Summary

California Department of Veterans Affairs: Changing Demographics and Limited Funding Threaten the Long-Term Viability of the Cal-Vet Program While High Program Costs Drain Current Funding May 2000 99139 rotiduA etatS ainrofilaC S T I D U A E T A T S F O U A E R U B The first five copies of each California State Auditor report are free. Additional copies are $3 each, payable by check or money order. You can obtain reports by contacting the Bureau of State Audits at the following address: California State Auditor Bureau of State Audits 555 Capitol Mall, Suite 300 Sacramento, California 95814 (916) 445-0255 or TDD (916) 445-0255 x 216 OR This report may also be available on the World Wide Web http://www.bsa.ca.gov/bsa/ Alternate format reports available upon request. Permission is granted to reproduce reports. C S A ALIFORNIA TATE UDITOR MARY P. NOBLE STEVEN M. HENDRICKSON ACTING STATE AUDITOR CHIEF DEPUTY STATE AUDITOR May 25, 2000 99139 The Governor of California President pro Tempore of the Senate Speaker of the Assembly State Capitol Sacramento, California 95814 Dear Governor and Legislative Leaders: As requested by the Joint Legislative Audit Committee, the Bureau of State Audits presents its audit report concerning the Department of Veterans Affairs’ Farm and Home Purchase Program (Cal-Vet program). This report concludes that the Cal-Vet program will substantially lose its ability to offer low-cost home loans to veterans beyond the next decade due to the restrictive eligibility requirements attached to one source of program funding and limited availability of its two other funding sources. In addition, the Department of Veterans Affairs is eroding Cal-Vet program funds by charging the program for unrelated administrative costs and by operating the program less efficiently than it could. Respectfully submitted, MARY P. NOBLE Acting State Auditor BUREAU OF STATE AUDITS 555 Capitol Mall, Suite 300, Sacramento, California 95814 Telephone: (916) 445-0255 Fax: (916) 327-0019 CONTENTS Summary 1 Introduction 7 Chapter 1 Shrinking Eligibility and Diminishing Funding Threaten the Cal-Vet Program 17 Recommendations 25 Chapter 2 Poor Cost Controls and Inefficient Administration Erode Funds and Delay Loan Applications 27 Recommendations 42 Chapter 3 Inadequate Management of the New Information System Increases Costs and Creates Doubt About the Reliability of Program Data 43 Recommendations 54 Appendix Cal-Vet Program Benefits 57 Response to the Audit California Department of Veterans Affairs 61 California State Auditor’s Comments on the Response From the California Department of Veterans Affairs 101 SUMMARY RESULTS IN BRIEF A rapid decline in the population of eligible California veterans and limited funding threaten the long-term viability of the California Veterans Farm and Home Audit Highlights . . . Purchase Program (Cal-Vet program) of the Department of Veterans Affairs (department) while high program costs erode Our review of the California current funding. Because federal restrictions severely limit Veterans Farm and Home eligibility for the Cal-Vet program’s major source of funding for Purchase Program revealed that: loans, proceeds from tax-exempt Qualified Veterans Mortgage Bonds (QVMBs), demand for these loans will drop dramatically (cid:1) By the end of the decade, over the next 10 years. The program has two other sources of eligibility for one type of loan and the limited funds funding, Qualified Mortgage Bonds (QMBs) and unrestricted available for the two funds, but approval to issue QMBs is difficult to obtain and remaining types of loans unrestricted funds are drying up. At the same time, excessive will severely diminish the program costs are eroding available loan funds due to improper program’s value to most veterans. charges of administrative costs, lengthy loan processing times, and mismanaged implementation of an integrated information (cid:1) Poor budget controls, management system. Coupled with the program’s short-term improper administrative lending strategy, these conditions will dramatically diminish its charges, and inefficient and inconsistent value to most veterans by the end of the decade. operations have driven up program costs and further The Cal-Vet program provides loans to thousands of qualified eroded funds otherwise veterans at below-market interest rates. Federal income tax available for loans. restrictions limit eligibility for the more plentiful QVMB loans to (cid:1) Mismanagement of the veterans who actively served prior to January 1, 1977, and apply implementation of a new for their loans within 30 years after leaving the military. We integrated information system resulted in its project these restrictions will cause a steep 90 percent drop in failure to meet the needs eligibility over the next 10 years and that in 2010 only about of the program without 100 of the remaining eligible veterans will actually apply for a an additional investment loan based on the Cal-Vet program’s current market share. The of time and program funds. department has lobbied Congress over the years to modify these restrictions, but it has so far been unsuccessful. Younger veterans who do not meet the requirements of QVMB loans may be eligible for two other types of loans: those backed by proceeds from QMBs and loans from the department’s unrestricted funds. Both of these loans rely on smaller pools of funding, which are also provided by the sale or refinancing of tax-exempt bonds. The department’s current lending strategy is to increase the total value of its loan portfolio. For the eight-month period of July 1999 through February 2000, the C A L I F O R N I A S T A T E A U D I T O R 1 program loaned $361 million, $25 million above its goal for the entire fiscal year. During this period, the program charged 5.95 percent interest for QMB loans and 6.65 percent for both QVMB and unrestricted loans. Because the program’s interest rates are as much as 2 percent below market interest rates, it is attracting many loan applicants; however, the frequency at which the department is now making loans will substantially exhaust the available QMB and unrestricted funds by 2006, with only residual recycled principal and interest from unrestricted funds available for loans. In light of the limited sources of funding available for two of the three types of loans that the Cal-Vet program grants, the department should consider other strategies that would benefit more veterans over a longer period of time. Rather than offer loans at the lowest possible interest rates to eligible veterans, the department could raise interest rates to follow rises in market rates, keeping them slightly below market, or provide additional funding by blending taxable and tax-exempt bond proceeds to grant loans at rates higher than they are now but less than commercial rates. These strategies maintain the program’s loan portfolio but would depress current demand for loans and preserve funds to enable more veterans to benefit from the program over time. Unfortunately, these options are limited in their continued ability to provide below-market interest rates and, therefore, are not a permanent solution to the continued viability of the program. For the department to continue to grant below-market interest rate mortgage loans to California veterans over the long term, the Legislature would need to create a new program funded by a new source. Additional concerns in the Cal-Vet program are poor budget controls and a lack of consistency and efficiency in program operations. Improperly charged administrative expenses and inefficient loan processing deplete the already limited funds available for loans to veterans. Most significantly, department records indicate as much as $1.3 million of Cal-Vet program funds in a single year were paid for the costs of administrative staff who do not provide service to the program or for staff whose service to the program has not been documented. Based on a 1974 decision by the California Court of Appeal, the department has violated state bond law by charging the Cal-Vet program for such costs. 2 C A L I F O R N I A S T A T E A U D I T O R Over the last four years the department has improved its efficiency in many areas. These improvements include centralizing loan contract servicing, adopting new loan underwriting standards, instituting mortgage insurance, and improving its management of delinquent and foreclosed loan contracts. However, the department has not fully implemented other reengineering changes in the Cal-Vet program that it has identified as necessary to efficiently operate the program. Because the department has not completed its reengineering efforts, which include the centralization of its loan processing operations and implementation of workload standards for its field and headquarters offices, the average cost to process loan applications has increased, costs vary significantly by field office, and loan applications take longer to process than is common in the industry. Another obstacle the department faces in controlling excessive program costs is implementing the Cal-Vet program’s integrated information system (system). This system is intended to provide reliable program and financial information needed to operate the Cal-Vet program. However, even though the department has devoted significant time and money to get the system running, the system still does not meet its needs. The department has not tracked expenditures in a way that allow it to control costs by comparing its actual spending to bring the system on-line to its $2.8 million budget for these one-time costs. Moreover, the significant resources being spent on this extended project eat away at funding the department could otherwise use to support loans to veterans. The implementation project has also been marred by problem- atic management. When key staff left in the middle of the project, management abandoned its original implementation plan and did not ensure staff adhered to prudent project imple- mentation practices. Fifteen months after its original target date for implementation, the department must spend additional program resources to modify this system further to prevent errors. Even so, it cannot be certain that the system will properly maintain borrowers’ file records and accurately accumulate program and financial data because it has not completed necessary testing. It failed to complete over half of the 358 tests it had originally planned to ensure the system worked properly, and it did not maintain adequate documenta- tion of the conversion of loan files from the old system to the new. The department’s continuing efforts to implement the system disrupted the Cal-Vet program’s normal operations C A L I F O R N I A S T A T E A U D I T O R 3 and caused additional workload for staff to correct errors in the accounting records and in borrowers’ files maintained in the system. Furthermore, the department has not adequately safeguarded the data stored in its system by following prudent procedures for approving, testing, and documenting changes to the