Our industry has been fraught with
technology blunders. At the same time, a select minority of new solutions has
triumphed to become successful mainstream applications. There's a lot to be
learned from our past mistakes, and we can use history to help tell the
future. One thing is certain: Many lenders and others in today's challenging
business environment will continue to waste money on the wrong technology--you
want to make sure your company isn't one of them.
There seem to be two core beliefs that drive investment in unproven
technologies: 1) Technology changes will revamp entire segments of the
marketplace, and 2) Lenders mistakenly believe that not leading with technology
implementation can be costly or even lead to a loss of competitive advantage.
If the technology train really is leaving the station, lenders are afraid of
missing it . Yet it's rare in our industry that
The smart players are those that allow a few small companies to experiment with
and prove a solution, and then they jump in full-force. With such patience, a
model can be better understood before it’s adopted and money isn't wasted
trying to perfect the solution. Many large lenders understand this concept
fairly well. Still, the recent debacle with HomeAdvisor(symbol)
(TM) Technologies Inc. (HTI) proves even the largest industry players can be
misled by nothing more than a business plan on paper.
Before
making any
technology prognostications or
assessments, let me
share a little of my background in mortgage technology. More
than 20 years ago, I developed
the industry’s first PC based LOS,
Contour Software. Since I formed the
company, Contour was bought and sold
twice (each time with growing valuations.) Additionally,
I have reviewed every
significant mortgage technology firm (usually
meeting with the CEO) and their solutions
for the past two decades. Although I have
written several articles predicting mortgage
technology trends, my
most notable prediction being the failure of the Value Added
Networks (VANs) to gain
widespread industry adoption.
Today, I define strategy for Ellie Mae, the largest mortgage origination
technology firm.
Now, let's review the past 20 years of mortgage technology and consider how
some
perceived technology marvels ended in failure. This article looks at a number
of past "fad" technologies that took the
industry by storm and yet, ultimately, only appear to have caused significant
losses. In this year alone, nearly a half-billion dollars is being wasted in
just one technology category (ASP solutions for lenders). What about the
latest offerings? Some show real promise and others predictably will flop.
Finally, let's look
out on the horizon and see what technologies of tomorrow are the most likely to
become mainstream applications.
Video conferencing
One of the classic mistakes we see underlying many
past technology blunders is the idea that an aspect of the industry can be
completely reshaped by a new technology. The worst mistakes are caused by
technology solutions designed to replace what are at their core people
solutions. For example, loan origination, at its core, is about personal
relationships, and someone on a screen doesn't replace personal contact.
In the mid-1990s, the concept of video conferencing was embraced as an
appealing application for loan origination. Articles were published predicting
that origination channels would be redefined with the power of this technology.
Video kiosks were showing up in shopping malls across
At least a dozen startup companies produced and marketed these solutions to the
mortgage industry. All told, I estimate that approximately one-half billion
dollars was invested by video conferencing firms, Realtors, lenders and
investors—in 1993, the largest firm alone had received more than $100 million
in funding This wasn't just
a fringe movement by a minority faction in the business, either. There was
hardly a mortgage technologist or executive who wasn't on board with this
compelling application. Yet, in the end, it was all a waste.
E-lenders
Thinking the Internet could replace what a loan
officer does is one of our most recent mistakes, in my view. Once again, like
lemmings, we all followed one another down a path of perceived technology
perfect solutions.
At the Mortgage Bankers Association of America's (MBA's) 1999 mortgage
technology conference, I asked the audience to estimate what percentage of loan
originations it thought would come through e-lenders by 2003. By a rough count
of the show of hands, approximately 10 percent thought the number would be less
than 20 percent, around one-third thought it would be 40 percent and the
remaining felt more than 60 percent of originations would come exclusively from
a Web site.
Once again, many were wrong. But why?
X.12
A March 1996 Mortgage Banking article written by Dan McLaughlin and Leilani Allen, PhD, entitled, “EDI” Setting the Record
Straight” said, "The primary value of X.12 is that it provides a unique
opportunity for all parties involved in a given transaction to
together define requirements. . . . This means that new values can be easily
added to suit the needs of one industry without having to define a completely
new format. . . . It is also useful to remember that only the first conversion
is difficult. Once people are familiar with the approach and technique, it is
merely a repetitive task."
