Printer Friendly

Information power.

Mortgage banking has gone from the data retrieval dark ages into the present. The power of the PC, coupled with new software, has put crucial management information at our fingertips.

ON JUNE 3, 1979, RICKIE UPSTART AT OUTFRONT Mortgage was pestering his boss, Ben Beenaround, to let him use the company's computer. He wanted to research whether the company had an overconcentration of servicing in Texas. Beenaround was hesitant--he knew it would mean writing up a proposal, having his boss review it and then sending it to the programming priority committee.

The data processing department had a backlog of requests including two of Beenaround's own and five from people reporting to him. He would have to fight to get a priority, and he wondered what impact adding another request would have on the seven requests already in the data processing queue. What's more, by the time Upstart's answers came out, Beenaround figured nobody would remember the question. Besides, Beenaround knew the mortgage business probably better than anyone. Texas was a great area--always had been.

He told Upstart to find something productive to do, like opening up the Oklahoma market. Oklahoma--now, there was a booming territory.

If this scenario takes you back to what times were like in the mortgage business at the start of the go-go 1980s, then you're not alone. For many in the industry, the absence of accessible, readable data on portfolios in the last decade meant that flying blind was, in too many cases, the name of the game. At least a few bad decisions were inevitable. Today, we almost take for granted the wealth of new information we have to help us avoid those bad decisions that in the old days could all too quickly take down a whole company.

For many people, "data" and "information" are interchangeable words. But mortgage servicing professionals have a different perspective. They can be drowning in data--5,000 to 10,000 characters of information per loan on hundreds of thousands of loans. But if mortgage executives are unable to condense this mass of data into summarized form, they have no useful information with which to make decisions. During the past decade, it has become increasingly easy for the end users to manipulate this all important data with less and less involvement from data processing professionals.

Up until the early 1980s, distilling information out of the masses of computerized data collected in the mortgage banking business was a lengthy process, unless the specific report you wanted was already programmed into the system. Someone had to communicate in writing the information desired, obtain approvals and usually have a committee assign a priority to the request.

When the request finally reached the top of the priority list, a systems analyst would design the solution and communicate with a programmer exactly what needed to be done. The programmer then had to write and test the program.

Assuming the communication process worked correctly and the program was well written, the information was produced within weeks, if a high priority was assigned; in months, if not. Such an approach to getting usable information out of your system engendered considerable costs. Such costs included management time, systems and programming costs and waiting time.

The lost time could be the most expensive item, if the information was needed for an urgent business decision. Rapid economic changes or sudden business developments that proved conventional wisdom wrong were difficult to detect until too late. Only the most far-sighted firms, and, of course, the lucky ones came through the volatility of the early 1980s unscathed.

But the only alternative to adhoc programming (or working your way up to the top of the programmer's priority list) was virtually useless. This alternative was really no alternative at all. It consisted of attempting to plow through a loan-by-loan listing several feet high that contained the raw data. The user could take a pad of paper and a calculator, write down the information loan by loan, add up all the columns and die of old age before getting an answer.

Within a couple of years, managers became able to get much better access to their information by the creation of commercial languages. Let's look at a scenario that shows what this might have meant to a mortgage company interested in buying some servicing about six years ago.

On August 22, 1987, Meg Crunchitt turned back to her project of evaluating servicing portfolios offered for sale. She had four interesting portfolios to review, all of which had come with master file extracts for her to analyze with her new commercial language. She was generating the preliminary reports and compiling the statistics produced--weighted-average interest, concentration by loan type and age, etc.--and predicting the expected life of the groups of loans, which was crucial to the value she would assign.

The reports were piling up rapidly, and she would be done soon--would have been done already, really, except for the constant interruptions. A few minutes ago, the tax department manager had been begging her for help. Now Meg heard footsteps approaching her door again. She tried not to look up, but no use. The collections manager plopped down in a chair and demanded help forecasting delinquency rates for New England. Would Meg just run her a report showing delinquency by product type, state and origination year for New England? Meg knew she would be at work late again tonight.