system software, or provided adequate security over unauthorized system access to prevent the loss or misuse of information in the system. Adhering to proper procedures for making changes to system software and restricting employee access to the system is necessary to ensure all alterations are authorized and correct and to safeguard the program’s information and assets. Most of the shortcomings we noted in the department’s efforts to implement the system can be traced to a lack of commitment on the part of management to ensure the implementation was successful. To date, the department remains unable to generate the program and financial information it needs to operate the program effectively and efficiently. RECOMMENDATIONS The department should determine how to use its remaining limited funding to best serve California veterans in purchasing farms and homes. If it decides to continue its present strategy of using available funds to provide loans at the lowest possible interest rates, it should plan its operations for the future curtail- ment of new loan activity. If the department determines that veterans are best served with interest rates closer to market rates and expands its pool of funds accordingly with alternate financ- ing methods, it should maintain current demographic data to identify veterans eligible for, and likely to participate in, the Cal-Vet program and adapt the program to provide home loans to the greatest number of qualifying veterans for as long as possible. In the absence of sufficient tax-exempt financing to ensure the continued viability of the Cal-Vet program, the Legislature should consider using state funds to establish a new program to aid California veterans in purchasing farms and homes. 4 C A L I F O R N I A S T A T E A U D I T O R To stop further erosion of Cal-Vet program funds, the depart- ment should complete the steps necessary to ensure its direct and indirect administrative costs are properly and equitably charged to all programs served by administrative staff. In addition, the department should identify the amount of Cal-Vet program funds it has used for activities outside the program and seek reimbursement from other appropriate state funds. To further increase the efficiency and consistency in the Cal-Vet program’s operations, and thereby reduce costs and improve loan processing times, the department should complete its reengineering efforts. The department should convene a centralized implementation team to ensure that its system functions reliably. The team should include a project manager and a sponsor from the department’s executive management team with the authority to allocate the necessary resources. Additionally, the department should contract with an outside consultant with experience in project management to oversee the team. The team should gather all data from prior implementation efforts, assess which tasks remain incomplete, and identify the steps needed to properly test the modules and system. In addition, the team should obtain the training necessary to design the program performance reports and financial reports the department needs to efficiently and effectively operate the Cal-Vet program. The department should adequately safeguard program data and assets by implementing a security policy to limit system access to employees who are properly authorized and ensuring access is not incompatible with their other duties. Further, the depart- ment should follow proper procedures for making changes to system software. AGENCY COMMENTS While it disagrees with some of the conclusions we reached, the California Department of Veterans Affairs (department) concurs with the recommendations we made for Chapters 1 and 2 but believes it has adequately addressed our recommendations for Chapter 3. To provide clarity and perspective, our comments follow the department’s response. (cid:1) C A L I F O R N I A S T A T E A U D I T O R 5 Blank page inserted for reproduction purposes only. 6 C A L I F O R N I A S T A T E A U D I T O R INTRODUCTION BACKGROUND S ince its inception in 1921, the California Department of VeteransAffairs (department) has assisted thousands of California veterans by providing low-cost farm and home loans through its California Veterans Farm and Home Purchase Program (Cal-Vet program). The Cal-Vet program is adminis- tered by the department’s Farm and Home Purchases Division (division) and is governed by the Veterans Farm and Home Purchase Act of 1974. As of June 30, 1999, the program had a portfolio of 32,243 loans totaling over $2 billion. In addition to the Cal-Vet program, the department currently administers the Veterans Services Program and the Veterans Homes Program, as shown in Figure 1. The department’s law, personnel management, information services, and administra- tive services divisions support all three programs, while its internal audits unit reviews their operations. As depicted in Figure 2, the Cal-Vet program is organized into three major sections: headquarters operations, planning and program operations, and statewide operations, which comprises the program’s 10 field offices. The California Veterans Board (board) and the California Veterans Finance Committee (finance committee) oversee the Cal-Vet program. A seven-member board is responsible for determining department policies; with the governor appointing six members and the department secretary serving as the seventh member. The finance committee is responsible for reviewing and authorizing bond sales to finance the Cal-Vet program. The governor, state controller, state treasurer, director for the Department of Finance, and the department secretary sit on the finance committee. In comparison to other government lending programs and conventional lenders, the Cal-Vet program offers several benefits, including below-market interest rates and low premi- ums on fire, hazard, and disaster insurance. See the Appendix for a comparison of the Cal-Vet program’s offerings to those of other lenders. C A L I F O R N I A S T A T E A U D I T O R 7 FIGURE 1 California Department of Veterans Affairs Office of the Secretary California SECRETARY Inspector General Veterans UNDERSECRETARY Board Women Veterans Home Law Office Operations Veterans Affairs Programs Veterans Internal Information Business Bond Cal-Vet Services Audits Services Services Finance Program Program See Figure 2 Legislative/ Administrative Personnel Public Affairs Services Management Accounting Office Reimbursements Accounting Systems Veterans Loan Accounting Contract Management Budgets 8 C A L I F O R N I A S T A T E A U D I T O R FIGURE 2 Cal-Vet Program Cal-Vet Program Headquarters Planning and Program Statewide Operations Operations Operations Escrow Unit Marketing Unit Customer Services Unit District Offices Management Analysis/ Home Protection Unit Satellite Offices Administrative Services Unit Collections/Foreclosure Unit Loan Processing Unit Data Processing Liaison Unit Loan Contract Services Unit PROGRAM ELIGIBILITY REQUIREMENTS AND FUNDING LIMITATIONS The eligibility requirements for the Cal-Vet program are estab- lished by state and federal restrictions on the three sources of program funding. These sources are proceeds from the sale of Qualified Veterans Mortgage Bonds (QVMBs), Qualified Mort- gage Bonds (QMBs), and from “unrestricted” funds. Loans from all three sources are limited to veterans who have served at least 90 consecutive days of active duty, have received an honorable discharge, or are currently serving in the military. Table 1 shows other applicant eligibility requirements. As veterans apply for loans, the department matches their eligibility against the most restrictive bonds, QMBs, first. Those veterans who do not meet QMB requirements are then matched against QVMB require- ments. Those who do not meet either QVMB or QMB require- ments are funded from the unrestricted pool of funds. C A L I F O R N I A S T A T E A U D I T O R 9 TABLE 1 Distinguishing Eligibility Requirements for Each Program Funding Source Eligibility Requirements QVMB QMB Unrestricted Must be a first-time home buyer (cid:2) Must have separated from service less than 30 years ago (cid:2) Must have service prior to January 1, 1977 (cid:2) Must be a wartime veteran (cid:2) (cid:2) Must meet certain income limits based on region (cid:2) Must not exceed certain property purchase price limits based on surrounding area (cid:2) California is one of only five states in the nation eligible to sell tax-exempt QVMBs. To sell these bonds and make loans with the proceeds, the department must first get the approval, or the authority as it is called, from California voters by placing a measure on the ballot. The department currently has remaining voter authority to sell $176 million in QVMBs and plans to place a bond measure on the November 2000 ballot to obtain the authority to sell an additional $500 million. The department also makes home loans to veterans using the proceeds from QMBs. However, to sell QMBs it must have permission from the California Debt Limit Allocation Commit- tee (CDLAC). CDLAC was created in 1985 in response to the 1984 Tax Reform Act, which imposed an annual limit on the dollar amount of tax-exempt private activity bonds (including QMBs for the Cal-Vet program) that may be issued in the State. CDLAC allocates approximately 20 percent of all the QMBs the State is authorized to sell each year for single-family housing among a variety of lenders. Once the department loans out the QMB funds it has plus the proceeds from bonds it is currently authorized to sell, it will need to obtain an additional allocation from CDLAC to sell more bonds. Because QMB-backed loans are 10 C A L I F O R N I A S T A T E A U D I T O R available to the public as well as to veterans, in order to obtain the allocation to sell QMBs from CDLAC, the department must compete for the ability to sell these bonds with nonprofit, local, county, and other state agencies. Unrestricted funds are also used to make loans and consist of proceeds from the refinancing of high-cost tax-exempt debt issued prior to January 1, 1981. The restrictions on these funds are less stringent, which allows the department more leeway in interpreting eligibility for loans backed by these funds. The department can only add to this source of funds using prepay- ments and repayments of loans made from these funds; it cannot generate more through future bond financing. Once it pays the related debt-service costs of the refinanced debt, the department can use any residual money from prepayments and repayments of loans made with unrestricted funds to replenish the fund source indefinitely; however, the funds available for this source of loans to veterans will diminish gradually because of the debt service commitments. The Cal-Vet program is funded primarily through the sale of tax-exempt bonds and related interest and operating revenues. Because the interest earned on tax-exempt bonds is not subject to state and federal income tax, bond purchasers are willing to accept a lower rate of return on their investment. The Cal-Vet