Mortgage Focus 2006: participants in Fannie Mae's latest annual benchmarking study saw the average cost to originate fall for all channels--except retail. The Internet/call-center channel posted the biggest drop in per-closed-loan origination costs.

Mortgage banking is cyclical and constantly buffeted by market, demographic and technological winds. Last year was no exception. Lenders that participated in Fannie Mae's seventh annual Mortgage Focus benchmarking study bid farewell to the days of "order-taking" their way through a queue of calls from eager refinance borrowers. Sky-rocketing home prices, narrowing spreads between adjustable-rate mortgages (ARMs) and fixed-rate mortgages (FRMs), and a shift to a purchase environment forced lenders to dust off time-honored skills of salesmanship, innovation, outreach and execution. [??] Successful lenders kept costs down and profits up by calibrating their product mix, focusing on staff and customer retention, and making strategic investments in technology. [??] Data collected for Mortgage Focus 2006 cover calendar year 2005 and, for the first time, the study includes an analysis of servicing. The origination study includes detailed information on origination costs, productivity and loan production for five channels: retail, Internet/call center, wholesale, correspondent and credit union. [??] In this year's origination study, 186 lenders participated, representing one-third of the top 100 U.S. lenders by production volume in 2005. Twenty-eight servicers participated in the pilot servicing study (see sidebar, "Spotlight Shines on Servicing"), including 10 of the top 100 servicers by portfolio size in 2005. Research focused on portfolio volume and composition, financial statements, technology costs, staffing and outsourcing.

Participating lenders use Mortgage Focus as both a benchmarking tool and blueprint for future change. "One of the reasons we find the Mortgage Focus study so useful is that it provides detailed breakdowns of income and expense statements. Using these, we can benchmark our performance against our peers, identify year-over-year changes and adjust our strategies where needed," says Rick West, senior vice president in the San Diego headquarters of Union Bank of California's mortgage division.

Market realities and lender response

As rates rose slightly in 2005, consumers were more likely to pursue purchase or equity transactions such as seconds, home-equity lines of credit (HELOCs) and cash-out refinances and less likely to pursue rate-term refinances. According to Fannie Mae's economics department, industry origination volume increased a modest 6 percent to $2.89 trillion--the second-highest figure ever. For the first time since 2000, purchase originations edged out refinances--51.1 percent of total originations versus 48.9 percent, respectively.

Affordability suffered due to soaring price appreciation and modest income growth. Accordingly, ARM share rose as home-buyers realized fixed-rate financing would price potential purchases out of reach. Lenders originated more interest-only (10) loans, payment-option ARMs and hybrid ARMs, and fewer conforming and government loans. (Note: In the Mortgage Focus study, alternative-A loans are segregated from conforming loans.)

Generally, these forays into new product offerings contributed to higher profitability. For example, among retail channel lenders for which alt-A loans represented more than 5 percent of their total product mix, profitability averaged 33 basis points--17 basis points higher than the channel average.

Successful lenders placed greater emphasis on recruiting and retaining experienced loan officers and account executives (AEs) able to sell and support expanded product menus. Technology use greatly influenced productivity in every channel. The capability to view and sign documents over the Internet, for example, significantly improved cycle times and pull-through.

Recognizing historic interest in non-owner-occupied properties, retail lenders originated 10 percent of their volume from these loans, wholesale lenders 16 percent, and even the typically refinance-oriented Internet/call-center channel originated 13 percent.

Emerging markets, too, were the focus of successful outreach strategies by many lenders. Seventy-five percent of retail top performers pursued an emerging-markets strategy, compared with 50 percent in the retail channel overall.

Performance highlights

By using their traditional mortgage banking skills, lenders were able to maintain or reduce their costs. The average cost to originate decreased in every channel except retail (see Figure 1), which increased a mere 1.48 percent. Credit unions had the highest productivity and lowest costs in the consumer-direct channels. Due to lean staffing and the nature of its business model, the Internet/call-center channel's performance began to close in on the credit union's lead (see Figure 2). Internet/call-center participants originated the highest percentage of conforming loans in the study at 73 percent (see Figure 3). Refinances accounted for 59 percent of their volume.

