If you want a reliable VAR calculation tool...wait for the next version!

By Andrew Webb

FourFifteen is an Excel-based VAR calculator and report generator produced by JP Morgan and named after the bank's own internal risk report, which is produced daily at 4:15 pm. It uses volatility and correlation data available from JP Morgan's Web site for its calculations.

The product originated in response to a gap that JP Morgan had observed in the market. There had been a number of systems available for some time that were potentially suitable for the larger banks. At the other end of the market, there were some very simple spreadsheet solutions. But the middle ground for smaller banks and corporates remained empty.

The bank had also been receiving many requests from clients to provide a low-cost, user-friendly application that would make use of its established RiskMetrics methodology. While FourFifteen provides BIS standard reporting, the development team decided to cast a wider net and produce something applicable to corporate treasurers, fund managers and insurance companies as well. To improve ease of use, a dedicated Web site was established specifically for FourFifteen users that would supply the necessary correlation and volatility data in its own specific format.

Version 1.0 of the product was launched in early 1996 in both Windows and Macintosh formats. Version 1.1, which included significant enhancements, was released in October 1996. Among these were support for options (which had been a significant hole in Version 1.0) and a considerable beefing up of the reporting functions (both standard and customized). At the same time, a new regulatory data set was made available from the FourFifteen Web page to speed up the process for clients to produce VAR figures that complied with BIS.

Product Structure and Principal Features

The main FourFifteen screen is started by opening the file "415.xls" in Microsoft Excel version 5.0c. The product literature states "Excel 5.0c or higher," but I quickly learned you can forget the "or higher"-Excel 7.0 didn't want to know. A string of interesting error messages was the only result; I learned later that an Excel 7.0 version was not yet available, but is likely sometime in the first quarter of 1997.

The only machine I could find with Excel 5.0c happened to be a 486DX66 laptop, which did actually meet the spec required (according to the manual) to allow FourFifteen to run "effectively." Most of the actual work in FourFifteen is done by three Excel add-in files located in the "System" sub directory.

The spreadsheet offers three types of menu setup-beginner, advanced user and power user. The assumed default is advanced user, which configures the screen so that only the FourFifteen menu is available. Beginner uses a Microsoft Wizard to take users step by step through the basics of producing a VAR report (only one menu command), while power user gives users the same menu options as advanced, but also leaves the standard Excel menu bar available at the same time.

Setting up VAR parameters is straightforward enough. A single dialog box allows users to configure time horizon, base currency, the option to use input prices or yield curve valuation, and risk threshold (as a percentage). A group of radio buttons and check boxes are used to select the appropriate data set and whether to correlate across asset classes, and offers the option to use BIS parameters. If users select this last option, all the relevant parameters in the rest of the dialog box are automatically set.

FourFifteen offers four separate methods of building portfolios for it to analyze-manual, generic, database and import. The first two are broadly similar in approach. The Manual method requires users to enter full details (up to 15 entries in some cases) for each security, while the generic method (only available for some types of instruments) cuts the number of entries required in half by automatically using standard market data to supplement entries. The database method allows users to select the more popular securities from a pre-installed list. It's only applicable for bonds, strips, bond futures and EuroCurrency futures. Frankly, none of these methods have much appeal if users have a large or rapidly changing portfolio-the man-hours required to ensure that the portfolio is accurate will be colossal. I did actually build my own sample portfolio to check this, and discovered that once I strayed from the most straightforward securities (options are particularly bad news), life got quite slow indeed.

The import method of portfolio building appears to be the best alternative, as it claims to allow users to import portfolio files from external position management systems. Unfortunately, the manual only allows this method four lines of text. It states that further information on the structure of FourFifteen portfolio files is available in a separate publication. Infuriatingly, it also proved to be the one FourFifteen publication not available on the web site, so I was unable to check just how realistic a proposition it was. (JP Morgan has promised to correct this.)

The actual reporting functions are reasonably comprehensive. Apart from the FourFifteen report itself, which gives a simple two-sheet outline of multiple-portfolio VAR and VAR against a benchmark portfolio, a number of other standard VAR and PV component reports are available. The default set supplied covers the 12 or so most popular requirements, such as VAR by instrument type, maturity or currency. Accompanying graphics are also available.