Such was the state of things when the ability to access and shape the raw data into meaningful reports still remained in the domain of just a few of the more analytical members of the mortgage management team within a typical company. But this too was eventually headed for change.

It was really in the mid-1980s that the more progressive mortgage banking companies, such as Meg's, began to use commercial languages that were less technical, higher level languages. This involved extracting from loan servicing master files key elements of information and storing these elements in a data file, with specialized software that could be used directly by the analytical staff.

This removed the need for highly trained data processing personnel between the information user and the information. The analyst could become proficient in using the software in a week or so. The time it took to get adhoc reports was reduced from days or weeks to hours.

This huge improvement was accompanied by a new problem--it was not uncommon for a line to develop outside the door of the person who had the skills to use the software. Waiting time began to build again, and power struggles over priorities began to arise. The analysts, like Meg Crunchitt, trained to use the programming as a secondary tool to further their research capabilities, resented being pressed into service as programmers.

In the meantime, the need for information was increasing. Interest rates were fluctuating wildly. Runoff rates were increasing. No longer did mortgage bankers make a loan at 6 percent, service it for 15 to 30 years and then process the payoff. Foreclosures became a much bigger problem, the frequency of sales and purchases of servicing began increasing rapidly and mortgage banking became attractive to large industrial conglomerates.

Mortgage banking became a highly complex business in a highly volatile stage of development. And volatile complex businesses need information to remain profitable.

PC power

While the need for information was increasing, the technology of data processing underwent a revolution through the development of personal computers (PCs). Their power has increased dramatically, while costs have tumbled. The mainframe computers used by midsize mortgage companies in the 1970s were less powerful than the PC systems available today for less than $4,000.

Perhaps more importantly, the mainframe world has tended to concentrate on repetitive applications that need to be done precisely, thousands or millions of times a day. The PC world has concentrated on the quality of the user interface to make the computer easier to use for audiences that are not expert with computers. The development of the mouse and "point and shoot" menus are examples of such user-friendly adaptations.

Let's travel forward in time for a moment and get a glimpse of a possible future scenario.

On September 14, 1997, Gayle End-User came back from a brief honeymoon. She greets her computer with a cheery, "Morning, Alpha." The Alpha-23 verifies Gayle's voice-print, consults its access and security clearances, loads her personal nickname and abbreviation codes and responds immediately, "Good morning, Gayle. What can I do for you today?"

"What's new?" Gayle asks. This user-specific question triggers the computer to produce the routine daily statistics needed for the operation of the customer service unit--number of calls taken, hold-time information, comparative results from the previous day, week and month and the number of agents currently on calls.

You're skeptical. It sounds a bit too far-fetched? Today, perhaps, but it will happen, and the speed at which we are approaching this technology in the mortgage banking workplace is accelerating rapidly.

The scenario is most notable for the speed of interaction and ease of use of the system for Gayle--the end user. Some sophisticated programming must be done by specialists before handing the system over to the user, but from that point on, no special computer language is needed. The data will be available immediately. We are beginning to approach Gayle End-User's enviable work-place situation.

In the real world of mortgage banking today, three broad software techniques exist that are major improvements over what poor Rickie Upstart contended with in the late 1970s and early 1980s.

Commercial languages

Commercially available high-level languages that work with data base files have eliminated the need for systems analysts and programmers to be involved with most report requests. The skill level needed to produce reports can be achieved by most mortgage company employees. The advantage is that complex reports containing highly computational elements can be programmed. The disadvantage is that it takes several days or a week to learn the syntax of the language.

Also, the time invested in learning the language is lost if the person does not use the program regularly. Furthermore, the functionality of this is limited by the fact that some people just don't like that kind of work or don't have the kind of minds suited for this.

The speed at which today's PC can provide the information when using a high-level commercial language is breathtaking compared with a few years ago. The average time to process and report on a 200,000-record file is now usually less than 10 minutes. The time will vary depending on the specific language and computer hardware, but the important factor here is that we are now talking in terms of minutes, not days.

For applications where the number of records is not very large, commercial languages make a giant step forward in terms of speed. While we have seen definite improvements in terms of ease of use of commercial languages, still too much special computer language knowledge is required for infrequent users or senior management.