program benefits from selling tax-exempt bonds because the lower interest rates it pays to investors allows it to lower interest rates for veterans. Tax-exempt bonds retain this status only when the department uses the proceeds for loans complying with various federal restrictions. Table 2 shows additional state and federal limita- tions on the funds the department may raise. MORTGAGE INTEREST RATES The interest rates the program charges veterans are approved by the board and the finance committee. The department uses loan principal and interest to pay the debt that finances the loans and to cover the program’s operational costs. Prior to January 1, 1999, state law required the department to charge a uniform interest rate for both new and existing loans. Statutory changes to the law now enable it to charge rates on new loans made after this date that are more closely aligned with market rates. Current laws allow the department to adjust interest rates C A L I F O R N I A S T A T E A U D I T O R 11 TABLE 2 Cal-Vet Program Funding Sources and Limitations QVMB QMB Unrestricted Source of Funds Proceeds from the sale Proceeds from the sale of Original pool of funds of QVMBs, plus any QMBs, plus any prepay- created by savings prepayments and ments and repayments associated with replacing repayments remaining remaining after debt- high-cost debt with low- after debt-service service commitments cost debt by refinancing commitments have have been met. pre-1981 general been met. obligation bonds; plus prepayments and repayments of loans made with unrestricted funds. Federal Funding Limitations The department is Each state can sell a With refinancings, restricted to selling no limited amount of maximum lifetime of more than $340 million revenue bonds per these bonds is 48 years. in QVMBs per year. year—amount is calcu- With refinancings, lated by multiplying maximum lifetime of federal per-capita these bonds is amount ($50 in 1999) 48 years. times that state’s population. State Funding Limitations Because QVMB bonds CDLAC allocates the N/A are general obligations State’s QMB allotment to of the State, the various local, county, and department must state organizations and obtain voter approval agencies. Because QMBs for the authority to sell are backed by the these bonds. residual equity of the program, the depart- ment must obtain approval from the finance committee to sell the bonds. With refinancings, maximum lifetime of these bonds is 32 years. within certain ranges without the board’s and finance committee’s approval if the department receives approval of the method from these two entities. To reduce the number of veterans who were turning to more competitive lenders for refinancing and to attract more borrow- ers, the department lowered its interest rates. In April 1998, because of the savings associated with refinancing some of its 12 C A L I F O R N I A S T A T E A U D I T O R high-cost debt with lower-cost debt, the department was able to lower the rate on loans made prior to January 1, 1999, to 6.95 percent. On those loans made after January 1, 1999, the department charges 5.95 percent for QMB-backed loans and 6.65 percent for loans financed with unrestricted funds or QVMBs. The department’s goal is to meet the statutory limit on bond yield—earnings above the interest rate it is paying on the bonds it has sold—while maintaining a benefit to veterans by offering below-market interest rates. In determining interest rates, the department must consider federal income tax restrictions, which prohibit it from earning more than 1.125 percent above the bond yield from its investments, including interest income from loans to veterans. In calculating the limit, the department can “blend,” or average, the cost of all of its outstanding bonds to arrive at a single interest rate for loans. The department’s stated goal is to establish rates that are as close as possible to 1.125 percent above bond yield, and 0.5 percent or more below commercial rates. Before the department adjusted the interest rates, its fiscal consultants prepared cash flow analyses to verify that the new rates were feasible and would not drain the department’s funds. The consultants projected that the Cal-Vet program’s fund balance would drop by $201.8 million over the next 10 years, an amount that still enables the department to meet its debt-service commitments. THE LEGISLATIVE ANALYST OFFICE’S 1998 REPORT ON THE CAL-VET LOAN PROGRAM In 1998, the Legislative Analyst’s Office (LAO) issued a report titled Rethinking the Cal-Vet Loan Program. The report made the following key points: • The program was not competitive with other public or private sector lenders. • Past mismanagement of the program had eroded the earnings retained by the program. • The department had initiated a program reform effort that had merit but still raised concerns. C A L I F O R N I A S T A T E A U D I T O R 13 • The need for Cal-Vet loans was declining while the need for other veterans’ services was growing. The report recommended that, in the short term, the Legislature not only strengthen oversight of the program but also restruc- ture it so future Cal-Vet loans could be issued at rates different from those established for previous borrowers. For the long term, the LAO recommended that the Legislature amend state law to phase out the Cal-Vet program by the year 2007 and, subject to voter approval, direct any surplus funds to programs benefiting aging war veterans and state taxpayers. Based primarily on the LAO’s findings, the governor requested and the Joint Legislative Audit Committee approved a fiscal- and program-compliance audit of the Cal-Vet program conducted by the Bureau of State Audits. SCOPE AND METHODOLOGY The Joint Legislative Audit Committee (committee) requested the Bureau of State Audits to conduct a fiscal- and program- compliance audit of the Cal-Vet program. Specifically, the committee was concerned about the program’s future, given the aging population of eligible veterans and the conclusions in the LAO’s 1998 report that the program was not competitive with other loan programs. To determine the future ability of the Cal-Vet program to make loans, we obtained data from the U.S. Department of Veterans Affairs (VA) on California veterans. We compared this data to demographic information extracted from the Cal-Vet program’s loan database. To assess the amount of funds available for these loans, we obtained cash flow projections from the department’s bond finance unit. In addition, we received infor- mation from the program about the history of its loan origina- tions. To understand the program’s loan strategy, we interviewed its staff about their activities and reviewed the department’s business and strategic plans. We also contacted other housing authorities to learn of alternative funding methods employed around the country. During our fieldwork, we reviewed and evaluated relevant laws, rules, and regulations, as well as previous audit work and related reports. To determine the financial condition of the program, we interviewed the department’s independent financial auditor and 14 C A L I F O R N I A S T A T E A U D I T O R reviewed the audit work it performed on major accounts in the department’s records. We found that for those major account balances as of June 30, 1999, we agreed with the auditor’s con- clusions and the presentation of those accounts in the audited financial statements for fiscal year 1998-99. In addition, we verified whether direct and indirect Cal-Vet program expenditures were reasonable and appropriate. We also evaluated the success of the department’s reengineering efforts in improving program efficiency by reviewing Total Quality Management (TQM) reports and interviewing department staff to measure the extent to which the recommendations for improvement had been implemented. Furthermore, we analyzed data on loans originated since July 1, 1999, to help us assess the department’s performance. We reviewed the department’s lending procedures to determine whether it prohibited unnecessarily risky loans. In addition, we tested a sample of loans to see whether the department adheres to appropriate regulations and its lending standards when granting loans. Because the program is currently a VA-guaranteed lender for many of its loans, we also spoke with VA staff, who review the Cal-Vet loans that the VA guarantees, to obtain their opinion on the quality of the program’s underwrit- ing. We found that the department follows its lending guidelines and the guidelines of its loan guarantors when granting loans to veterans. To determine whether the department exercised adequate controls when implementing its new integrated information system, we obtained and reviewed documents relating to system implementation. We also interviewed department staff and system consultants to obtain an understanding of the project management methodology, progress reporting, change management procedures, security issues, and payment records related to the implementation effort. (cid:1) C A L I F O R N I A S T A T E A U D I T O R 15 Blank page inserted for reproduction purposes only. 16 C A L I F O R N I A S T A T E A U D I T O R CHAPTER 1 Shrinking Eligibility and Diminishing Funding Threaten the Cal-Vet Program SUMMARY T he future of the California Veterans Farm and Home Purchase Program (Cal-Vet program) is uncertain due to a projected decrease in the number of veterans eligible for its largest source of loan funds and to the diminishing amounts currently available from its other two loan sources. The Cal-Vet program can offer below-market mortgage rates to qualified veterans because the California Department of Veterans Affairs (department) has the authority to sell tax-exempt bonds to interested investors. However, during the next decade, fewer and fewer veterans will benefit from the Qualified Veterans Mortgage Bonds (QVMBs) that provide the program’s major funding for loans, since federal income tax restrictions have limited eligi- bility for loans backed by these bonds to a shrinking population of older veterans. Based on the average age of veterans meeting the requirements, we estimate that the number of veterans eligible for these loans will drop 90 percent by 2010. For the veterans who are still eligible at that time, we project that as few as 100 a year will actually apply for a loan based on the Cal-Vet program’s current market share. Although QVMBs finance a large portion of the farm and home loans the Cal-Vet program currently offers, the program is still able to assist younger veterans not meeting requirements for these loans through two other sources of funding: Qualified Mortgage Bonds (QMBs), which are also tax-exempt bonds, or the department’s unrestricted funds. However, both these sources of funds are limited and, at the current rate with which the department is granting loans backed with these resources, the amount available will disappear by 2006. If the department rethinks its current lending strategy, it may be able to extend the life of the Cal-Vet program beyond the next decade. One