than the default AND-relation. Constraints have their own activation formulas; this enables the client to activate and de-activate constraints automatically. Constraints can be turned into assertions by setting their test-only switch. Assertions do not affect the optimization, but they can be queried by the client as to passing or failing just like constraints.
To specify the objective function to an optimizer model, the client requests a model parameter for each component of the objective function and sets its “objective function coefficient” property to the appropriate number; a positive number implies maximization, negative minimization. If there is more than one parameter with a coefficient, the optimizer will optimize the sum of the values multiplied by their respective coefficients.
The solving of an optimization model is done in phases. During the first phase, the formulas involved in constraints (and the objective function) are visited to determine the set of parameters that contribute (directly or indirectly) to constraints or the objective function. During the second phase, the definitions of all these parameters are visited to determine which are variable. The next few optimization phases are designed to determine array sizes and array element variabilities for those parameters with array values. This is done by generating an auxiliary set of parameters whose formulas are cloned from the original. The auxiliary parameters corresponding to parameters that the client has specified as variable are given special place-holder values which have a trivial pass-through arithmetic. When the other auxiliary parameters are evaluated, the result is normally a place-holder value, or an array of such. Wherever a place-holder value is encountered, that represents a variable (column) to be created in the optimization model.
Next, objects representing the basic variables (columns) of the model are created, and the coefficients, or rows, of the matrix are generated. Some rows correspond to the definitions of parameters, others to the constraints as supplied by the client. To generate the coefficients, a special type of value is used which is basically a linear combination of variables. Once an arithmetic of these values has been programmed, then they can be passed through the normal expression evaluator, and the rows created from the results. A second set of auxiliary parameters with cloned formulas is used to do this. Only client constraints that cannot be interpreted as bounds on the variables generate rows.
During coefficient generation, OR-groups of constraints, and some non-linear functions are linearized by creating extra binary variables to “take the decisions”. Functions that cannot be linearized generate non-linear errors.
After the model has been generated this far, a lot of information is available to the client such as linearity of the model, lists of constraints and optimizable parameters. The model is finalized by removing any egregious scaling problems (for example caused by the client defining some parameters in dollars and some in percents), and replacing some symbolic large and small constants (introduced with the extra binary variables) with real numbers estimated from other constants in the matrix.
The arrays that Cplex expects as inputs are created and the model transmitted to the Cplex server. The engine waits until that server has completed, forwarding any progress messages back to the client. The results are obtained from the Cplex server, and put back into the file wherever the client has specified that parameters are variable. Binding information and shadow prices are obtained from Cplex and stored in the corresponding constraints. To assist users in tracking down infeasibilities, only the values that the user has specified as optimizable are put back into the model; any dependent parameters and constraints are then recalculated from their formulas, and lists of failing constraints are available the client (user) to display.
If there are any optimization search parameters, then the model building and solving steps are repeated using different guesses for the search parameters until the best objective value is found. Optimization search parameters are formula parameters with a search prefix in which only the first three arguments have been specified (low, high, accuracy). This kind of search prefix has no effect outside of optimization. It is used by the client to solve for variable parameters which cannot be specified as variable to the optimization model without sending the model non-linear. The algorithm for choosing the guesses is a custom quadratic-fit algorithm for finding the maximum of a function.
For the purposes of generating the optimization model, any non-optimization search parameters are frozen at their current values. If one of these has a new value as a result of the optimization in such a way as to invalidate the optimization, then a message about this is generated and sent to the client, and a special status condition set.
There are two optimization actions provided for the client to execute. The “optimize” action will perform all of the above steps. The “build model” action performs only the first few steps, enough to get the client data on linearity, constraints, etc. This last action is performed by the Advantage optimization chapter to provide visual feedback while the data to display the chapter is being prepared.
The engine can export its model to Microsoft's EXCEL software, creating as far as possible a working spreadsheet model of all the parameter formulas. The exported model is limited in that it cannot handle date changes or changes in frequency or term that require re-dimensioning of arrays. It does a good job of formatting the resulting sheets to reflect the data hierarchy of the model. The client has control over the sheets and what is included in them.