The Custom reports dialog allows you quite a bit of flexibility in your report design. Apart from allowing users to assign up to six different categories (c-party group and rating, currency, maturity, risk type and instrument), users can further customize these individual categories by regrouping their components into buckets (maximum of eight). As with the standard reports, reporting values can be expressed in a base currency or as a percentage of the present value of the portfolio.

Three additional reports are available-correlation, market rates and position sensitivity. The correlation and market rates reports allow you to extract and view a selection of data extracted from the data sets collected from the Web page. I found the correlation report unstable; it caused Excel (and Windows) to crash, when I attempted to run it. The position sensitivity report was one of the best features of FourFifteen. Users can modify a portfolio by adding or deleting instruments and then calculating the revised VAR or PV values to assess the impact of the changes. If desired, the modified portfolio can then be saved separately, while the original portfolio remains unaffected.

My view

In attempting to fill the middle ground, I think FourFifteen has instead fallen between two stools. Granted, there isn't much in the way of head-to-head competition out there, but there will be soon. $25,000 per year ($35,000 in the first year, which includes setup and training) for three add-ins and a spreadsheet that will currently only run on an obsolete version of Excel is not a bargain. It's also not particularly stable; apart from the problem with the correlation report, there were intermittent run-time errors as well. They weren't fatal, but they were irritating as they terminated the current task.

Data entry will also be a major headache with FourFifteen, even if users succeed in setting up the conversion and import of portfolio files from other applications. I'm sure there are probably some corporate users around with simple and small enough portfolios for FourFifteen to handle effectively, but whether they will feel inclined to pay the price is another matter.

There are also some significant holes in the instrument coverage. While the options situation is much improved over Version 1.0, it's Euros only-American exercise is not available. The recommended hardware specification for FourFifteen to run "effectively" is also on the optimistic side. Despite having the requisite amount of system RAM (16MB) on the test machine, it was evident from the turgid work rate and grinding hard drive that the swap file was still taking a fair hammering.

It's not all bad news-some features are impressive. Apart from the sensitivity analysis already mentioned, the new delta-gamma methodology for measuring VAR on portfolios containing options is worthy of note. As a pragmatic solution to the problem caused by the skewed distribution of returns of options, whilst avoiding the computing overhead of a full simulation, it is quite outstanding. (A full treatment of the method by Peter Zangari is available from JP Morgan's Web site). The FourFifteen documentation was also exceptional; clear and superbly laid out, with a good general introduction to the subject of VAR as well. Technical support was helpful and enthusiastic.

I think that JP Morgan needs either to move the decimal point in the price tag one place to the left or do some serious building work on this product, particularly with respect to database entry and maintenance. As yet, it's not in the right place on the price/functionality curve.

Peter Vinella
President, Peter Vinella & Associates International Inc.

Before I get too far into this review, I should come straight out and admit that I am not overly fond of software produced by broker/dealers, especially when it is intended to be used by their customers. In fact, I hate it.

Software is a difficult business at best, and most broker/dealers just won't make the necessary commitment over the long haul to be successful. During the past few years, my firm has been engaged in dozens of software evaluations for major global financial institutions. Based on this experience, we have discovered that in the case of most mission-critical applications, the questionable viability of the vendor-created product far outweighs all the other decision factors, such as functionality or ease of use. Even in the case where there is only one contender with a strong functionality set, given a weak vendor, most firms would prefer to develop the application in house.

I have managed customer research/software businesses at two major U.S. broker/dealers. In both cases, the service was discontinued, not because of a lack of success. On the contrary, they were both well-received-more than 250 clients in one case. They were discontinued because senior management didn't consider them critical components of their revenues or operations. Despite their success (and revenues), management wanted to focus on trading and not software.

What about FourFifteen? Is it a viable product, one that can overcome my prejudices and low regard? In a word, no. I am truly sorry that they lived up (down) to my expectations. The people at JP Morgan are extremely nice and attentive. They seem to care about the product and appreciate its potential value to their customers. The system itself is targeted-in theory, at least-at a market that has been generally ignored by the major risk-management systems vendors. I even began to feel somewhat guilty about the general negative bent of the review.