Even strong industry leadership organizations such as the Mortgage Bankers
Association of America (MBA) are capable of misjudging the potential staying
power of a given technology. In 1995, I warned the industry against the
pitfalls of X.12 – I took a lot of heat for it, too! However, it seemed
everyone believed the vision that X.12 standards would change the
industry—making transactions more streamlined, saving hundreds of dollars per
loan. From the largest lenders to the smallest technology startups, everyone
was planning to adopt X.12. Unfortunately, when it came time for
implementation, the
reality was significantly different.
With X.12, the problems stemmed from a misunderstanding of the impact
technology could have on our complex environment. The original design of X.12
was intended for facilitating transactions in industries (such as shipping and
invoicing) not nearly as complex as those of the mortgage industry. Those in
our industry involved in the decision-making process to adopt X.12 had little
to no experience with actually creating electronic data interchange (EDI)
solutions. In hindsight, it's hard to say how we could have avoided this. The
positive impact of this failure is that it gave us the experience
needed for the Mortgage Industry Standards Maintenance Organization (MISMO)
project. So the exercise has proved somewhat beneficial.
Origination systems reinvented
About once a decade we see new origination systems
come to market from new vendors. These "revolutionary" loan
origination systems (LOS) promise far greater efficiency from their new technology.
This was first apparent when Windows(symbol) (R) was
introduced, then imaging became available and finally the promise of the
Internet arrived on the scene.
In each case, five to 10 new vendors would arrive with solutions based on these
new platforms. Some lenders eventually purchased these systems, but within two
to three years, the vendors' funding-well ran dry and the LOS were gone. This left lenders with systems that couldn't be
properly supported and maintained, and eventually they were forced to purchase
new systems.
The mistakes in this case were twofold. First, there was a belief that the new
LOS were radically more efficient. Most often, these
systems were less efficient as they didn't have the refinement that comes with
years of actual use. The top 10 LOS today all took at least a decade to design,
test, implement and continuously improve. New platforms rarely make a dramatic
improvement in productivity.
Second, existing LOS vendors will migrate their
applications to any proven platform. Thus, the predominant vendors today are
the same ones that existed in the days of the DOS world. Further, these very
companies will be the ones that migrate their
applications to any future new platform that proves its market
acceptance.
Work flow
Returning to the 1990s, work flow was a big
buzzword. Many lenders saw it as the solution to high labor costs. Today,
in my view, workflow has proven of little value to the typical branch
origination office In fact, many companies actually increased their costs of
originating loans with workflow solutions. The cost of implementing and
maintaining these systems outweighed the staff hours they saved, from what I
have seen. There are a few success stories, but such cases involve certain
types of origination environments,
including call centers and other high-volume operations.
Once again, the problem arose with mortgage technologists and other decision
makers becoming enthralled with a new technology without fully comprehending
its impact and benefits. Because it is difficult to evaluate these solutions,
many origination companies have been sold on theory alone.
Transaction systems
During the last two years, there have been more than a
dozen companies attempting to build the industry's standard transaction system.
The attempt is to create a system that would work with all parties involved in
the mortgage transaction.
Again, the industry spent more than a half-billion dollars. Many companies like
NetOriginate, Ocwen and nCommand have closed down. Others like OpenClose
and ALLTEL's Interchange never saw the industry
adoption among mortgage originators they had expected.
At one point, more than a dozen companies were vying for the market niche to
drive transactions from originators and vendors. Now, following a shakeout, a
few companies are in a position to profit from this market including Ellie Mae
Inc.,
The reason the other companies failed is simple: They all built systems that
stood apart from the LOS. Both Ellie Mae and RealEC understood from the
beginning the need to integrate tightly with originators' mission-critical
back-end systems (LOS). The others failed to see that transactions were
currently being handled by the LOS (though in a more manual fashion), and thus,
originators expected e-commerce transaction functionality to occur from within
the LOS. When done from outside the LOS, data integration issues
were too significant and it was too cumbersome using two separate systems.
Given the overall downturn in the capital markets (stock markets and venture
capital), it's unlikely the mortgage industry will see so much funding wasted
on technology in the future. Instead, the invested capital will require a far
greater chance of success by seeking a proven model. Further, there will be
much more reliance on experienced management teams with direct and proven
experience in mortgage technology.
Most of the firms that failed were led by managers without long-term experience
in mortgage technology. Our industry is very complex, and it takes years to
understand both technology and the mortgage industry—and years more to
understand how they should work together. The mortgage technology firms
thriving today have management with at least 10 years of direct experience.