course it could take is to slightly increase the interest rates it offers for loans backed by QMBs or unrestricted funds. The department could also add to its scarce funding for C A L I F O R N I A S T A T E A U D I T O R 17 these types of loans by selling taxable bonds and blending their proceeds with proceeds from its traditional tax-exempt bonds. Under these alternatives, the program could serve a greater number of veterans for a longer period of time while still providing them with lower interest rates than they could find elsewhere. However, these options will not provide veterans with lower interest rates indefinitely and, therefore, are not a permanent solution to the continued viability of the program. If the department is to continue providing loans at below-market interest rates to California veterans beyond the next few years, the Legislature will need to create a new source of program funding. CAL-VET LOANS WILL BENEFIT FEW VETERANS AFTER 2010 Federal restrictions on loans backed by QVMBs, currently a significant source of funds for the Cal-Vet program, will reduce demand for these loans over the next decade. Based on the Cal-Vet program’s current market share, we estimate the number of veterans eligible for QVMB-backed loans will drop by almost 90 percent during this period, leaving just over 100 veterans that year who might still apply for loans. Although many veterans are eligible for loans backed by the department’s two other funding sources, QMB proceeds and unrestricted funds, limited money is available. At the frequency with which the department currently grants these loans, the money will be exhausted within six years. As depicted in Table 3, declining eligibility, and therefore demand, for QVMB-backed loans, coupled with limited QMB and unrestricted funds, threaten the program’s long-term operations. TABLE 3 Factors Affecting the Cal-Vet Program QVMB Loans QMB Loans Unrestricted Loans Eligible population Rapidly declining; loss of Significant; likely to Nearly all wartime of veterans 90 percent by 2010 remain high veterans in the State Availability of funding Voters can approve more At current rates, all At current rates, all for loans funding available money will be available money will be lent out by 2003 lent out by 2006 18 C A L I F O R N I A S T A T E A U D I T O R Cal-Vet’s Largest Loan Program Faces Declining Veteran Eligibility Historically, the Cal-Vet program has granted more QVMB- backed loans than any other type. From fiscal years 1984-85 through 1996-97, these loans accounted for 69 percent of all Cal-Vet loans. Even with the introduction in fiscal year 1996-97 of loans backed with unrestricted funds, QVMB- backed loans still accounted for 45 percent of new loans as of December 31, 1999. California voters can continually replenish QVMB funds by approving the sale of additional bonds. However, since July 1984, federal requirements have limited QVMB-backed loans to those veterans who had at least one day of active service before January 1, 1977, and apply for their loan within 30 years of their discharge from active service. Because the department has not done a detailed analysis to project how quickly veterans in the State are losing QVMB eligibility, we estimated the reduction using projections of the veterans’ population from the U.S. Department of Veterans Affairs and demographic information from the Cal-Vet program integrated information system, including data on the percentage of veterans who currently have loans with the program. As shown in Figure 3, by 2010, federal requirements on period of active service and timeliness of loan applications will cause nearly 90 percent of the 618,000 qualified veterans in California to lose their eligibility for QVMB-backed loans. Of the 69,000 veterans who remain eligible in 2010, we project that just over 100 will actually apply for a loan based on the current Cal-Vet market share. While Congress could loosen federal tax laws and increase the number of veterans eligible for QVMB-backed loans, attempts to change these laws have thus far proven unsuccessful. Attempts to Change Laws Restricting Eligibility for QVMB Funds Are Thus Far Unsuccessful The overall implication on the Cal-Vet program from a decline in the number of eligible veterans has prompted the department to work to change existing federal laws to expand the eligible veteran base for QVMB-backed loans. The department, along with veterans’ representatives from the four other states that may award QVMB-backed loans, is encouraging Congress to pass federal legislation allowing five states to expand the definition of a “qualified veteran.” However, the current version of the bill, HR 1215, is still in the House Ways and Means Committee. C A L I F O R N I A S T A T E A U D I T O R 19 FIGURE 3 The Number of Veterans Eligible for QVMB-Backed Loans Will Decline Rapidly 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0 20 C A L I F O R N I A S T A T E A U D I T O R snareteV elbigilE 2 0 0 0 2 0 0 2 2 0 0 4 2 0 0 6 2 0 0 8 2 0 1 0 2 0 1 2 2 0 1 4 2 0 1 6 2 0 1 8 2 0 2 0 2 0 2 2 2 0 2 4 2 0 2 6 2 0 2 8 2 0 3 0 Year According to a senior legislative assistant for the representative who introduced the bill, the bill likely will not pass out of the committee any time soon. A similar bill the five states backed in 1997, HR 1241, was introduced in Congress but died in committee. QMB and Unrestricted Funds Are Limited Although the QVMB-backed loans face declining eligibility In just six years, the and demand, the two other sources of funds for the types department will exhaust of loans the Cal-Vet program offers, QMBs and unrestricted its two sources of loans funds, lack sufficient financial support to continue to benefit for younger veterans. veterans with low-interest loans for more than a few years. As of January 31, 2000, the program had $140.1 million available for first-time home buyers through QMB-backed loans. According to its monthly projections, the Cal-Vet program expected that all available QMB funds would be loaned out by May 2003. The source of unrestricted loans for wartime veterans is also limited. While many veterans are eligible for unrestricted loans, as of January 31, 2000, only $53 million was available to make these loans. In February 2000, a consultant to the program indicated that it would exhaust its current supply of unrestricted funds in three to six years. As stated in the Introduction, because the California Debt Limit Allocation Committee (CDLAC) determines the amount of QMBs available to housing programs in the State, the depart- ment must compete for the ability to sell these bonds with other nonprofit and government agencies, such as the California Housing Finance Agency (CHFA) and local housing authorities. All of these organizations provide loans with similar federal restrictions, although, unlike the department, these other entities are not confined to serving only veterans. The depart- ment has not assessed the difficulty of obtaining additional allocations from this state committee. According to CDLAC, the department would probably receive some of the allocation to sell additional QMBs if requested. However, any increase it would receive would reduce the allocations given to competing lenders. Because CDLAC does not have the power to increase the total QMBs the State is authorized to sell each year to provide loans for single-family housing, it can only shift the yearly allocation for this type of QMB among all of the housing programs in the State. The department reserves unrestricted funds for wartime veterans Having already who do not qualify for the more plentiful QVMB funds. Funding refinanced all of its old for unrestricted loans originally came from the refinancing of bonds, the department general obligation bonds sold before 1981. However, the depart- has no way to create ment has already refinanced all of its pre-1981 bonds, making it additional unrestricted impossible to create additional unrestricted funds. Although funds. many younger veterans are currently eligible for them, as of January 1, 2000, the program had only $53 million in unre- stricted funds to make this type of loan. At the rate it currently grants these loans, its consultant estimates the Cal-Vet program will use all of its unrestricted funds within three to six years. After these funds are depleted, the department may continue to make unrestricted loans using recycled principal and interest payments that exceed the refinanced bonds’ debt-service costs. However, according to its consultant’s estimates, the annual number of loans the Cal-Vet program will be able to make beyond 2006 using only these recycled funds will not exceed 700—less than half the annual number the program now makes using these funds. Moreover, these recycled funds will continue to ebb away each year because of the department’s debt-service commitments until they are completely gone in 2028. C A L I F O R N I A S T A T E A U D I T O R 21 THE DEPARTMENT’S CURRENT LOAN STRATEGY WILL SERVE FEWER VETERANS THAN WOULD ALTERNATIVE STRATEGIES The department has some options in the way it uses QMB and unrestricted funds, but it has thus far not determined how best to serve eligible veterans with the limited available funding. The department’s current strategy is to focus on boosting its overall loan portfolio by expanding its market share of veteran home buyers and increasing the aggregate value of its loans. Although the department has succeeded in making many new loans, these loans are being made at rates that limit its ability to conserve scarce funds to benefit the maximum number of veterans for as long as possible. However, the department has the option of providing veterans with QMB-backed loans and unrestricted loans at rates that are above what it now offers but still lower than current market rates. Increased revenues from higher interest rates would pro- vide more veterans with home loans over a longer period of time while still affording them a benefit. To be effective, this option would require that the department change its interest rates more quickly than it has in the past to maintain its desired competitiveness with changing market rates for home mort- gages. The department could also use higher-cost taxable bonds as an alternative to its traditional tax-exempt bond financing to extend funding for QMB-backed and unrestricted loans. Unlike tax-exempt bonds, taxable bonds do not carry restrictions from federal income tax laws on the use of proceeds. Although both options will allow the Cal-Vet program to continue a bit longer, neither will provide a permanent solution for the program’s continued viability. The Cal-Vet Program Focuses on Expanding Its Loan Portfolio Without Regard to Long-Term Consequences The department’s strategy has been to grant loans from all three of the Cal-Vet program’s funding sources at interest rates low enough to attract a sufficient number of applications to The department’s increase the size and value of its loan portfolio. The program’s strategy uses below- most recent business plan, its plan for