Correspondent lenders benefited from variable cost structures that offered flexibility in a slowing market (see Figure 4). Wholesale lenders had the highest per-loan origination costs, but realized cost improvements compared with last year due to lower yield-spread premium (YSP) and servicing-released premium (SRP) costs, fewer government and subprime loans, and wholesalers' increased use of the Internet to do loan locks and receive loan files from their brokers.

The Internet/call-center channel was by far the most profitable of all originator categories in the study, earning an average 81 basis points versus 15 basis points for retail lenders, their closest non-credit-union competitors (see Figure 5). These lenders also experienced the greatest improvement in profitability from last year, at 197 percent.

Factors that contributed to the Internet/call-center channel's remarkable profit surge included the highest percentage of conforming and refinance loans in the study, lean staffing and new originations driven by increased marketing expenditures. (see Figure 6).

With its smaller staff requirements, the correspondent channel had the highest productivity as measured in closed loans per direct full-time equivalent (FTE) (see Figure 7). Internet/call-center channel lenders were almost 50 percent more productive than retail lenders.

Wholesale lenders closed the lowest percentage of conforming loans of all channels, and their alt-A production grew substantially. Retail lenders concentrated on more complex and labor-intensive purchase transactions, which made up 65 percent of channel volume.

Business model influenced product strategy in consumer-direct channels

In the consumer-direct channels, retail lenders turned to customer retention, a broadened product menu and best-execution strategies to succeed. Internet/call-center lenders maintained a simple product menu and submitted more loans through an automated underwriting system at the point of sale.

For Mortgage Focus participant Wachovia Mortgage Finance, Charlotte, North Carolina, this meant rolling out a new mortgage processing system more closely matched to customer needs, according to Scott Muenger, senior financial analyst.

"Wachovia had a very busy year in 2005, implementing a new mortgage processing system and growing our mortgage sales force," says Muenger. "While we still have room for improvement, we were recognized by [Westlake Village, California-based] J.D. Power and Associates as third for overall customer satisfaction among 13 mortgage lenders in the recently released 2005 J.D. Power and Associates Primary Mortgage Origination study. I think this shows that focusing on our customers, despite the aggressive growth initiatives under way, has really paid off for us."

Where did channel lenders find their profitability "sweet spots?" Retail participants expanded their product horizons; those that originated more than the channel average of 19 percent of payment-option, negatively amortizing and hybrid ARMs were 21 basis points more profitable than the channel average.

Conversely, standardization drove the Internet/call-center channel's profits. Lenders that originated a higher-than-average percentage of conforming or refinance loans reported profitability greater than the channel average.

Lenders whose loan officer tenure exceeded the channel average benefited from increased originator productivity; in the retail channel, this gain was 23 percent. Competitive compensation was the top employee-retention strategy.

In the retail channel, lenders with a closing-to-delivery cycle time of fewer than seven days, compared with the channel average of nine days, closed an average of 43 loans per direct FTE--11 percent more than the channel average. Internet/call-center channel lenders that closed their customers' loans in fewer than 30 days (compared with the channel average of 41 days) were 39 percent more productive.

Wholesale and correspondent lenders sought greater efficiencies

Both wholesale and correspondent channel lenders responded to market changes with technology applications designed to increase efficiency, new products to enhance profitability and improved processing options to lower costs.

Automation and Internet delivery spurred productivity and reduced costs. In the wholesale channel, closed-loan files received via the Internet more than doubled from 11 percent to 29 percent.

Wholesale lenders that used electronic data transfer or that required brokers to use proprietary technology had lower average costs to originate than the channel average. In their search for efficiency, some lenders turned to Web-based business process management (BPM)--a technology workflow solution with the ability to escalate decision-making through routing and rules management. These lenders saw their average productivity increase from 86 closed loans per direct FTE in last year's study to 120 closed loans per direct FTE in this year's study.