In the current marketplace, there are several ideas in which companies are
investing and yet it is my prediction that many of these will fail or have very
limited success. At the same time, I foresee some of the new solutions
succeeding. Here's a look, from my perspective, at each type of solution along
with my general rating (from hot to cold) as to their viability from today's
market perspective.
Transactions systems: Hot
While we have just finished noting that the last few
years brought over-investing in transaction systems, still we believe
transaction systems are clearly the future of our industry. Ordering credit
reports electronically has been popular for many years. Still, other
transactions like appraisal, flood and title orders have yet to gain widespread
use. In
the months to come, the industry will experience a significant increase in
ordering all types of services electronically--in fact, it's already happening.
Electronic orders will become a de facto standard for all transactions in our
industry. All LOS vendors are and will continue to be the driving force for the
development and implementation of these new electronic transaction solutions.
Application service providers for LOS: Cool
Loan origination systems that are ASP-based (application service provider) will
find a cool reception among mortgage origination companies. Perhaps their
biggest weakness is they'll never work well on a portable laptop (at least not
until we have high-speed, wireless Internet access). Further, when all aspects
of an ASP loan origination system are considered, it will be more expensive and
less productive.
On the surface, the ASP model looks very intriguing. But look under the hood,
and you'll see a long list of problems. At this point, an ASP solution is only
viable for a small minority in the industry, in my view. Also, I believe it
will be the entrenched LOS vendors that will supply these systems once they are
proven attractive.
It should be noted that many other ASP solutions have the
potential to become hot, including e-commerce solutions for
wholesaler-to-broker communications, management-reporting solutions, providing
loan status to third parties and others.
The primary issue with ASP-based systems is that working in a browser
environment is generally less productive than working in a similar Windows
application. A good example is how e-mail, the ultimate Internet application,
is primarily handled via a Windows application such as Microsoft Outlook™ or
Eudora™ rather than a browser-based solution such as Microsoft's Hotmail™ or
AOL™. This is certainly the case for large-volume users that rely on constant
e-mail usage.
Note that high-production environments with mission-critical applications, such
as an LOS, are more likely to be desktop-based solutions. Other problems
associated with browser-based solutions include data-entry inefficiencies, slow
printing and problems with imaging. The future is actually strong for an ASP
LOS, but an ASP LOS is much farther away than anyone expects.
Digital signatures: Cold
Homebuyers want their hands held as they pour through the 100-plus pages of a
closing package. A digital signature and an electronic signing room is a far
cry from a loan officer handing you the keys to your new home at the closing
table. There are also many technical issues to overcome, which are roadblocks
to consumer and industry acceptance.
With that said, there are some uses for digital signatures. For home-equity
loans and upfront disclosure packages, the concept is terrific. Even refis could be handled this way. However, there's one major
problem: Borrowers will have to obtain a digital certificate, which still
requires a notary signature and completing a somewhat arduous process. Consumer
acceptance of digital signatures is years down the road. Nevertheless, lenders
will be investing money preparing for its eventual arrival.
Rate distribution systems: Cold
Since the early 1980s, the mortgage industry has seen several technology
providers attempt to create a wholesaler-to-mortgage broker solution. Such a
solution would distribute rate and product
Dig deeper and you'll find that in the last 20 years more
than 25 companies have attempted to build this solution--and all but two have
gone out of business. Further, the founders and investors sunk more than $100
million combined into this business model. I'm unaware of a single company that
ever earned a profit trying to distribute rates and mortgage product
While rate distribution systems have failed, there is an alternative likely to
take hold. Prequalified rate sheets (PRS) are a newer solution that will likely
be a winner and replace the need to distribute rate sheets. With PRS, a broker
sends data about the borrower to a wholesaler and the wholesaler's e-commerce
solution returns a PRS. It is in essence a price sheet for a particular
borrower and will likely include risk-based pricing.
Pasadena, California-based IndyMac
Bank's eMITS(symbol) (R) system is a good example, and
has certainly proven this business model. Soon, these systems will all be
accessed from the LOS so that a mortgage broker can obtain a PRS from virtually
any lender with the click of a mouse. The borrower data will flow from the LOS
to the lender's Internet servers, where the data will be crunched and a PRS
will be returned. The one surprise here is how slow the wholesalers have been
to adopt this model, given its powerful potential to shift market share
significantly (witness the climb of IndyMac's market share since eMITS was introduced).