fiscal year 1998-99, market interest rates established the program’s goals as granting $336 million in to increase the size and loans and increasing market share to 10 percent of the loans value of its portfolio of made to veteran home buyers throughout the State. This goal loans. was set high enough so new loans would exceed the loans paid off through regular monthly mortgage payments or through 22 C A L I F O R N I A S T A T E A U D I T O R prepayments. The only goal outlined in the department’s latest strategic plan concerning the Cal-Vet program is to make the program more competitive. The strategic plan does not address the challenges the program faces regarding the decline in veter- ans’ eligibility for QVMB-backed loans or the limited QMB proceeds and unrestricted funds. Thus far, the department seems to have been successful in its loan strategy. For example, from July 1999 through The department exceeded February 2000, the Cal-Vet program has made loans totaling its annual goal in the $361 million in just the first eight months of the fiscal year, first eight months of the surpassing its annual goal of lending out $336 million. During fiscal year by granting the time these loans were made, the Cal-Vet program was $361 million in loans. charging interest rates of 5.95 percent on its QMB-backed loans and 6.65 percent on its unrestricted loans—1.4 percent and 2 percent, respectively—below prevailing market rates. However, offering such low rates will provide less of a return on these loans and will limit the department’s ability to offer the same benefit to veterans in the future. Alternative Strategies Could Extend QMB and Unrestricted Funds Other options are available to the department that would extend the Cal-Vet program and serve more veterans. One option would be for the department to set interest rates for QMB-backed or unrestricted loans above current rates but still below market rates. Higher interest rates would slow the number of new loans and provide more revenue that would allow the program to serve more veterans and provide new loans for a longer period of time while still giving them a lower interest rate than they would find elsewhere. It could also facilitate this option by responding more quickly to changes in mortgage interest, which would help control the loan volume. The department also has the option of adding to its scarce supply of funds for QMB- backed and unrestricted loans by blending taxable bond pro- ceeds with its traditional tax-exempt bond proceeds. To explore the first alternative, we studied prevailing interest rates over a 15-year period. We noted that the smaller the difference between Cal-Vet and market rates, the lower the volume of new loans. For example, between fiscal years 1989-90 and 1991-92, the Cal-Vet program’s rates shrank from 2.1 percent below market to 1 percent below market. At the same time, new loans dropped from 2,600 to 1,500 per year. If the program increased its rates but kept them below market, it C A L I F O R N I A S T A T E A U D I T O R 23 could replenish funding sources while still providing a benefit to veterans and, over a longer period, eventually grant more loans. Additionally, as the volume of new loans slows, the program will not deplete funding for the loans as rapidly. It could then use money from repayments and prepayments of outstanding loans in excess of its debt-service costs to further replenish these funds. According to a cash flow analysis a consultant performed for the department, if the volume of unrestricted loans slowed from the current rate of $300 million to $180 million a year, the Other strategies would Cal-Vet program could extend the life of unrestricted funds by temporarily extend the eight years and double the number of loans it grants. For this Cal-Vet program and option to be effective, the department must adjust the Cal-Vet serve more veterans. program’s interest rates with changes in the market rates. Under its current policy, the department cannot respond rapidly to rate changes because it must first seek approval from both the California Veterans Board (board) and the California Veterans Finance Committee (finance committee), a somewhat lengthy process. The last time the department sought approval was in November 1998, so the current interest rates do not reflect the rise in market rates over the last year. State law allows the de- partment to adjust interest rates without seeking prior approval from either the board or the finance committee if it does so using a method approved by both entities. Such a method would allow the department to change its interest rates up or down within a predetermined range. It could then promptly adjust its interest rates according to fluctuations in market rates and maximize its returns on loans. The department also has the option of adding to its scarce supply of funds for QMB-backed and unrestricted loans by blending taxable bond proceeds with its traditional tax-exempt bond proceeds. According to the department’s bond counsel, current tax law allows the department to add to its existing funding for these loans with proceeds from the sales of taxable bonds. Such a strategy would enable the Cal-Vet program to make more loans to veterans, even though the department would have to charge higher interest rates for loans funded with these blended proceeds. Blending proceeds of taxable and tax-exempt bonds is a method successfully used by other hous- ing authorities, such as the CHFA. In addition, Texas issues taxable bonds for unrestricted loans and when available, uses program resources to subsidize veterans interest rates. 24 C A L I F O R N I A S T A T E A U D I T O R Although either option would extend QMB-backed and unre- stricted loans beyond the six years now projected for their viability, neither is a permanent solution for the long-term survival of the Cal-Vet program. Under the first option, as the program raises interest rates nearer to market rates, more and more veterans will obtain loans from other lenders. Similarly, if the second option were employed, when enough taxable bonds were sold that the rates from blended proceeds The Legislature would would approach market rates, the program would no longer need to create a new loan offer a unique benefit to veterans that would not be available program funded from a from commercial lenders and other housing programs. new source to provide below-market interest For the department to continue to provide below-market interest rate loans to veterans rates to veterans, the Legislature would have to create a new over the long term. program and a new mechanism for funding loans. As described in the Military and Veterans Code, the State’s policy is to provide veterans with the opportunity to finance purchases of farms and homes. The State Constitution also authorizes using state funding to help war veterans buy farms and homes. In the absence of sufficient tax-exempt bond financing, the traditional source of program funding, the Legislature may wish to create a new program funded through appropriation of the State’s funds. RECOMMENDATIONS The department should determine how to best serve California veterans in acquiring farms and homes using its remaining limited funding. If the department decides to continue its present strategy of using available funds to provide loans at the lowest possible interest rates, it should plan its operations for the future curtailment of new loan activity. If the department determines that veterans are best served with interest rates that are closer to market interest rates and expands its pool of funds with alternate financing methods, it should consider the follow- ing actions: • Maintain current demographic data necessary to identify the population of veterans eligible for, and likely to participate in, the Cal-Vet program. • Seek approval for an interest rate methodology to allow the Cal-Vet program to quickly adjust interest rates in accor- dance with changes to market rates. C A L I F O R N I A S T A T E A U D I T O R 25 • Explore methods of additional funding for loans, such as blending taxable bond proceeds with traditional tax-exempt bond proceeds. • Based on the above steps, adapt the Cal-Vet program to provide a home loan benefit to the greatest number of qualifying veterans for as long as possible. In the absence of sufficient financing to ensure the continued viability of the Cal-Vet program, the Legislature should consider using state funds to aid California veterans in purchasing farms and homes. (cid:1) 26 C A L I F O R N I A S T A T E A U D I T O R CHAPTER 2 Poor Cost Controls and Inefficient Administration Erode Funds and Delay Loan Applications SUMMARY T he Department of Veterans Affairs (department) has not adequately controlled its spending of California Veterans Farm and Home Purchase Program (Cal-Vet program) funds for program operations or ensured its loan processing is standardized and cost-efficient. Most significantly, it has directly charged the Cal-Vet program for the costs of administrative staff who do not provide service to the program or for whom it cannot document service. Using bond proceeds or related rev- enues for other than their authorized purpose violates state bond laws. In addition, the department has not updated its method for properly allocating indirect administrative costs to the different programs it operates. In addition, because the department has not made improve- ments to the Cal-Vet program that it has identified as necessary for consistent and efficient operations, average loan processing times not only vary significantly by field office, they exceed loan processing times reported by commercial lenders. Lengthy processing times also increase the personnel costs associated with processing a loan and are an obstacle for some loan appli- cants who have to arrange for interim financing while waiting for their Cal-Vet loans. The department has not centralized its loan processing operations or implemented workload standards for its field and headquarters offices to ensure uniformity of operations and standardized processing times. Because the department has not implemented improvements to reduce loan processing times and costs, and because it has not properly allocated administrative costs to its programs, fewer funds are available for loans to veterans. C A L I F O R N I A S T A T E A U D I T O R 27 EXCESSIVE COSTS ERODE FUNDS AVAILABLE FOR VETERANS Over the past three years, the program’s overall costs per loan have increased because its operating costs are being spread over fewer loans. While the number of loans in its portfolio declined by 6,009 from fiscal years 1996-97 to 1998-99, the overall cost of administering the program remained constant. When measured on a per-loan basis, and adjusted for inflation, the average annual administrative cost per loan increased from $508 in fiscal year 1996-97 to $599 in fiscal year 1998-99. Figure 4 illustrates this increase in costs, which results in fewer funds being avail- able to veterans. FIGURE 4 Annual Administrative Costs Per Loan Are