The client can create other types of actions in worksheets. A marksman is a targeting action which works similarly to the search prefix. Being an action, however, it can be executed asynchronously with the client displaying progress. Each time round the loop, the variable parameter is guessed, and a specified list of actions performed, which could include optimization and other marksmen. This is repeated until a target variable reaches a specified value, or is maximized or minimized. A matrix is a repetitive action designed to vary a parameter over a range of values, perform a list of actions which could include optimization and marksmen at each iteration, and store a set of results. These results are collated into a two-dimensional matrix of values assigned to a result parameter. An action sequence is a set of actions which is executed serially; this could be used to implement a primitive macro language.
The engine preferably has a background thread which steals idle cycles to perform tasks in advance of the client requesting them. For example, evaluation of parameters, including searches, and generation of the optimization model. This can markedly improve interaction speed as the client may well find that whatever it needs to create a display has already been prepared by the engine. For example, the engine may submit background optimizations to Cplex so that should the user decide to press the optimize button, he would get an instantaneous result.
As explained above, the instant invention includes a powerful formula language which can be used in the worksheet, as well as in other chapters of the invention, to provide scenario information. This formula language includes a library of predefined functions and keywords which can be used by the user when using the tool. An exemplary set of these functions and keywords which comprise the formula language, together with an associated syntax, is provided below.
Returns the absolute value of an argument.
Number is any numeric value or expression that is a real number.
Tells the application to recognize values as amounts to be allocated uniformly to each day in their time periods.
Accrues Advance or Arrears payments into a set of user-defined intervals.
Accumulates values from an indexed array into the overlapping periods under a different index.
An activation switch is a formula that evaluates to True or False (e.g. x=y) in regard to a heading or a constraint.
An attempt to reference an inactive parameter returns an error unless:
Makes a parameter mimic the behavior defined by the formula for another parameter.
This is one of four calendar methods you can choose to evaluate intervals between dates for the purposes of calculation.
This is one of four calendar methods you can choose to evaluate intervals between dates for the purposes of calculation.
AddYears returns a date based on the calculated interval between a given date and given number of years, using the day-counting metrics of the current calendar.
Defines a payment stream in which each payment applies to the period following the date on which the payment occurs.
Defines the first value in a date stream or maps the first value in a sorted numbers stream.
Alignment keywords anchor the anniversaries in continuing dates to a month/day or the same day each month at a given frequency. There are two ways to express alignment:
Align frequency mm/dd
In this expression, you anchor continuing dates (at the given frequency) to the relative date mm/dd rather than the first date defined in the formula.
Align Monthly On dd
In this expression, you anchor monthly continuing dates to a day of the month other than the starting date.
Use the Also keyword at the end of a date stream formula to insert a date that will not otherwise occur according to the preceding formula.
An anchor date is the point of departure in the calculation of an anniversary or other fixed date.
AND returns a logical value that asserts if a set of conditions are true or false.
An anniversary is a date that occurs on the basis of a given anchor date and frequency.
Returns the last anniversary of an anchor date up to (and including) a cutoff date. Anniversaries occur at the specified frequency beginning on the anchor date.
Annual (or Annually) is a frequency keyword used to define the interval between continuing dates in a date stream.
Returns the values in a source array that correspond to a given index position.
An array is a type of parameter that contains a series of one or more values. The values are arranged according to a formula. There are three types of arrays:
ArraySum returns the total of all the values in an array.
The Arrears prefix means that each payment applies to the period that precedes the date on which the payment is incurred.
A badge is an internal ID that associates certain data to a list member within a category (“category=list member”). For example:
Defines the last date in a date stream or defines the date on which the last value in a numbers stream occurs.
A bookend date is a date index element that does not represent the beginning or end of a period. The period prefix used in the date index determines the bookend date.
The tool relies on a user-specified calendar method to evaluate intervals between dates for the purposes of calculations such as:
The active calendar method in a calculation affects payment values because there are more or fewer days in a given interval based on the calendar.