XML: Warm
The industry would surely love a standard, and nothing
has come closer to it than extensible markup language (XML) and the MBA's MISMO
group (MISMO defines the data within the XML format). Scores of talented
industry leaders have put their heart into creating these standards, and some
technology companies in the industry have begun using them. Still, the
standards don't yet work for all transactions or in all circumstances for a
wide variety of reasons. Thus, XML's successes will continue to be a bit of
hit-and-miss.
The industry is beginning to see some significant usage (unlike X.12, which
never saw much). To the mortgage originator, XML really isn't that important
because it operates behind the scenes. For originators, the only issue is if
the data can move from company to company without fail. For this reason and others, alternative data
formats will continue to see widespread usage—such as Fannie Mae's Desktop Underwriter(symbol) (R) (DU) file format. Also, existing
vendors are creating comprehensive loan file formats with XML, and use data
tags created by the MISMO group. Thus, some companies will use
aspects of the MISMO work even if the predefined MISMO transaction sets may not
be a perfect fit.
E-lenders: Cool
Today it's no secret that the dot-com e-lenders didn't take over our world. The
wasted capital was considerable--likely more than $1 billion. In my view, pure
online lenders will continue to have limited success. Mortgage originators will
instead use the Internet to support their existing operations. Sure, some
aspects will become streamlined and Web sites will improve, but the demise of
the loan originator is at least a decade away (if
it's coming at all).
There are a couple of areas realizing true success. Home-equity loans and
Second mortgages are prime candidates for pure Internet origination systems.
Refinancing will also be a hot application for e-lending. In the next refi
boom, the mortgage industry will see a much larger share of Web-based refi
originations. Loan originators don't add much
value in the refinance transactions, and this business will most likely belong
to the large servicers. Today's large lenders with their huge servicing
portfolios will eventually figure out how to refinance their own customers
without paying for the added costs of loan origination channels.
Web site builders: Warm
Most Internet service providers (ISPs) are getting hit hard in the public
markets, with their stock prices 80 percent to 95 percent off their highs.
Still, those companies that host Web sites will always be around, even if
consolidation dramatically reduces the number of ISPs out there.
In our industry we've had experiences with hundreds of ISPs that could build a
Web site for a mortgage company. Their ranks have been reduced dramatically (in
the last 12 months I’d guess this number has been cut in half), and most
mortgage companies now realize they need to work with an ISP that understands
the mortgage industry--knowledge that comes from ISPs that serve the mortgage
industry exclusively.
There are dozens of this type of vendor, but their ranks have also been
shrinking. In 2002, it should be clear who the remaining winners are in the ISP
battle for market share. Eventually, there will only be a few. To determine
which Web developers will survive, simply look for the market leaders that have
the greatest number of customers. They are the ones to offer more "bang
for buck." This area will always be around--it's simply
maturing.
Internet portals for mortgage professionals: Hot
This is a new aspect of our market and it's just
getting started but appears to be coming on fast. Think of MyYahoo! for the mortgage industry and you'll get the picture. Each
worker will have his or her own Web page that contains all the
These portals are designed to handle all the transactions
that each worker needs. For example, loan officers can find the loan products
they need for a borrower, such as a PRS. Loan processors can use the site to
order credit, title and appraisals, with a click or two, from their favorite
vendor. Such sites will also contain an abundance of general
Mortgage Internet portals that are integrated with the LOS provide the most
promise as data moves seamlessly between the LOS' database and the vendors'
back-end systems. The new technology terminology for this is "supply chain
management"--when back-end systems of both the supplier and consumer
(i.e., origination company) become tied together.
Portals allow both the mortgage company and the vendor to know each step the
other is making on any given loan. A major problem for originators is that when
a borrower calls asking for a general status report on their loan, the
originator might have to make several calls or check several Web sites.
Instead, an originator should be able to pull up the loan file in the
portal-connected LOS and instantly know the status of the loan from the view
of the loan processor, appraiser, underwriter, closing agent and so forth.
Additionally, when a mortgage broker locks a loan with a wholesaler, both the
wholesaler and the mortgage broker should know if at any time the interest rate
or lock expiration date differs between the wholesaler's systems and the
mortgage broker's LOS. This requires real-time data movement, something a
portal can effectively handle.
PDAs: Cold
Every few years, we get a new wave of ultra-small
computing devices. The latest is the personal digital assistant (PDA), from
companies like Palm and Handspring. While these devices are far better than
anything that preceded them, they still aren't up to performing the tasks a
typical mortgage originator needs.