My guilt was very short-lived, however, once I learned that FourFifteen was not free, but commanded a $25,000 annual license fee plus a nominal $10,000 for one day of setup and training. Let's assume that the $25,000 doesn't bother some readers, and they are still interested in FourFifteen. What about the product?

Before you can run the application, users will need some basic desktop technology. They will need an IBM compatible PC (486 or higher), Windows 3.11 and Microsoft Excel version 5.0c-although the documentation leads you to believe that you can run it on version 5.0c or higher. We tried running the software under Windows 95 and Excel 7.0 on one of our "nitro-fueled" 200-mhz Pentiums, but no luck. Finally, after cranking up a Toshiba 486 notebook still running Windows 3.11, we got it to run, sans beginner mode, but it was slow. It took three more days of system tuning to make the application ready to run.

I spent a great deal of time reviewing FourFifteen's documentation. This proved to me that this is not commercially produced software. The documentation is superb. It features a two-color, glossy cover and lacquered pages, is quite detailed, offers loads of formulas and graphics, and is professionally laid out. This cost a lot of money-I only wish that they would have included a technical reference as well as all the risk management theory. On page 1 of the manual, it states almost proudly that FourFifteen is not part of JP Morgan's internal risk management systems. It is new software that is merely named after an internal report used at the bank.

From a risk theory point of view, the system is indeed very consistent with the RiskMetrics methodology. VAR is computed via cash-flow exposures to parameterized volatilities and a simple covariance matrix of currencies and indices. The risk matrix is computed by JP Morgan and downloaded via the Internet - neat! Although JP Morgan claims that this is a multifactor approach in the vein of APT models, it is actually an application of a modified Markowitz covariance matrix that is based on the returns of indices rather than those of individual securities and commodities. In fact, the indices are so linearly dependent that I believe the model vastly overstates the number of explanatory variables. Moreover, this approach is not unique and was originally implemented by BARRA in the mid 1980's.

The RiskMetrics approach has many fans-it is a widely accepted standard and conforms to the spirit of the 1995 BIS/Basel accord. So what exactly does FourFifteen do? It prints reports. Users load in their portfolio, configure the application, download the requisite-and free-data sets and out come their reports.

Of course, there isn't a true database, so historical or as-of analyses of your risk/performance is impossible. I wonder what users would do if they needed to rerun a risk analysis that accounts for as-of corrections, fails and data-entry mistakes? The lack of a true database also makes maintaining the system painfully difficult, and users can choose only the RiskMetrics data set and nothing else. Woe betide the person who wishes to compute a Monte Carlo-based VAR or employ volatilities, prices, rates or yield curves.

One other small point. Users first have to get their portfolios into FourFifteen and then maintain the positions and security/deal cash flows. To be fair, users can write an interface from an external system into FourFifteen and load/maintain the positions semi-automatically. Unfortunately, this is not an intelligent interface via a true API. This, in turn, necessitates the need for extensive manual reconciliation and intervention. If users trade quite a bit, or if they run a large portfolio, especially one with a lot of private placements, OTC or synthetic securities, this product is not optimal.

I should point out that this is a major problem with most risk management systems. They assume the data is readily available, correct, timely, consistent and complete on some existing legacy system somewhere in the firm. At least three-quarters of a risk management system involves complex data management.

Another major deficiency in the system is the lack of a canned utility that can easily export data and reports from FourFifteen into other systems. Probably the biggest drawback of the system from a risk management standpoint is the lack of counterparty and status information associated with a position or cash-flow. For instance, it is impossible to define a position as traded, but not settled. All positions are considered "in the box." In addition, reports can only reflect trade-date positions. This makes the system unusable for broker/dealers who must perform risk management on both a trade- and settlement-date basis.

In terms of technology, the product is well-conceived and nicely implemented, but is naive and lacks requisite automated administration tools and interfaces. Functionally, the system is a very limited implementation of the RiskMetrics risk management methodology. It lacks the flexibility and depth to be useful for even mid-sized institutions outside of generic, vanilla risk reporting. As a product, it offers nothing truly unique from other existing, more mature packages. At $25,000 per year, it is easily one of the most costly spreadsheet solutions, even if they threw in the training. This is especially true when one considers the actual cost of ownership, which includes the necessary labor. Without JP Morgan's financial backing and market presence, FourFifteen, in its current state, could not stand on its own as a viable commercial product.