Multiple Calendars in the Same Case
You can specify different calendar methods for different requirements in the same case. If a case has more than one calendar method, the application uses a decision tree to determine the method that applies to a particular calculation.
Decision Tree for Calendar Selection
When a case contains multiple calendar methods, the application applies the closest method it can find based on this order:
Cash is the default prefix for a payment stream when neither the Advance prefix nor the Arrears prefix is given. Use the Cash prefix (or omit a prefix) to specify that each payment applies to the date on which it is incurred.
Categories and List Member Reference
A badge is an internal ID that that comprises a category and one of the list members within that category. Certain functions use the list member portion of a badge as an argument to collect or extract data.
Returns a number rounded up to the nearest multiple you specify.
Returns a character value from the ANSI code you specify.
Returns a value from a list of arguments based on an index number.
A classification is a financial label. It tells the application how to retrieve values given a context such as instrument, outcome, party/role, and so on.
Guidelines for Working with Classifications
There are three ways to see a list of payment (income/cash) classification members.
In Smart Paper
Returns an ANSI code that specifies the first character in a text string.
Collects data from parameters that match up to one or more badges.
Collects income from parameters that match a set of one or more badges. Each argument position represents a category and takes a user-specified list member name.
Collects payments from parameters that match a set of one or more badges. Each argument represents a category and takes a user-specified list member name.
Links multiple text strings into one string.
A constraint/assertion is a formula you can add to a parameter along with the formula that generates the parameter values.
Returns the full pathname (i.e. SheetName.Heading.SubHeading.Parameter) to the given reference.
Counts the number of values in an array.
Returns an array in which each value is the accumulated value of all preceding values in the argued array.
Returns the daily present value of a set of cash flows.
Returns a date value from numeric values that represent a year, month, and day.
Returns a date value in format “dd mon, yyyy” (including the quotation marks).
A date expression is a date included in a formula. The date can be given in any of the following formats (refer to example):
A date index is an array parameter that contains an ascending series of dates. The dates are used to coordinate series of values according to when they occur over time.
Returns a date stream consisting of dates used to key the argued numbers stream(s) without regard to the StartDates or EndDates prefix designation of the source.
Converts a text string representation of a date to a date value.
Returns an integer value (between 1 and 31) for the day of the month in a date value.
Returns the number of days between two dates on 360-day calendar (twelve 30-day months).
Returns the number of days between two dates based on a 365-day year.
Returns the number of days in period according to specified calendar.
Returns the number of days in a calendar year.
A destination parameter is one in which the formula refers to another parameter in order to use the referenced parameter as a source of data.
Returns the difference between each successive value in an array or series of values.
Returns an elapsed time value in whole numbers for years, months, and days.
Indicates that each date on the index (except the first date) represents the LAST day of a period.
Returns the combined dates used to organize a set of number streams, first converting them to end dates if necessary.
Derives period start dates from one or more numbers streams and/or date indexes, then creates a date index of dates that reflect the end dates of the underlying periods. Also incorporates dates or date streams given as arguments.
Defines the last date in a date stream.
European—30 —360 Function
This is one of four calendar methods you can choose to evaluate intervals between dates for the purposes of calculation
Returns a number rounded away from zero to the next even integer.
Allows for a break in a sequence of values being mapped to an index.
Returns the natural logarithm base e (2.71828182845904) raised to the power you specify.
Returns the logical value False.
Finds and returns the starting position of a text string within another text string.
Locates a value given its context within the case.
Refers to the first value in a given array.
Refer to the first date in a date stream in order to use the date itself in another date stream.
Returns a number rounded down to the nearest multiple you specify.
Use the For keyword to repeat a numbers stream value or extend a date stream sequence.
Use the Forever keyword in date stream in lieu of an ending date phrase if you want to specify that the stream never ends.
GetResult (Result, Party, Outcome, Context, SubsidiaryParameter)
Locates a SetResult parameter flagged with a matching set of arguments and returns the parameter value. Or, if optional arguments are included, the function can return a subsidiary parameter value.
