Personal Investment Management Using XML

June 26, 1998

The Business Problem
The Role of XML
Solving the Problem with Chrystal Astoria and XML
   Situation Overview
   User Scenario
   Behind the Curtain
XML Example
Information Flow Diagram

The Business Problem

Investors need access to personalized information regarding their portfolios. Currently, financial institutions hand-craft such reports only for large companies or very important individual investors. Now the demand has increased for such reports from a broader range of individuals participating in group investment plans. To meet these demands, institutions must develop an automated service that combines data from various sources such as Moody's Investors Service, Standard and Poor's, EDGARS, financial analysts, and personal profiles. In addition, customers are demanding the ability to perform "what if" scenarios with the data and demand more charts, spreadsheets, and graphs to better understand the data. They are also looking for more up-to-the minute information about their investments.

The Role of XML

We are in the midst of a fundamental shift in the way that people store information -- away from the traditional use of documents and files to thinking about a document as a container into which various pieces of data are combined or "poured" as required. This kind of document may be stored for an extended period of time or it may exist only as long as a person interacts with it.

In this scenario, XML defines the vehicle for describing and delivering data that enables retrieval, reuse, exchange, and combination in new ways. Specifically, the system uses information encoded in XML for querying and analyzing financial information. The results are dependent on the individual's profile.

Solving the Problem with Chrystal Astoria and XML

Situation Overview

With Astoria from Chrystal Software (http://www.chrystal.com ), the financial institution is able to create on-demand reports containing up-to-date information, tailored for each individual investor. Since the report is generated automatically from existing data, multiple views can be developed with no incremental cost. These reports provide an instant view of the data, and provide financial advice based on various "what if" scenarios.

Using Astoria, the financial institution's fixed information assets become liquid. With no programming ("out-of-the-box"), the financial institution loads both the XML data and the non-XML components into Astoria where it can be accessed, tracked, and versioned independently, providing a central repository for on-demand service. With Astoria's Software Development Kit (SDK), the financial institution streamlines and automates its data assembly and delivery processes.

User Scenario

Two employees in a Fortune 500 company investment plan wish to see their current portfolio and make some changes based on various scenarios and criteria. They are in the same plan but have very different financial needs.

The first employee selects the internal financial channel and identifies herself. With stocks soaring, she wants to get the latest analyst reports to see if she should buy or sell based on the funds that are available in her plan. When her portfolio is displayed, she's asked what action she wants to take. First she must decide if she wants to buy or sell immediately or set up instructions to track certain stocks and buy or sell when they reach a certain price. She is also interested in evaluating some long-term options.

To make these decisions, she poses various scenarios to the system. She has two children to send to college and wants to ensure she'll have enough income for her own retirement. So she asks the system to evaluate what kind of mix she needs based on the latest stock fund ratings and analyst reports from Moody's and Standard and Poor's. She asks the system to tell her (based on what's available in her plan) what portfolio will meet, exceed, or miss her expectations over 18 months, 5 years, and 20 years. The system returns the data on her current portfolio and makes some recommendations for long-term investment. A report is returned to her online, in a series of views showing the combined data along with the recommendations.

A second employee begins the same way but sees a completely different portfolio -- generated from the same financial data but based on his "stock picks" and his personal profile. He is an aggressive investor, young, with no children. He chooses to have the system remember his requests and send him a similar report updated on a weekly basis via e-mail.

For no additional charge, the financial institution has created multiple custom reports that meet customer demands.

Behind the Curtain

Integrating data from multiple sources

Chrystal takes advantage of the self-describing nature of XML to allow users to quickly and precisely locate information. To accomplish this, Chrystal enhanced the Verity search engine to understand XML content, structure, and meta-data. This engine lets users search for document content, structure, attributes, and version information. For instance, users looking through their portfolios can now search for data contained within tags, within elements in context, and for elements within structures. They can also retrieve data contained in meta-data -- information about the information.
Data about the individual portfolio is stored in an object database and accessed via the Astoria content management system.
Orders specifying transactions in the future use an XML-based form. Specific event-based actions are built with the Astoria SDK. Astoria provides an XML-based form with constraints based on the profile for the "what if" scenarios.
Human Resource (HR) data about the employee is stored in a relational database, on another server. Data is extracted from the HR database through the Query feature and transformed into XML data by one of the XML-compliant transformation products like Omnimark, XML and scripting, or the Astoria SDK.
Some of the report and analysis data is also stored in the Astoria component manager where it has been enhanced with XML meta-data.
Other financial data is retrieved from external Web servers, like Standard & Poor's. Here the HTML data is "scraped" and converted to XML. If any of that data is stored, it is enhanced with XML meta-data.
Data is combined and presented in a view -- a "document. " The scripts that implement the business rules that govern how this integration occurs are constructed with the Astoria SDK. For instance, a business rule defines that for single mothers with teenage children only safe investments with long-term growth be recommended..

Getting information updates

Instant updates are provided by the Chrystal Astoria "differencing" engine, which monitors data within specified XML tags and reports changes to the data.

Delivering data for display in the browser

Data is delivered to the client via HTTP as XML data where it can then be manipulated according to the user's preference via the XML Data Source Object (DSO) and the parser in Microsoft® Internet Explorer. For display, it is then moved to Dynamic HTML (DHTML) on the client

XML Example: Financial Portfolio

The following XML example shows a financial portfolio combining local structured analyst data and remote relational investment data.

<portfolio name="Mary Weber" risk="balanced" return="medium">
  <profile name="Weber, Mary Ann" employee="82340" source="HRdatabase"/>
  <analystreports>
    <Moodys source="http://www.moodys.com/profiles/webermary"/>
      <description source="Astoria.Moody.Description.complete"/></Moodys>
    <SP source="http://www.ratings.com/funds/profile/weberma">
      <description source="Astoria.SP.Description.partial"/></SP>
  </analystreports>
  <investments>
    <shortterm>
      <moneymarket type="taxexempt" name="Phoenix"/>
        <bond type="taxable" name="BioTech"/></shortterm>
    <longterm>
      <retirement>
        <ira contribution="maximum"/>
        <beforetaxsavings source="HRdatabase"/>
        <collegefund name="NY Aggressive Fund"/>
   </retirement>
    </longterm>
  </investments>
</portfolio>