Four Fifteen

Contact: John Sarapata
Phone: (212) 648-3759
Fax: (212) 648-5373

Pricing: $25,000 per year ($35,000 for first year includes setup and staff training)

Since this review was written, JP Morgan has announced a joint agreement with The Mathworks to develop FourFifteen further. A new version of the product, based on the MATLAB 5 development platform, will be available later this year. JP Morgan will concentrate on the risk methodology part of the project while The Mathworks will focus on the software development. The use of MATLAB 5 is designed to allow clients to rapidly develop and add their own customized modules to FourFifteen. Apart from enhancing the existing FourFifteen risk analytics, it is anticipated that the new system will include an improved database interface.

By Andrew Webb

Cash management is a new module within Renaissance's Opus suite that provides an infrastructure for the viewing, reporting, settlement and management of cash flow within a trading environment.

History and Background

The Opus Cash Management Module has its earliest origins in the experiences of Sean Togher, Renaissance's director of product strategy and customer support, who worked for two major trading banks in the 1980's. He and other Renaissance developers noted that there were no systems on the market that provided a "catch all" solution for the comprehensive flow monitoring of cash items both internal and external to the business. In some cases, the substantial funding cost of certain traders' transactions were completely detached from the actual trades. Management therefore lacked an accurate picture of the true profitability of individual traders and trading teams.

The structural plan for the cash management module was consequently designed with the core idea of ensuring that everything would be swept into it. (Togher points out that it's always the oddball transaction that gets omitted that inevitably comes back and bites you later when you're not looking.) Because so much of the structure was already clear in his mind, the timeframe (for what was by any standards a complex development project) has been tight. The product spec was defined by the last quarter of 1995 and has taken slightly more than nine months from start to finish.

Product Structure and Principal Features

The cash management module, like the rest of the Renaissance Opus suite, currently runs on UNIX (though there are plans for a Windows NT version in the pipeline.) It makes extensive use of drag and drop between itself and other Opus applications. For example, you can enter a trade into another Opus module and then drag it into the cash management module window of your choice for immediate processing. ("Immediate" is the apposite word; the demonstration had to be switched at short notice from a Sun UltraSparc server to a rather antiquated SparcStation 10. Despite this handicap, and the substantial size of the demonstration portfolios, processing speeds were actually quite respectable. On the right hardware the cash management module should fly.)

The SQL-based Opus Organizer allows the user to quickly drill down (and sideways) from any single item or group in the cash management module to any other Opus module in order to examine transaction details and linkages. A nice GUI touch here is that windows expand as you drill down rather than just proliferate and clutter the screen.

The accounts department will be happy. Apart from straight cash management, reporting and analysis are also available for P & L as well as accounting and clearing activities. Apart from more than 10 standard reports, users can quickly create their own customized ones using the flexible drag-and-drop capabilities of the organizer.

This flexibility is a general feature throughout the cash management module, from choosing the level of data granularity to the journal voucher facility, which allows for the manual entry of miscellaneous cash flows. Because this allows for the creation of a detailed component cash flow database, it is a straightforward matter for a cash flow to be tied to its originating deal for an audit trail. Another flexible feature is the way that the user can define the trade status level that will trigger the production of a cash item into the system. Despite this level of flexibility, the cash management module has an effective error trapping routine that blocks the threat of data duplication. It was demonstrated effectively (and on one occasion inadvertently!) during the presentation.

Counterparties can be defined in the cash management module as internal as well as external, so intra-company trades from various books can be handled. Counterparty banking details, such as branch, SWIFT number and so on can also be entered and linked in to the cash management system for disbursements and receivables.

In order to facilitate global enterprise and risk management, a considerable amount of effort has gone into providing an API that allows the collection of data from external non-Opus systems. (Data can be received in either object-orientated or ASCII files.) This does more than just ensure that users are not beset with loose ends. When combined with the modules ability to decompose individual transactions into separate cash items, it will be of particular benefit to risk managers striving for that ever-elusive "global picture."