Returns one of two values depending on the results of a logical test for a true or false condition.
The application can return text if it is encased in quote marks.
An index is a type of parameter used to coordinate values.
Returns a value from any type of array based on the relative position of the value within the array.
Returns a number rounded down to the next integer of lesser value.
Tells the application to interpret each of its values as a lookup value that is linearly applied across its time period, creating a stepped value for each day in the time period
An interval is the length of a date index period calculated as a portion of a year.
Calculates the numbers of years between two dates based on the day-counting metrics of the given calendar.
A key is a point of coordination on an index that is used to sort values.
Returns the index used to coordinate the values in a numbers stream.
Refers to the last value in a given array.
Refer to the last date in a date stream in order to use the date itself in another date stream.
Returns one or more of the leftmost characters of a text string.
Counts the number of characters in a text string, including spaces.
Returns the natural logarithm of the number you specify.
Returns the logarithm of a number to a specified base.
Returns the logarithm of a number to base 10.
Returns the occurrence of values shared between two parameters keyed to different indexes.
Converts a text string to lower case.
Returns a scalar value or array of values, depending on the number of arguments given and if the Max parameter is indexed.
Returns array in which each position holds the maximum value found at the corresponding position within a set of arguments.
Returns the single maximum value found in a set of arguments.
Extracts one or more characters from within a string.
Returns a scalar value or array of values, depending on the number of arguments given and if the Min parameter is indexed.
Returns array in which each position is the minimum value found at the corresponding position within a set of arguments.
Returns the single minimum value found in a set of arguments.
Returns the remainder of a number after a division.
Returns an integer between 1 and 12 that represents the month in a date value.
Returns the last calendar day of the month for a given date scalar or stream.
Monthly is a frequency keyword used to define the interval between continuing dates in a date stream.
Monthly_MISF_Yield (Flows, SinkingFundRate)
Returns the nominal monthly MISF yield of a given array of cash flows.
Returns an array in which, relative to the current position, each value refers to the next sequential value in a source array.
Lets you specify that a parameter is not keyed to an index, even if the parameter appears to be within the scope of an index.
Reverses the logical truth or falseness of the argument.
Returns a number rounded up to the next odd integer away from zero.
Assigns a given value to the given date, or (with the Except keyword) lets you exclude a value on the given date.
Tells the application to calculate values to meet the objective function of a case. An example of an Objective Function is the total cash of a transaction.
Returns the logical value True if any of the arguments are true. If all of the arguments are false, returns False.
A parameter is data that represents a calculation and can be used in other calculations.
Returns the first or last day in a period. You can specify whether you want the day to be in the period, or before or after the period.
Returns the end date of a period defined by a set of bookend dates.
Returns a period stream based on the intersecting dates of two other period streams.
Maps a numbers stream into a set of user-defined intervals.
Returns the portion in years of a period according to specified calendar.
Returns a stream of periods to which values in the given numbers stream are keyed.
Returns the start date of a period defined by a set of bookend dates.
Creates a period stream from a pair of bookend dates and an existing date stream.
Returns a number raised to the power you specify.
A prefix is a keyword you use at the beginning of a formula. The prefix applies to all the values in the formula and determines how the values apply to time periods.
Previous (Array, FirstValue)
Returns an array in which, relative to the current position, each value refers to the preceding value in a source array.
Quarterly is a frequency keyword to define the interval between continuing dates in a date stream.
RegularAscendingArray (Start, Step, Number)
Creates an ascending array of in which each value increases by a specified increment.
Creates a date stream or date index.
Replaces a portion of a text string with another string.
Repeats a text string the number of times you specify.
Returns one or more rightmost characters of a text string.
A role is the name for the set of properties that defines Party interaction with an Instrument.
Returns a number rounded to the number of digits you specify.
Returns a number rounded down to the number of digits you specify.
Returns a number rounded up (away from zero) to the number of digits you specify.