Information systems

A second development that has aided mortgage banking managers is the informational system that pregroups the data and uses software that lets the user work within preprogrammed reports. The user selects the report and data options, then the system combines the detailed groups of data correctly. An example of such a report would be one that shows delinquency rates by state and by product type. Well-designed systems are very easy to use and require little knowledge of computer syntax.

The speed of the system depends on the complexity and number of detail groups. If the number of groups is large, even simple reports can require 45 minutes or more. This is acceptable to an analyst doing large projects but can discourage line personnel and senior management from using the system for their purposes.

The major disadvantage of using a grouping technique is that the user is limited to the preset groupings--he or she may not be able to get the data the way it's really needed. For example, you need to consider how detailed the grouping is for geographic identification. If you group by state, what if the user wants the data by standard metropolitan statistical area (SMSA)? Or, if you group by SMSA, what if the user wants the information by city?

Also, with these systems, loan-by-loan detail is not available. For example, if a report showed that a certain five-digit ZIP code had a very high delinquency rate, these information systems would not allow you to get a list of the individual loans that caused the high delinquency rate. Upfront programming time can be extensive, but it is insignificant compared with the investment of several hours of mainframe time to update the data to reflect the current situation, especially if done daily. These disadvantages are a trade-off with ease of use for the end users.

Newest technology

The latest development to aid mortgage executives in their pursuit of more instantaneous management information is new software technology. This technology significantly increases speed (compared with commercial languages), is very easy to use (similar to information systems) and, at the same time, allows for loan-by-loan listings.

This new technology is being used in the McDash Data Base Download, which currently contains data on more than 6 million home mortgages currently being serviced. This software analyzes loans at an average rate of about 2 million loans per minute, thus having output in three minutes. When installed on a server, which is a central PC that is connected to employees' individual desktop computers, several people can use the system at the same time.

The design of a requested report is done by "point and shoot" menu selection. A new user can be using the system with less than one hour of training. Because almost no special computer knowledge is needed, personnel who use the system infrequently do not lose their ability to use it when needed.

This same technology is employed in the new McDart system, which can prepare both summary reports and a loan-by-loan lister for more than 400 fields of information contained in the individual mortgage company master file. The ease of use is comparable to the McDash system, and the speed of reports is measured in seconds--about 15 seconds for the average large mortgage company. While this technology does not include voice-response capabilities of the kind Gayle enjoyed with Alpha-23 in our high-tech scenario of the future, the speed is matched, and any required special computer knowledge is nearly eliminated.

Remember when Meg Crunchitt was working on her servicing valuation project, and the collections manager appeared in her office and demanded help? She needed a report showing delinquency by loan type, by state and by origination year for New England. With the newest technology available today, the collections manager could have prepared the report herself in a couple of minutes using only six keys on her computer keyboard. All of the options appear on the screen in English, not computer jargon.

We decided to test just how easy to use this latest technology really is. In a test in this author's office, an infrequent user of the newest technology, running on a 486 33-MHz machine (state-of-the-art PC hardware in 1991) took 2 minutes and 15 seconds to generate and print the report that the collections manager asked Meg to help create--a status report on delinquency rates by product type, by state, by year of origination. Designing the report took 25 seconds. Computer time to cycle through a 6 million loan data base and to collect the data on the company's 300,000 loans was 70 seconds. Including the time it took the user to notice the computer was done, in 40 seconds the report was printed.

Many companies today have one or more of these methods of preparing ad-hoc reports. This gives these companies the ability to get the information they need for business decisions within minutes or almost always in less than an hour. Mortgage managers today need to know that it no longer should take hours or days to get the information needed to make key decisions.

Back at OutFront Mortgage--still in business despite some mean encounters with Oklahoma no-bids, it's lunchtime on August 4, 1993. Rick Upstart, now a vice president, has just gotten a call from the executive suite of the parent organization. They are interested in buying a servicing package of California loans. They want to know what the delinquency and runoff trends have been. Is Southern California the same as up north and how about the Los Heuvos area in particular? They needed answers fast, and Rick went looking for help.