My view

While cash management may hardly be a topic to set the pulse racing, it is something that has gained enormously in importance in the last few years. The general narrowing of spreads in the industry has meant that efficient resource management has had to start to become a profit center in its own right. Couple this with the regulatory and business requirements of risk management and you begin to appreciate the timeliness of this product launch.

I have to admit to being seriously impressed by what is clearly a well designed and implemented application. Functionality is all there, as is flexibility. But above all else, I think its scope is the most significant point. A great deal of thought has clearly gone into ensuring that the cash management module can cope with any type of cash item, whatever its source. (The effort expended on the handling of cashflows from external systems alone bears witness to that.) The rigor of this approach has spawned an outstanding product.

The release of the Opus cash management module is also a tidy tactical move on Renaissance's part. Not only does it have the potential to enhance the effectiveness of some of the other modules in the existing Opus line up (the Risk Manager in particular), but it also puts down something of a strategic marker for Renaissance.

While the company has striven for some time to be seen as a front-to-back-office solution provider, the fact remains that many people still think of it as a provider of smart pricing tools. The cash management module could be the watershed where that all changes.

Peter G. Kane
EDS Global Securities Industry Group

At a time when business moguls and operating managers alike declare that "cash is king," cash management and risk management solutions have become de rigueur. This message has not been lost on Renaissance Software Inc. as it recently unveiled a new cash management module to the successful Opus product. The cash management functions build on the strong infrastructure established with Opus and extend the process to deliver a rich and intuitive suite of position, settlement and simulation results.

The module breaks down transactions into individual "cash items" and introduces a more flexibly defined "book" for linking common elements. The book grouping presents all settled cashflows derived from an instrument, including various fees, into a cohesive view. An enhancement that illustrates the thoroughness of the design allows input of non-standard settlement instructions at both the book and transaction levels. This feature is vital to multicurrency account clearing operations. A further refinement is the user definable option for both a base currency and funding currency for each account.

Counterparty cash-flow management is enhanced by the definition of counterparties as internal (that is, branches, subsidiaries) or external. In addition, the counterparty cashflow screen displays the entity for those counterparties that have multiple booking entities. The importance of maintaining the correct counterparty booking entities is essential to understanding an organization's exposure to legal and tax accounting issues particular to special purpose vehicles and other structured entities. Once the counterparty information is set up, a screen display and report can be generated for all accounts under that counterparty organized by ABA number, SWIFT number, CHIPS number and so on.

A unique feature is the ability for user-definable rules that define when cash items are generated based on the trade status. A user or systems administrator with the proper security can set the table to generate cash flows when a transaction is moved into a trade status of "booked" or "confirmed" depending on the accounting rules and revenue recognition preferences of the organization.

The Opus module includes the ability to ignore "implied" cash flows of outlying floating legs created from the yield curve and merely display the calculated or "known" cash flows depending on the user preference.

NOSTRO management is included in the cash management module. NOSTRO account balances are displayed based on the netting effect as required by the International Swaps and Derivatives Association (ISDA) rules.

Bowing to the accountants, the module allows a user to redefine a cash item as a suspense item. Once the definition is changed, a new journal voucher is automatically created for the item and the updated posting is made to the activity and balance positions.

Funding

A flexible funding application in Opus presents funding positions by either book or NOSTRO, depending on how the organization manages its obligations. If funding is done by NOSTRO, then a money market loan or deposit trade ticket for the suggested funding amount is generated. If funding is managed by book, a suggested trade is also generated. If the book currency and the funding currencies are different, Opus creates the F/X trade ticket and also generates the money market side of the funding deal for reconciliation.

For those accounts that require a minimum balance, the application incorporates this information into the suggested funding transactions.

Expanding Window Views

Rather than relying on opening multiple windows to display transaction detail, Opus uses an expanding window to add views to the existing window. This has the effect of maintaining the association and context of information. For example, a view of the account balance window displays the various balance line items that make up the total balance. By double clicking the mouse on the balance line item, the detail for that item is added to the expanded display window. Going the next step, a double click on the balance item detail will yield the item activity for that item detail. Further investigation would require viewing the balance sub-application. By double clicking on the balance item activity, the application is automatically launched. This design approach is not only well-conceived for users of Opus, but it represents a more efficient technical solution to a memory-intensive process.