A scalar parameter defines a single quantity that does not change over time or other direction. A scalar parameter is not attached to a key position (such as a date) on an index, even if the scalar parameter happens to be located in the scope of an index.
The Search prefix tells the application to perform search and repetitive calculations to determine the value of a scalar parameter.
There are three ways to use the Search prefix:
In an optimization search, you set up the Search parameter formula, then you perform an optimization with an objective function.
Searches for a value for which the target expression equals the target value. For example:
Semiannual (or Semiannually) is a frequency keyword used to define the interval between continuing dates in a date stream.
Semicolon (;) Keyword
The semicolon symbol (;) stops the current sequence of dates or values in a stream. The stream then continues or ends according what follows the semicolon in the formula.
The SetResult prefix lets you define a three-argument identity of sorts to bookmark a value. You can then retrieve the value, when needed, with the Get Result function.
Returns the sign of a number. If the number is positive, the application returns 1; if the number is 0, the application returns 0; if the number is negative, the application returns −1.
A source parameter is one that contains data you want to use in another parameter.
Returns the positive square root of a number.
Indicates that each date on the index (except the last date) represents the FIRST day of a period.
Returns the combined dates used to organize a set of number streams, first converting them to start dates if necessary.
Derives period end dates for one or more numbers streams and/or date indexes, then creates a date index of dates that reflect the start dates of the underlying periods. Also incorporates dates or date streams given as arguments.
Defines the first date in a date stream or assigns the first value in a numbers stream to the given date.
SubArray (Array, Start,Count)
Returns a portion of an existing array as of a specified key position.
SUBSTITUTE (String, OldString, NewString, InstanceNum)
Substitutes any number of instances of a text string with another text string.
Returns the total of all values in one or more indexed parameters of the same name under a given heading.
Returns the sum of all values in all of the arguments.
Returns a single number that is the sum of an array up to, but not including, the specified date.
The Table prefix tells the application to interpret each value in the parameter as a simple lookup value along a date index.
When you reference the Table parameter in the formula to calculate values for a destination parameter, the application applies the table value in effect as of the Table parameter period that contains the destination parameter date.
Extracts a value or values from a table.
Stops the current sequence of continuing dates in a date stream. The stream re-continues or ends according what follows in the formula.
Assigns the given value to each remaining period on the current index according to the payment prefix, unless:
Returns a date stream using given dates and dates from another date stream.
Changes multiple blank spaces between words to a single space.
Returns the logical value True.
Converts a number to an integer by removing the fractional part of the number. The application rounds the number down to the next integer toward zero.
Returns a numbers stream stops as of the index date that precedes the given truncation date.
Combines all values found within the given arguments and returns an ascending array with duplicates removed.
Assigns the last value in a sorted numbers stream to the key position that contains the given date.
Unsorted Array (Definition)
An unsorted array is a type of array parameter that comprises a series of independent values. That is, the values are not keyed to an index (even if the unsorted array appears below an index).
Converts a text string to upper case.
This is one of four calendar methods you can choose to evaluate intervals between dates for the purposes of calculation
If the first argument is active, the function returns the value of the first argument. If first argument is inactive, the function returns the second argument.
Determines the year for a given date.
As is readily apparent from the description of the invention above, the instant financial modeling and analysis tool provides a user friendly, effective and efficient tool for modeling financial or other mathematical scenarios of almost any kind. The graphical user interface combined with the powerful engine provide a greatly improved modeling tool as compared to the prior art.
It is noted that the invention is not limited to modeling financial scenarios or deals, but, instead, can be used to model any scenario involving mathematical values over time.
The implementations described above illustrate the characteristics, features and advantages of the present invention. These implementations, of course, are not exhaustive, and other implementations within the scope and spirit of the present invention will be apparent to those skilled in the art. In other words, while the invention has been described in connection with what is presently considered to be the most practical and preferred embodiments, it is to be understood that the invention is not to be limited to the disclosed embodiments, but on the contrary, is intended to cover various modifications and equivalent arrangements which fall within the true spirit and scope of the appended claims.