Both channels (wholesale and correspondent) profited by broadening their product mix. Wholesale lenders with above-average alternative-A and subprime volume had average profitability of 59 basis points and 72 basis points, respectively, compared with the channel average of 10 basis points. For correspondent lenders, an above-average number of subprime loans generated a profit of 81 basis points compared with the channel average of -4 basis points.

Credit unions continued shift to second liens 

Key developments in the credit union channel included a shift in loan mix; evolution of a more sales-oriented culture; technology-driven process improvements; and increased outsourcing.

"As the Florida real estate market heated up and home-price appreciation accelerated in 2005, more and more of our members decided that they would prefer to invest in their current homes rather than buy up," says Kim Ray Yarnelli, vice president of real estate services for GTE Federal Credit Union, Tampa, Florida. "Fortunately, we had already created a fully automated application and underwriting process on the Internet, and were able to easily handle the transition to a HELOC and seconds market."

Changing loan mix resulted in fewer refinances and more seconds and HELOCs. HELOCs and seconds represented more than half of loan volume for nearly 60 percent of credit unions in the study. This group had 39 percent higher productivity, and its average cost to originate was 27 percent lower than the channel average. However, the average profitability for this group was just under 4 basis points--far below the 38 basis points reported by the credit unions where HELOCs and seconds volume represented less than half of their origination volume.

To foster a more sales-oriented culture, credit unions focused on competitiveness and experimented with variable compensation. Credit union lenders with above-channel average incentive costs reported average profitability of 59 basis points; conversely, credit unions that did not pay incentives reported an average loss of 13 basis points.

Internet originations for credit union lenders in the study nearly doubled from last year. Moreover, productivity among credit unions that communicated loan status to borrowers via the Internet grew 36 percentage points per FTE from last year's study.

The average outsourcing cost per closed loan rose significantly as more credit union lenders adopted outsourcing practices. On average, credit unions that outsourced any function reported net profit of 33 basis points--more than double that of lenders that kept all of their functions in-house.

Fraud concerns continued

Lenders' concerns about mortgage fraud persisted for a third straight year. Sixty-eight percent of respondents viewed mortgage fraud as a significant threat to their balance sheets and income statements (see Figure 8).

The wholesale channel expressed the greatest degree of concern and resource commitment--96 percent of channel lenders reported that fraud is a serious threat, and 68 percent allocated 1 percent or more of their budget for fraud management.

The highest usage of fraud-detection tools was reported by the wholesale (64 percent) and Internet/call-center (60 percent) channels. Fraud was less troubling for credit unions, which have the closest relationship with their customers; just 36 percent of these respondents felt fraud was a serious or important issue.

Looking ahead

In anticipation of rising rates and decreasing volume in 2006, lenders plan to emphasize service delivery, technology adoption, human resource (HR) initiatives and product mix as competitive differentiators. Retail lenders identified increasing efficiency as their most important business objective, followed closely by hiring, training and retention.

In Mortgage Focus 2006, however, lenders who ranked market share as their first- or second-strongest competitive differentiator also had considerably higher profitability than lenders who chose a different focus, showing that strategies focused on market share had the strongest correlation with profitability.

Internet/call-center participants ranked the extension of Internet capabilities as their most important business objective for the year ahead. While 90 percent of Internet/call-center participants had Web sites and actively monitored site usage, only 10 percent captured leads online. Internet/call-center channel participants ranked customer service as their most important distinguishing strategy.

More than one-third of credit unions participating in Mortgage Focus 2006 named efficiency increases and staff training as their top business objectives for 2006, while rate ranked third. Interestingly, credit unions that ranked rate as their top strategy were nearly three times as profitable as those that ranked service as their top priority.

Top business objectives for lenders in the third-party origination channels include emerging-market outreach, increased efficiency and customer service. While conforming loans generally lowered costs and increased productivity for both wholesale and correspondent channel participants in 2005, a product mix that favored subprimes and hybrid ARMs proved more profitable.