This design feature is found throughout the Opus product line.

Importing External Cashflows

The key to an effective risk management system is its ability to receive data feeds from external systems. The new cash management module of Opus is architected to accept ASCII or object oriented (OO) formatted files of cashflows generated by other systems. This feature means that the Opus can be the central cash management repository across trading locations, business units and product areas. The greatest value of disciplined cash management is the ability to capture all cash inflows and outflows with offsetting of netting and projections of future exposures for flexibility in determining funding decisions. Armed with this information, a capital markets participant can nimbly select from several options to ensure the lowest cost of funding and the lowest risk factor.

Intelligent Report Generation

Renaissance has developed a technically proficient but simple to use report generator called "The Organizer," which is based on structured query language (SQL) design. It permits users to point and click to select, reorder and format tables and data attributes into customizable reports. The interface is quite intuitive and does not require any knowledge of SQL. Once customized, users can define their own default views of the report for future reporting cycles.

In addition, the cash management module has a dozen or so standardized reports for immediate use and modification. These include book funding projections, currency funding projections, balance details, NOSTRO projections, suspense items and so on.

The organizer also delivers report previews in PostScript before sending to print. Reports can be stored and archived in either PostScript or ASCII format for historical and audit purposes. Printing can also be scheduled for off-hours generation after the end-of-day processes are completed. The ASCII file format means that Opus cash item data can be uploaded into a spreadsheet for further analysis.

Currently, graphical representation of cashflows and reporting is not provided by Opus. Renaissance has identified a contained graphical capability as a priority for future releases. There is nothing to preclude users from incorporating Opus data into their own graphic applications.

Conclusion

Renaissance Software Inc. has delivered a well-conceived and elegantly crafted enhancement to Opus for cash management. While cash management has often been considered one of the more mundane aspects of risk management, it is just these types of basic controls and systems that differentiate world class risk managers from the mere pretenders.

Mr. Kane is a senior business manager in EDS Global Securities Industry Group. He specializes in developing technology treasury, trading and risk management solutions for multinational banks and brokerage houses. Before joining EDS, he has held technology positions at Lloyds Bank PLC and at Sumitomo Bank Capital Markets, Barclays Bank PLC and Merrill Lynch.

Mr. Kane has a BS in finance and international business from New York University.

Wolfgang Porada
Senior principal at AMS American Management Systems

There is no shortage of supply with IT systems supporting derivatives trading. Most of the solutions provided by software boutiques are focusing on front-office analytics support. Solutions covering the full end-to-end functionality from dealer trade-entry through middle office to back office are rather rare. Getting a deeper look into the cash management module of the Opus Derivatives System was an exciting experience.

The software clearly demonstrates that it's design was driven by practioners in ALM and cash management. As an IT consultant, I'm used to the typical problem-that a bank's internal hierarchy and booking structures do not fit into a black box provided by a software vendor. Either customization of the product or add-on tools are necessary.

The new Opus module shows that it is possible to develop a completely flexible system, where nearly everything concerning accounting is parameterized and user-definable. I never have seen a system capable of booking every single cash flow item of a multiple cash flow instrument, like a swap, into separate accounts! I would love to see this functionality becoming part of a limit and enterprise-wide risk management system.

The cash flows produced through the booking process of the Opus system (or imported from external systems) are accessible through a SQL tool, a GUI called "Organizer." This tool is much more powerful than standard GUIs for data access-it is possible to launch an application simply by double-clicking within this presentation layer.

One of the applications that can be launched directly from the SQL tool was the FX funding tool. This application automatically creates the necessary funding deals or the appropriate loans for the stream of currency cash flows (or for single cash flows as well). The user defines the appropriate parameters for this functionality as minimum balances, rounding or settlement lags.

This high grade of flexibility is consequently maintained in the reporting area of cash manager. Eleven turnkey reports are provided by the system. Because all reporting is done through standard SQL, it takes little time to customize the existing standard reports or set up additional individual ones.