For example, a new loan file can require hundreds of data elements to be
entered. On a PDA, entering this amount of
Many other issues exist as well. Loan officers need to show charts, loan
amortization schedules and loan product comparisons side by side. These are the
tools of the profession, and a PDA could never do justice considering today's
laptops. A PDA might be fine for a loan officer to keep track of appointments
and phone numbers, but every loan officer should be well equipped with a laptop
capable of handling every request a prospective borrower might make. If a loan
officer must always carry a laptop, then a
PDA loses much of its value. Keep PDAs as personal organizers, but don't expect
any real productivity gains from them for loan officers.
Broadband access: Hot
In 2002, we'll see the majority of the mortgage industry adopt broadband access
to the Internet. This could be ISDN, Digital Subscriber Line (DSL), cable
modems, T1 or even
some of the brand-new, high-speed wireless solutions via satellite. As we all
become far more reliant on the Internet for day-to-day business, we need to
ensure that the employees of the company are operating at peak performance.
Lenders just can't expect their loan processors to use a portal proficiently
when communicating at 56kbps (kilobits per second) or less. If lenders aren't
using high-speed access today, they should be soon. Internet access has now
become so much of a requirement for our industry that even small mortgage
brokers need broadband access so as not to lose money on staff inefficiency.
LOS: Hot
Some in the industry have wondered if eventually the
LOS might be replaced by solutions like automated underwriting systems (AUS)
that become much more comprehensive. There has been some hypothesizing about
this but it's not going to happen.
Instead, in my view, the LOS will continue to grow in importance as the industry
becomes more interconnected. The LOS contains the master database from which
all other parties in the industry must communicate. E-commerce will continue to
become a core function for the LOS.
LOS also involve much more than just printing loan
applications and related documentation. They are the central database for
everything a mortgage origination company does. Their uses include tracking all
customer correspondence, providing management reporting, tracking all
documentation, performing all compliance work, tracking customers' escrowed
fees, managing the staff and a variety of other
services. The LOS will continue to be the dominant application used in the
origination industry for the foreseeable future.
Electronic delivery of documents: Warm
While it will take a long time before signatures will
be digital, everything else will be. Sending documents via the Internet is a
great way to make our industry more efficient. Already, many lenders are doing
this on a regular basis, and you can expect it to go mainstream
in 2002.
There's not much to dislike other than the fact that some of these files can
really clog up an e-mail server or Internet connection. Electronic documents
(especially when using a PDF format) can get really large. Also, this is only a
one-way transaction, since the signed documents still must be physically
delivered.
ASP solutions for wholesalers/lenders: Hot
If there is one mistake many wholesalers have made in
the last couple of years, it's dragging their heels when it comes to building
fully enabled Web sites for mortgage brokers. Ironically, this has occurred
even after it has become obvious that successful implementation has had major
advantages for origination firms such as Calabasas, California-based
Countrywide Home Loans Inc. and IndyMac.
While these two companies are leading the industry, even they can make further
improvements.
Today there are several companies offering ASP solutions to
allow other wholesalers to keep up with what Countrywide, IndyMac and a few
others have done on their own. It makes sense for a wholesaler to rely on these
ASP solutions. These turnkey solutions can offer rate-lock systems, PRS
features, loan submission, risk-based pricing and the like They
will continue to thrive as wholesalers seem to need outside vendors to remain
competitive on the technology front. Perhaps it's their sheer size, but many
wholesalers seem incapable of building world-class e-commerce engines.
Conclusion
One theme of this article is the cost of failed technology to our industry and
its investors. Looking to the past, we can better predict the future. Examining
each of the past's failures, we can make smarter decisions for today and
tomorrow and avoid repeated losses.
Time and time again, history has proven the need for experienced management
with specific mortgage technology background. Firms that navigate these
perilous waters carefully can find the right solutions. Those solutions
can dramatically propel the success of their firm. Increasingly, technology decisions,
both wise and foolish, will determine a firm's ultimate fate. MB
bio:
Before joining Ellie Mae, he founded Contour Software, a
California-based provider of mortgage loan software systems, one of the one of
the earliest LOS companies. Cooley has spent the last 20 years near the center
of mortgage technology. During his
tenure in the industry, he has met with founders of virtually all significant
mortgage technology firms in the last 20 years