Even in an environment of diminishing volumes and higher rates, lenders can prosper. Certainly, sustainable success calls for strategies well-aligned with risk appetite, resources and business models. However, putting time-honored, entrepreneurial skills to work may well provide an extra competitive edge.

Sal Mirran is senior vice president of business and strategy development for Fannie Mae in Washington, D.C.

RELATED ARTICLE: SPOTLIGHT SHINES ON SERVICING

In an environment of price depreciation, regulatory tightening, default-prone products and fraud, servicing may be returning to its importance as a "profit protector." Clearly, servicing divisions' influence on lender profitability warrants closer evaluation. Mortgage Focus' pilot servicing study analyzed servicer cost, productivity and profitability, categorized by portfolio size (small, medium and large) as well as for servicing participants overall.

The pilot study found average total cost to service among participants was $106.69 per loan. Costs were driven by customer service (payoffs, taxes and customer inquiries), default and indirect costs. Surprisingly, small-segment servicers (those with fewer than 50,000 loans in portfolio) reported the lowest total cost to service, at $91.84 per loan. Because many smaller lenders had low-cost portfolio characteristics (plain-vanilla product types, smaller volumes, single-state portfolios), they were able to achieve both service and cost-control objectives.

The average direct cost to service--a subset of the total cost to service, which includes only costs and associated expenses for staff directly involved in the servicing function--was $61.93 per loan. Large-segment servicers (those with 100,000 or more loans in portfolio) reported the best performance, with an average direct cost to service of $59.77. The customer-inquiry function appeared to hold the greatest potential for cost savings, as there was a variance of more than $80 between high-cost and low-cost servicers.

Servicers that outperformed their peers in this study focused on effective call management and supporting technology, especially robust Web site functionality.

While large-segment servicers outperformed other servicing segments, bigger was not always better. In the default-management function, for example, 86 percent of servicers with the highest levels of productivity were in the small segment due to their low delinquency ratios.

The customer-service function had the lowest productivity of the functional areas in all servicing segments. In the small segment, 56 percent of servicers handled routine calls using a live operator instead of a voice-response unit (VRU) or Web site.

Study results indicated that the benefits of technology on servicer productivity are significant, if still largely untapped. For example, servicers in the medium segment (those with 50,000 to 100,000 loans in portfolio) that used imaging were found to have productivity nearly four times as high as servicers that maintained paper files.

Total revenue averaged $250.30 per loan for all servicers in the study. Top performers started with higher service fee revenue per loan, which offset write-downs. Servicers with larger portfolios had larger write-downs in total dollars and in average amount per loan.

The amount of ancillary income collected was proportionate to portfolio size. Large-segment servicers collected an average of $71.26 per loan in ancillary income--more than double the $32.47 per loan of ancillary income that servicers overall reported. Less than one-third of study participants reported revenue from insurance and bi-weekly programs.

Competitive rates, good service and convenience can produce customer loyalty and deter portfolio runoff. "Tracking the reasons why customers pay off is a good barometer of how competitive our rates are and how well we serve the customer," says Karen Reed, vice president for CUSO Mortgage Corporation, Hampden, Maine.

Nonetheless, only a small minority of servicers have formalized customer retention and measurement programs. Improving service metrics may help cultivate profitable portfolios--the top priority of the majority of this year's servicing participants.

RELATED ARTICLE: TOP PERFORMERS FOCUSED ON INFORMATION TECHNOLOGY INVESTMENTS, STAFF EXPERTISE

This year, Mortgage Focus includes a "top performer" category. Each channel review of top performers analyzed the performance and strategies of low-cost, high-productivity (LCHP) lenders--models of efficiency that outperformed 75 percent of their peers in cost and productivity--and highly profitable LCHP lenders and servicers.

Retail and wholesale top performers relied on the Internet more extensively than their peers. One-third of retail channel top performers offered electronic signing of loan disclosures, compared with 9 percent of the overall channel.

Average staff tenure for retail channel highly profitable LCHP lenders was 46 percent higher than the channel average. Wholesale channel top performers had average account executive (AE) tenure that was more than twice that of their peers.

Figure 1 All Channels Emerged Unscathed

                      Average Cost to Originate Per Closed Loan
                      Mortgage Focus 2005  Mortgage Focus 2006

Internet/Call Center  $2,920               $1,958
Retail                $2,703               $2,743
Credit Union          $1,206               $1,081
Correspondent         $2,588               $2,512
Wholesale             $3,925               $3,581

SOURCE: FANNIE MAE

Note: Table made from bar graph.

Figure 2 Credit Union and Internet/Call-Center Channels Were Least
Costly

                                                 Average Cost to
                      Average Cost to            Originate Per Closed
                      Originate Per Closed Loan  Loan (basis points)
                      Mortgage Focus 2006        Mortgage Focus 2005

Internet/Call Center  129                        $1,958
Retail                176                        $2,743
Credit Union           99                        $1.081
Correspondent         139                        $2,512
Wholesale             177                        $3,581

SOURCE: FANNIE MAE

Note: Table made from bar graph.

Figure 3 Conforming Loans Dominated Internet/Call-Center Channel Volume

                         Percentage of Total Loan Count
               Internet/Call          Credit
Product Type   Center         Retail  Union   Correspondent  Wholesale

Conforming     73%            61%     53%     69%            44%
Government      2%             5%      1%      9%             6%
Jumbos          4%             6%      3%      4%             9%
Alternative-A   8%             4%      0.5%    9%            22%
Subprime        0%             2%      0.5%    1%             2%
Seconds         5%             7%     17%      3%             9%
HELOC           7%             8%     23%      3%             5%
Other           1%             7%      2%      2%             3%

SOURCE: FANNIE MAE

Figure 4 Correspondent Lenders Benefit From Variable Cost Structures

                          Average Cost Per Closed Loan
                  Internet/
                  Call               Credit
                  Center     Retail  Union   Correspondent  Wholesale

Direct Personnel  $882       $1,363  $663      $357           $808
Incentives        $239         $531  $159       $69           $242
YSPs and SRPs                                $1,664         $1,841
Other Direct      $385         $475  $203      $141           $361
Indirect          $500         $534  $170      $303           $286

SOURCE: FANNIE MAE

Note: Table made from bar graph.

Figure 5 Internet/Call-Center Channel Was Most Profitable

                      Channel Profitability Basis Points

Correspondent         -4
Wholesale             10
Retail                15
Credit Union          24
Internet/Call Center  81

SOURCE: FANNIE MAE

Note: Table made from bar graph.

Figure 6 Internet/Call-Center Channel Enjoyed Greatest Profit
Improvement

                                  Basis Points
                      Mortgage Focus 2005  Mortgage Focus 2006

Correspondent         -0.16                -4
Wholesale              6                   10
Retail                34                   15
Credit Union          51                   24
Internet/Call Center  27                   81

SOURCE: FANNIE MAE

Note: Table made from bar graph.

Figure 7 Correspondent Business Model Allowed For Greater FTE
Productivity

                         Closed Loans Per Direct FTE
             Internet/Call          Credit
             Center         Retail  Union   Correspondent  Wholesale

Direct FTE      61             39      95     180             86
Originator     196            102     399
Account                                     1,426            466
Executive
Processor      315            234     435   2,454            771
Underwriter    806            709   1,490   2,681            522
Closer         988            645   1,008   2,166            685
Post-Closer  1,181          1,214   1,516   1,897          1,668
Shipper      3,070          1,895   2,552   1,478          2,529
Secondary    1,933          1,999   3,564   4,086          3,322
Marketing

SOURCE: FANNIE MAE

Note: Table made from bar graph.

Figure 8 Seriousness of Mortgage Fraud

Serious/Very Important Issue  68%
A Somewhat Important Issue    15%
A Minor Issue                 10%
Not an Issue                   7%

SOURCE: FANNIE MAE

Note: Table made from pie chart.