Although the system I tested was already beta-tested at a client site, the demo crashed while I was conducting a data insert. Obviously this was caused by an inconsistent system setup, because several people had played around with the database and system configuration before the demo run.

Renaissance continues to position itself as a vendor offering software driven by practioners' needs, an approach that normally applies more to development sites at the east coast near Wall Street than to the more academic sites in Silicon Valley. Hopefully last year's acquisition by SunGard does not emphasize short-term financial goals over long-term strategic development.

Wolfgang Porada serves as director in the risk practice of AMS Management Systems, a leading IT consultancy. He is responsible for the solution architecture for enterprise-wide risk management systems in the banking industry. He joined AMS from SunGard, where he served as head of sales for Continental Europe.

Dr. Jeanette Jin Chiang, Ph.D.
Vice President, Chase Manhattan Bank

Renaissance Software's Opus cash management system is a comprehensive, centralized cash-flow warehouse for maintaining and analyzing firmwide transaction and position data primarily involving current and future currency receives and delivers. It is designed as a general infrastructure for users who need to manage, analyze or forecast cash flows and perform other related activities such as gap management and treasury funding. To understand the extent of theses capabilities better, the advantages and disadvantages of the software will be reviewed against high-level criteria such as the completeness of the system, its efficiency, inclusiveness and, of course, its accuracy. This review will also stress the degree of flexibility the new module offers users to customize the system and to add new functionality.

The cash management module is a basic tool for analyzing cash flows, projecting cash flows and providing customized reports. It allows users to browse and organize both internally and externally generated data such as actual and projected receives, delivers and balances. Since cash flow is a key concept in financial risk analysis, it is important for any risk manager to have a system that is flexible enough to integrate cash flows and data from various sources. Opus offers this flexibility by allowing users to generate cash flows, accept external cash flows, add new instruments, and add new fields via application utilities and APIs. In addition, the application's cash flow structures permit the user to combine and recombine individual elements and build customized objects. Subsequent results can be manipulated in a variety of ways and shown via custom reports and displays. Any analysis can be saved and rerun as a batch file or printed as a PostScript document or a simple text file. In addition, reports can be generated either by book or by currency. The program was developed in C/C++ and provides users the additional flexibility to integrate their 'in house' programs. Opus is available for UNIX-based systems only and employs Sybase as its relational database management system.

The overall structure of Opus is quite comprehensive. It allows for extracts of global transactions and positions from external trading and operational systems to reside in a centralized data management facility. It also allows users to define and maintain static data associated with the various instruments on numerous levels. The layouts of the cash-flow maintenance utilities are particularly well done in terms of their flexibility to work with and display data in both summary and itemized levels. It also provides a mechanism that links money market and foreign exchange transactions using both trading book and funding structures to some of the analysis tools, such as cash analysis and settlement management, that was straightforward and took only a few seconds. On the whole, the funding applications provided efficient processing and user interface as well. Inexperienced users can navigate through the system via a well-conceived GUI. The software is window and menu driven and you only need to know how to use a mouse to get around the system. For example, it required only a simple click to accept a transaction, clear the screen, and be ready to enter a new transaction without the need to bring up the new transaction window again.

Although the system provides the user with a receive vs. deliver gap analysis, the system offers little support on how best to meet any deficiencies or invest any excesses. The software does not provide any significant tools to help optimize the level of the cash balance at different points in time such as portfolio optimization or artificial intelligence. (Tools for this type of analysis are available within the larger Opus package.)

Overall, as a cash-flow organizer, Opus provides an efficient vehicle to manage cash flow of complex instruments. It also provides flexibility for users to communicate from a trading floor to back offices and to extract cash flows into other risk management tools. It is a good software for the accounting-based cash management. Meanwhile, in order to meet today's high-level standards, the system needs to extend its capability of conditional analytic functions.

Jeanette Jin Chiang, a Ph.D. in finance, specializes in the modeling and quantitative analysis of mortgage-backed securities. She has published articles on the term structure of the eurocurrency market and on volatility in the equity markets. She is also a frequent guest speaker at banks and other financial institutions in Taiwan and mainland China. Her views are personal and do not represent the views of Chase Manhattan Bank.