Rick glanced into the mailroom; as he walked past, he saw Ben Beenaround. No use talking to him. Ben would just refer him to systems and programming, just like the old days. Rick knew that route would get him the information, all right, but way too late to suit the brass. Rick's first stop was at Meg Crunchitt's office. Meg's secretary told Rick that Meg was in a meeting at another company negotiating a sales agreement on the large package of loans OutFront was selling. The meeting would be over by 5:00 p.m. It wouldn't take long for Meg to run the reports, Rick knew, with her commercial language.

Continuing down the hall, Rick asked the manager of the planning group if their new information system could help out. Sure, said the manager. "I can run a couple of these package reports. It only takes a few minutes. Of course, I can only get you California as a whole--we can't break it out for just certain parts of California." Better than nothing, Rick thought, and quick, but not quite what I need. Finally, Rick poked his head into Jana State-of-the-Art's office to find her working with the new technology system. Explaining his project, Upstart asked if the new technology system could get him his answers. Sure, Jana said, in just a few minutes.

The PC is the future

No matter how your company is doing ad-hoc reporting today, improvements will be made during the coming years. Advances in hardware technology continue to be made, especially in PCs. The raw speed of PCs is increasing as 50MHz, 66MHz and even 100MHz models replace the older 25 and 33 MHz PC. The new Intel Pentium microprocessor will make PC speeds increase dramatically. The increased speed of computers will allow further improvements in user interface, meaning less training time and easier-to-use software applications for mortgage bankers.

For most mortgage companies, the master files are stored on mainframe computers. In the future, the mainframe will be used to create the file containing the fields of information to be included in the informational system and downloaded daily, weekly or monthly to the PC. All further work will be done on PCs. The people needing the information will intereact with their desktop computers.

If an in-house mainframe already exists, one could argue that it is cheaper to use the mainframe than to buy new PC equipment. It may be cheaper in the short term, but in the long run, it would be more expensive. Unless your mortgage company is shrinking, at some point, the mainframe will run out of capacity. To buy incremental capacity will be a major investment ($100,000 and up, most likely). PC capacity can be added in much smaller increments: $10,000 buys a lot of PC capacity.

The technology for our futuristic Alpha-23 scenario is actually available now. However, it is not yet economical to invite such an intelligent form of computer life into the mortgage banking workplace. But state-of-the-art systems are getting close to becoming a permanent fixture.

Whatever system you use for ad-hoc reporting, it can be judged against the following four standards:

* Ease of use--minimal special computer syntax knowledge, no extensive training and easy to remember.

* Speed--seconds for an individual mortgage company's data, a couple of minutes for multimillion loan data bases.

* Flexibility--user control over groupings and loan-by-loan detail listings available when needed.

* Need for computer professionals--only in establishing and operating the system; no involvement in individual reports.

For the professional analyst, like Meg Crunchitt, the commercial languages are very satisfactory unless the data files are very large and speed becomes an issue. For management, whether it be executive or line management, with a need to obtain information, the ease-of-use and speed criteria are overwhelmingly important. If technology is not easy to use, many won't or can't use it. As waiting time for the computer response increases beyond a minute for the average-size portfolio, management will begin to stop using it. Information systems may meet this criteria for high-level, limited-reporting systems. In most cases, the needs of executive and line management for flexible systems will be met by the new technology.

Alpha-23, I look forward to talking with you very soon.

Dale G. Poszgai is president of McDash Analytics, Inc., and McDash Software, Inc., Jacksonville, Florida.
COPYRIGHT 1993 Mortgage Bankers Association of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Poszgai, Dale G.
Publication:Mortgage Banking
Date:Aug 1, 1993
Previous Article:Partnerships bridged with technology.
Next Article:Electronic mortgage talk.

Related Articles
Standards for library media centers in schools for the deaf: an updated perspective.
Changing Your Management Style: How to Evaluate and Improve Your Own Performance.
'Traffic light' for computers. (Software Briefs).
Your Library Goes Virtual.
Collaborating to meet literacy standards; teacher/librarian partnerships for K-2.
An educator's guide to information literacy; what every high school senior needs to know.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters