S&P Reviews Servicer Escrow and Call Center Administration

One problem with drawing any conclusions from the Standard & Poor (S&P) Spotlight Report on mortgage servicers is the series short history.  The first of the three reports published covered the second half of 2007, a time when mortgage delinquencies, while they hadn’t yet hit the news, were beginning to be felt by the servicing industry.  One would expect that what is now a tsunami of delinquencies, defaults, and foreclosures would profoundly affect servicer operations and S&P continues to make the point that they have.  But the data is so limited historically those trends are not dramatic or sometimes even apparent.  For that reason we will, in reporting on some of their findings, use the results of the first report rather than the more recent one which covers the first half of 2008.

One of the metrics that Standard & Poor (S&P) looked at was the companies’ call center operation.  The assessment took into consideration the amount of time it took the center to answer the phone, how long the representative and the customer talked, the percentage of people who abandoned the call before they spoke to anyone, and the utilization of the companies’ web sites.

The report states that the average length of a call was expected to rise with the increase in delinquencies and the accompanying loan modification and other workout inquiries.  The length did rise, but not significantly, from 228 seconds to 243 seconds for prime loans and from 241 to 245 seconds for subprime loans.  It would have been interesting to see statistics for a more placid period in 2004 or 2006.

It took an average of 50 seconds for a servicer to answer the phone for a prime call and 40 seconds for a subprime call.  Loan type seems to be an odd way to segment this data; size of firm would have been more informative.   In spite of the hang time, less than 5 percent of all callers abandoned the task before reaching a call center representative.

Servicers have made a considerable investment in facilitating customer service both through automated voice response technology and interactive websites.  The report found that the average capture rate (the percent of calls not requiring human intervention) of the voice response systems was 50 percent for prime loans and 41 percent for subprime.  The authors speculate that the higher capture rate for prime borrowers reflects the more complex nature of subprime calls which may more frequently be about delinquencies and workouts.

S&P defines website usage as the percent of all borrowers in the portfolio who are registered with and actively using the company’s website.  From that standpoint the financial investment appears to be paying off.  Subprime usage ranged from 32 percent in small firms (defined as those with less than 100K of prime loans or 50K in subprime loans in the portfolio) to 70 percent in large firms (more than 1mm prime or 200K subprime).  Prime usage was lowest in medium sized companies at 39 percent and highest in large ones at 54 percent. 

While this was not strongly supported by the data, the report points out that escrow accounts, while commonplace for prime loans, have been minimal in subprime portfolios but that is changing.  Escrow accounts increased from 57 percent in prime portfolios in 2007 to 67 percent in the most recent study while subprime accounts which were at 39 percent in 2007, dropped to 34 percent before rising again to stand at slightly over 40 percent.  S&P attributes the increases requirements in new loan modification terms.

One interesting item in the escrow portion of the report is the handling by servicers of lender or forced placed insurance (LPI), something that is more likely to occur with subprime than prime loans and is disproportionally used by small servicers.  In the second half of 2008 small companies placed hazard insurance on 14.4 percent of the loans in their portfolios; the numbers for medium and large companies were 11.53 percent and 8.56 percent respectively.  LPI was a toll used much less with prime loans; the figures are 4 percent, 1.66 percent, and 1.51 percent.  S&P considers a high incidence of LPI an indicator that a servicer has not implemented appropriate reviews and oversight and notes that it is a major reason for customer abuses, litigation, and headline risk.

One finding that this reporter found surprising was the renewal rate of LPI.  Over the 18-month period covered by S&P’s reports, renewals of policies ran between 20 and 26 percent for prime loans and 20 percent and 27.5 percent for subprime.  S&P interpreted this as an indication that borrowers were satisfied with LPI products which these have historically been extremely expensive, possibly because they may not be able to secure coverage on their own due to location or underwriting guidelines.