Increasing 1996-97 1997-98 1998-99 $0 100 200 300 400 500 600 Cost In Dollars 28 C A L I F O R N I A S T A T E A U D I T O R sraeY lacsiF $508 $570 $599 Note: Numbers are adjusted for inflation using fiscal year 1996-97 as the base year. Calendar year 1997 inflation rates were used to adjust fiscal year 1997-98 costs. Likewise, calendar year 1998 inflation rates were used to adjust fiscal year 1998-99 costs. Our finding mirrors the conclusion in a 1998 report by the Legislative Analyst’s Office (LAO), which found that the admin- istrative costs of the program rose between fiscal years 1989-90 and 1996-97. Although the LAO found that the administrative cost per loan increased from $241 to $519, our estimate of administrative costs is slightly lower because we excluded costs associated with bond issuance. These costs are one-time expenses specific to a particular bond issue and are not within the program’s control. In addition, data from the Cal-Vet program indicates that staff time required to process a loan nearly doubled between fiscal years 1995-96 and 1999-2000. A report the Cal-Vet program prepared in 1996 indicates that it took 21.2 hours for staff to process a loan in fiscal year 1995-96. According to information in a budget change proposal (BCP) the Cal-Vet program prepared in September 1999, staff currently take 40.4 hours to process a loan. This figure is based on the average time it took a sample of loan processing staff to complete their work and represents an 91 percent increase in loan processing time. Using personnel expenses from July 1999 through January 2000 If field offices had been as for the field offices, we calculated that it cost the Cal-Vet efficient in processing program $827 on average to process a loan during this period. loans as they were in Had staff processed loans as quickly as they did in fiscal year fiscal year 1995-96, the 1995-96, each loan would cost only $434 to process. Given that Cal-Vet program would the Cal-Vet program processed 2,201 loans during this period, it have saved more than spent over $850,000 more to process loans than it would have if $850,000 in the first staff worked as quickly as they did in fiscal year 1995-96. seven months of fiscal year 1999-2000. Private-sector lenders estimate they require 10 to 26 hours to process one loan. Compared to these figures, the Cal-Vet program’s processing average appears excessive. According to the chief of headquarters operations and the program’s statewide operations manager, loan processing times have increased due to new underwriting procedures and a new integrated information system; however, as discussed later in this chapter, the increase may also be attributed to the need to implement the recommen- dations identified in the Cal-Vet program’s reengineering plan. Additionally, loan processing costs vary significantly among the program’s 10 field offices, indicating that some offices’ opera- tions may be more cost-efficient than others. Each field office may either be a district office or a satellite office. Because district offices pay the cost of an office manager and satellite offices do not, we expected district offices to have higher per-loan process- ing costs. However, we found that this was not always the case. Nevertheless, we compared average personnel costs separately for district offices and for satellite offices. Figure 5 compares costs among the 10 offices. From July 1999 through January 2000, costs per loan ranged from $510 to $1,803 for satellite offices and from $617 to $2,437 for district offices. As the figure indicates, the San Diego office had the lowest average cost of all district offices while the Redding office had the lowest cost for all satellite offices. Had all offices processed loans at costs comparable to these two offices, the department could have saved over $550,000 in just seven months. C A L I F O R N I A S T A T E A U D I T O R 29 FIGURE 5 Loan Processing Costs Vary Among Field Offices District Office Satellite Office $2,500 2,000 1,500 1,000 500 0 Fr S e a s c n r o a me n Sa t o n D S i a e n g t o a W Cl e a s r t a C o B v a i n k a ersfie C l d o nc o R r d e d d R i n e g dla n d V s e nt ura 30 C A L I F O R N I A S T A T E A U D I T O R naoL reP tsoC egarevA The Department Has Inappropriately Charged Program Funds for Administrative Costs Excessive administrative expenses are also partially attributable to incorrectly charged administrative costs. Even though the department does not track the activities of administrative staff, it charges the costs for many of these positions to the Cal-Vet program. For the month of March 2000, the department charged $298,000 in salary costs for 88 administrative staff to the Cal-Vet program. However, at least part of the cost of these staff should be paid from other state funds as their services benefit other veterans programs the department administers. For example, the department charges the program for 30 of the 52 positions in its information services division, including the chief of the division, even though not all of these positions work solely on Cal-Vet program activities. Only 11 are assigned to the farm and home support section, although others may work part-time on the program’s projects as well. Of the 15 positions in the department’s legal division, 7, including the chief counsel, are also charged 100 percent to the program, yet the department has no documentation of the time legal staff work on the program. Other administrative positions also charged 100 percent to the When it uses Cal-Vet program include the chief of the legislative and public affairs program funds to pay for office, the chief of employees’ rights and relations, the director direct and indirect of cemeteries and memorials, the chief of the budget office, the services that do not chief of contract management, and the chief of internal audits. benefit the program, the Clearly, it is unrealistic to think that all these administrative department violates state staff work exclusively on Cal-Vet program activities. bond laws. The Cal-Vet program is funded by general obligation and revenue bonds; therefore, the use of these proceeds is subject to state bond laws. According to a 1974 decision by the California Court of Appeal, using bond proceeds or related revenues for activities other than their authorized purpose is in violation of these laws. The department is attempting to correct this situation. It hired a consultant in January 2000 to perform a workload efficiency study and cost-benefit analysis of all centralized administrative functions and determine the correct funding source for each. It expects the consultant to conclude the study and establish an equitable cost-allocation system by June 30, 2000. The deputy secretary of operations stated that the department uses Cal-Vet program funds for other purposes because of its inability to allocate costs for the time of employees who work on multiple programs. He gave the position of deputy secretary of the Cal-Vet program as an example. The deputy secretary attends executive staff meetings with other department execu- tives where issues relating to all the department’s programs are discussed. As a key member of the executive management team, the deputy secretary must be involved in all agency issues, but it is difficult to properly separate out the time he spends on various programs. However, other state agencies, even quite large ones that administer many programs, regularly meet the challenges of equitably allocating administrative costs among their various programs. Documentation provided by the department shows it has a history of using Cal-Vet program funds to pay for administrative staff without sufficient evidence that they provide service to the program. Because the department does not require that administrative employees track their time, it cannot determine with certainty how much it has inappropriately charged to the program or know how many years this practice has continued. Information the department gathered during its 1996 reengineering efforts suggests that the funds paid for services unrelated to the program may be significant. At that time, C A L I F O R N I A S T A T E A U D I T O R 31 the project team that analyzed staffing recommended the department study the activities of 71 positions located outside the Cal-Vet program, but whose costs were charged to the program. In a 1997 BCP, the department stated that it charged the Cal-Vet program $1.3 million in fiscal year 1996-97 for adminis- trative staff that did not work on the program. The department requested that the Department of Finance include in the The department governor’s proposed budget for fiscal year 1998-99 a shift of acknowledged that it costs totaling $1.3 million from the Cal-Vet program to the improperly charged the State’s General Fund to ensure the department’s administrative Cal-Vet program for costs were properly charged. The department stated it had $1.3 million in developed its request after studying staff work flow during fiscal administrative costs in year 1996-97 and work-flow projections for fiscal year 1998-99. fiscal year 1996-97 alone. The department believed it had only two alternatives to granting the request for this shift. One was to redirect funding for veter- ans homes to the Cal-Vet program and thus reduce services to veterans; however, the department stated that this alternative could result in health and safety issues in the veterans homes it operates. The second was to continue to allow the Cal-Vet program to pay for unrelated services. The Department of Finance did not act on the request because the department failed to include either the necessary workload analysis to sufficiently demonstrate that the administrative staff were critical to its function or adequate rationale for why the positions should be supported by the General Fund rather than by Cal-Vet program funds. For its fiscal year 1999-2000 budget, the department proposed another BCP that addressed only positions in its information services division. It acknowledged that, due to program and budget changes, the Cal-Vet program was being overcharged for administrative support while the State’s General Fund was not paying its fair share. The department stated that, based on its analysis, 12.6 positions in the information services division were incorrectly charged to the Cal-Vet program. The Department of Finance did not include this BCP in the governor’s proposed budget but did not state its reason. The deputy secretary and the chief of financial services both stated that the department has no confidence in the studies used to prepare the BCPs because the Department of Finance has not yet agreed to a methodology for cost allocation and because the workload for cost allocation varies from year to year. The deputy secretary further stated that the prior studies were point-in-time studies that may not be relevant today. 32 C A L I F O R N I A S T A T E A U D I T O R In addition to incorrectly charging some administrative posi- tions directly to the Cal-Vet program, the department cannot be certain that it equitably allocates indirect administrative costs among its programs. Indirect costs include departmental administrative positions such as employees assigned to the California Veterans Board and the secretary’s office, and shared operating expenses such as building rent and utilities. For fiscal year 1998-99, the department allocated $2 million in indirect administrative costs to the programs it administers with almost $1.2 million charged to the Cal-Vet program. The department allocates these indirect administrative costs using percentages intended to represent the time worked by staff on each program. Its accounting administrator believes the cost-allocation percent- ages it uses were last adjusted sometime in 1998. However, changing cost-allocation percentages without an underlying rationale can result in programs absorbing more indirect costs than they should. A sound cost-allocation practice requires that the department conduct time studies of staff activities frequently enough to ensure that the percentages used match the level of benefit each program receives. This data should be used to develop a plan for allocating costs. According to the department’s fiscal year 1999-2000 BCP, its allocation plan was developed in 1991. However, it was unable to provide us with a copy of the plan or any evidence that the plan had been revised since 1991 to reflect changes in work activities. The Department Has Not Completed Implementation of Its Reengineering Plan The Cal-Vet program’s operating costs are higher, in part, because the department has not completed all of the Although the department reengineering steps it identified as necessary to increase the embarked on a plan to effectiveness and efficiency of its operations. In April 1996, the reengineer its Cal-Vet department began a comprehensive program improvement, program in 1996, several or reengineering project, based on the principles of Total improvements to increase Quality Management (TQM). It has implemented many of the efficiency remain recommendations developed through this project and made incomplete. many significant improvements to the Cal-Vet program’s business processes; however, several crucial improvements have yet to be completed. For example, the department has not completed centralization of its loan processing; updated its underwriting manual, staffing model, and annual training plan; or developed staff workload standards. C A L I F O R N I A S T A T E A U D I T O R 33 During the first phase of the Cal-Vet program improvement project, a team of consultants conducted a baseline review and assessment and produced a written report of findings and recommendations. In the second phase, staff were assigned to TQM teams facilitated by a contracted consultant. The teams developed recommendations for improving business processes, which were reviewed by department management and revised as necessary. The final phase of the project involved implementa- tion of the business process reengineering plan for the Cal-Vet program. In addition, the reengineering plan recom- The reengineering process mended the implementation of an integrated covered the following areas: information system (system). The implementa- • tion of this system is the subject of Chapter 3. Loan origination • Underwriting • Loan contract services Over the past four years, the department has • Delinquencies and foreclosures • implemented many of the recommendations Insurance programs • Financial management from its consultant and TQM teams and • Organizational structure improved efficiency in many areas. These • Data processing • improvements include centralizing loan contract Marketing servicing, adopting new underwriting standards, instituting mortgage insurance by contracting with a private company, becoming an approved lender for the U.S. Department of Veterans Affairs Loan Guaranty Program, improving management of delinquent and foreclosed contracts, and reducing losses from insurance programs. Table 4 summa- rizes the department’s reengineering achievements for the Cal-Vet program. In its reengineering plan, the department identified centraliza- tion of loan processing as a means of increasing efficiency. Although it created a small central processing unit during fiscal year 1998-99, it has yet to fully centralize its loan processing. The wide range of costs among field offices identified earlier in this chapter indicates that some offices may be more efficient than others. The cost variance among the offices also indicates that a large central processing office might be more efficient than many small offices that do not have standardized operations. The department has already centralized its loan contract servicing operations, which allowed it to cut back on staff. The chief of the Cal-Vet program stated that staff are working to obtain additional space to accommodate an expanded central loan processing unit. 34 C A L I F O R N I A S T A T E A U D I T O R TABLE 4 Cal-Vet Program Reengineering Achievements Area Achievements Loan Origination (cid:127) Developed mortgage broker program. (cid:127) Improved and streamlined loan application forms. (cid:127) Supported legislation to expand eligibility. Underwriting (cid:127) Improved documentation procedures. (cid:127) Implemented quality-control reviews. (cid:127) Improved credit-review policy. (cid:127) Adopted U.S. Department of Veterans Affairs underwriting standards. Loan Contract Services (cid:127) Converted from capitalization to impound accounting. (cid:127) Centralized loan contract service functions. Delinquencies & REOs (real (cid:127) Centralized collections and foreclosures. estate owned or foreclosed (cid:127) Reduced processing times and losses from foreclosed contracts.* properties) Insurance (cid:127) Increased deductibles for disaster, fire, and hazard insurance. (cid:127) Transferred life and disability insurance management to an outside contractor. (cid:127) Reduced losses from insurance programs. Statutes & Regulations (cid:127) Sponsored legislation to enhance program benefits. (cid:127) Sponsored legislation to increase flexibility when setting interest rates. (cid:127) Worked for passage of new bond issue. *We were only able to evaluate the Cal-Vet program’s performance in this area from fiscal years 1996-97 through 1998-99 because the program does not currently produce statistics allowing it to trend delinquent and foreclosed contracts. We also found that the Cal-Vet program does not have an updated underwriting manual or a training plan for program staff, both of which were included in the department’s reengineering plan and are necessary to ensure that staff are properly trained and that loans are processed efficiently and in a uniform manner. When we asked loan processing staff for their underwriting procedures, they gave us copies of a number of different manuals, several of which were outdated. One loan processor told us that he had assembled a manual on his own to use as a reference by combining pages from a variety of sources. Furthermore, the chief of headquarters operations reported that the Cal-Vet program has not updated its training plan for the current fiscal year. C A L I F O R N I A S T A T E A U D I T O R 35 The Cal-Vet program also lacks workload standards, even though the department’s reengineering plan states that developing industry operating standards and implementing a performance review system based on the standards is one of its goals. According to the chief of headquarters operations and the The absence of workload statewide operations manager, the Cal-Vet program has not standards and failure to developed formal function-specific workload standards. The use an optimal staffing chief of headquarters operations stated that managers currently model reduces monitor workload by determining whether work is completed accountability and by required deadlines or whether a backlog of work is building contributes to higher up. This management method differs from the practices of costs and inefficiencies several private-sector lenders we spoke with who stated that among field offices. their processing staff are held accountable for completing a set number of loans each day. Similarly, developing workload standards for Cal-Vet program staff would ensure that staff performing the same job are held accountable for completing the same amount of work. According to the chief of headquar- ters operations and the statewide operations manager, the Cal-Vet program is planning to hire a retired annuitant to assist in developing workload standards. We also found that the Cal-Vet program’s staffing model is outdated. The model the program currently uses was developed in response to a recommendation in its reengineering plan for district and satellite offices to maximize staff resources at a time when loan volume was low; however, loan volume has since increased significantly, and the Cal-Vet program has not up- dated its staffing model to reflect these changes. According to the statewide operations manager, the current model is based on a ratio of two property agents for every program technician. Property agents are responsible for underwriting loans, whereas program technicians process applications. According to the statewide operations manager, this ratio was appropriate when the loan volume was low because property agents are in a higher job classification and therefore have more flexibility than pro- gram technicians in the types of work they can perform during slow times. She also stated that when loan volume is high, a ratio of one property agent for every program technician is more efficient and that the department plans to hire additional program technicians to reach this ratio. This lack of an appropri- ate staffing model for current conditions has also contributed to high costs and to variations in costs among field offices because the use of staff is not optimal. 36 C A L I F O R N I A S T A T E A U D I T O R LENGTHY PROCESSING TIMES INCREASE COSTS AND ENGENDER DISSATISFACTION WITH THE PROGRAM The Cal-Vet program’s lengthy processing times contribute to increased costs and frustrates borrowers. Due to the nature of the real estate market, veterans may find themselves forced to pass up desired properties if sellers are unwilling to wait for the Cal-Vet loan to come through or are forced to find more costly interim financing while waiting for the program to process their loans. Our analysis of loan applications processed between July 1999 and February 2000 revealed that the department processed only 8.1 percent of all Cal-Vet loans in less than 60 days. According to the chief of the Cal-Vet program, it has a goal of processing all loans within 30 to 45 days. Figure 6 shows a breakdown of the program’s overall processing times. For the purpose of our analysis, we removed new construction loans from our calcula- tions because, according to the manager of the Cal-Vet program’s management analysis unit, processing times for these loans are often longer due to delays in construction. FIGURE 6 Cal-Vet Program Loan Processing Times 2% < 30 days 31 to 60 days 6% 61 to 90 days 91 to 120 days Over 120 days 32% 19% 41% C A L I F O R N I A S T A T E A U D I T O R 37 According to the department, our calculations of the Cal-Vet program’s processing times may be overestimated. The statewide operations manager reported that, until recently, the coding of new construction loans had been inconsistent, so many new construction loans were not coded as such. The department also performed its own analysis and found that the average process- ing times for loans funded between July 1999 and March 2000 was 71 days. To arrive at this statistic, it eliminated new con- struction loans as well as loans that had been in the system over 180 days under the assumption that they were either incorrectly identified and were really new construction loans or because reasons outside the department’s control caused the delays. However, the department did not include the basis for these assumptions in its analysis. Without reviewing these long unprocessed loans to determine their nature or why their processing had not been completed, we do not think it is reasonable to automatically eliminate them to arrive at an average processing time. According to the statewide operations manager, our analysis of loan processing times may also be overstated because the department did not include a data field in its new system to track the date when all paperwork for a loan application had been received. Therefore, some of the processing time we included is time staff spend obtaining additional documenta- tion. We believe including this time is acceptable, however, because lenders we contacted in the private sector include the time required to ensure all necessary documentation is in place Long loan processing in their considerably lower average processing time of 30 to times have cost the 45 days. Cal-Vet program business and reduced According to a BCP the department prepared in September 1999, the benefit received by extended processing times have caused the loss of potential loan some veterans. contracts as veterans’ dissatisfaction with the program has risen. Furthermore, a loan processor for the program reported that some veterans must take out bridge loans, a form of short-term interim financing, when their Cal-Vet loans take so long to process. Because such financing is more costly than what Cal-Vet offers, the overall benefit of the program is reduced for veterans who must use this mechanism to temporarily finance their purchases. Several private-sector lenders also noted that the program’s processing times were excessive. One mortgage broker even stated that he would not refer clients to the program for that reason. 38 C A L I F O R N I A S T A T E A U D I T O R FIGURE 7 Backlog of Loans In Process Is Growing 700 600 500 400 300 200 100 0 C A L I F O R N I A S T A T E A U D I T O R 39 dessecorpnU fo rebmuN snoitacilppA naoL A New Mortgage Broker Program and Exceptionally Low Interest Rates Have Increased Loan Volume The Cal-Vet program’s growing backlog of loan applications in process has contributed to long processing times. The backlog is due in part to the department’s lack of preparation for increased demand for Cal-Vet loans prompted by a new mortgage broker program that brings more applicants into the program. Based on the department’s weekly sales reports, we determined that the Cal-Vet program’s backlog of unprocessed loans grew by 699 between the beginning of October 1999 and the end of March 2000. Figure 7 illustrates the Cal-Vet program’s growing backlog. 9 9 9 0 0 0 9 9 9 0 0 0 Oct N ov Dec Ja n Fe b Mar We were unable to determine the status of the Cal-Vet program’s backlog prior to October 1999. Weekly reports were not available for the first three months of fiscal year 1999-2000 due to the installation of a new system. When calculating the program’s backlog, we assumed that 14 percent of the applications would never be funded. We based this drop-out rate on the figure the department used in its fiscal year 2000-01 BCP requesting addi- tional loan processing staff. According to the Cal-Vet program’s statewide operations manager, based on the current number of loans in the system, the total backlog may be greater than 699. A spreadsheet the department completed in May 2000 that included the loans backlogged prior to October 1999 shows 1,699 loans in process. Using this recent data and assuming a drop-out rate of 14 percent, the Cal-Vet program has an esti- mated workload of 1,461 loans. Given that it processed an average of 262 loans per month between October 1999 and March 2000, it may currently be backlogged with over five-and-a-half months’ worth of work. Factoring in the same drop-out rate, we estimate that the program’s workload increased to 379 new applications per month from October 1999 through March 2000. If its current average processing rate remains at 262 loans per month, the program’s backlog will continue growing at a rate of 117 applications per month. The department plans to hire 43 new Cal-Vet program staff during fiscal year 2000-01 to handle its increased workload; however, as mentioned earlier, the Cal-Vet program has not updated its procedures manual or its training plan, and it has not developed workload standards or a staffing model appro- priate to its current loan volume. Given the program’s present level of productivity, additional staff will be needed to reduce its backlog, but had the department previously taken steps to increase its efficiency, it could have already reduced the time required to process loans and, therefore, prevented some of the current backlog. The Cal-Vet program’s backlog can also be attributed to two recent changes for which the department did not adequately prepare: a new mortgage broker program and a sharp rise in market interest rates. In January 1999, the department imple- mented a program that allows mortgage brokers to originate Cal-Vet loans. Even though it should have anticipated an increase in loan volume, the department did not adequately prepare for it. As Figure 8 shows, between July 1999 and February 2000, the Cal-Vet loans brokers processed grew from 8.4 percent in July 1999 to 45.7 percent in February 2000. According to the manager for the Sacramento district office, the The quality of loan quality of loan packages the program receives from mortgage packages prepared by brokers varies. Some of these applications take longer to process mortgage brokers varies, because some brokers do a less thorough job of gathering all of sometimes contributing the information necessary to process an application. On the to longer loan processing basis of background research on the California Housing Finance times. Agency’s and the California Public Employees Retirement System’s lending programs, a report staff prepared as part of the Cal-Vet program’s reengineering efforts in 1996 recommended 40 C A L I F O R N I A S T A T E A U D I T O R FIGURE 8 The Percentage of Loans Originated by Mortgage Brokers Is Increasing 50 40 30 20 10 0 C A L I F O R N I A S T A T E A U D I T O R 41 srekorB yb detanigirO snaoL J ul 9 9 A u g 9 9 Se pt 9 9 Oct 9 9 N ov 9 9 Dec 9 9 Ja n 0 0 Fe b 0 0 that the department allow authorized lenders to originate Cal-Vet loans and that it contract with a manager to set up and administer the network of authorized lenders. In spite of this recommendation, the department has not contracted with a manager, nor does it monitor or train brokers, steps that could help to ensure the quality of loan application packages and cut back on processing times. According to the chief of headquarters operations, when the program first started, the department provided training to brokers; however, field staff are currently too busy to provide training. She also reported that the Cal-Vet program is considering contracting with a company that can train and certify brokers. The second reason for the Cal-Vet program’s backlog is that its interest rates are currently well below market rates; thus, the demand for the program has increased. For example, the department’s rates are currently 6.65 percent and 5.95 percent, respectively, whereas market rates averaged 8.2 percent as of March 2000. As discussed in Chapter 1, unlike other government loan programs and conventional lenders who respond more rapidly to fluctuating market rates, the department has not increased interest rates in response to increases in the market and narrowed this gap. RECOMMENDATIONS To reduce further erosion of Cal-Vet program funds, the depart- ment should complete the steps necessary to ensure its direct and indirect administrative costs are properly charged to those programs served by administrative staff. In addition, the department should identify the amount of Cal-Vet program funds it has used for activities outside the program and seek reimbursement from other appropriate state funds. To further increase the efficiency and consistency in its opera- tions, thereby reducing costs and improving loan processing times, the department should do the following: • Complete centralization of loan processing. • Track industry standards and its own operations and develop and implement workload standards it can use in staffing and budgeting decisions for its field offices and headquarters offices. • Develop a field office staffing model appropriate for the Cal-Vet program’s current loan volume. • Update its underwriting manual and training plan and provide uniform training to all Cal-Vet program staff. • Train mortgage brokers who receive and process Cal-Vet program loan packages. • Monitor application packages submitted by brokers, either internally or through the use of a contracted manager, and limit participating brokers to those who consistently provide loan packages that meet the program’s requirements. Before hiring additional staff to process loan applications, the department should carry out the above steps to identify workload standards and develop efficient and consistent loan processing operations. (cid:1) 42 C A L I F O R N I A S T A T E A U D I T O R CHAPTER 3 Inadequate Management of the New Information System Increases Costs and Creates Doubt About the Reliability of Program Data SUMMARY T he Department of Veterans Affairs (department) will require more money and more staff time to ensure that the new integrated information system (system) for its California Veterans Farm and Home Purchase Program (Cal-Vet program) provides necessary, reliable program and financial information. The department selected an off-the-shelf system that required some modification to meet its needs. However, 15 months after the original target date for taking the system into production, it is still not functioning properly. Because the department did not discretely track one-time expenditures to implement the system, it cannot exactly identify the amount of program funds or department staff time it has spent to acquire and install the system. As of March 2000, the department had spent approximately $2.7 million as well as dedicated a consid- erable but undocumented amount of staff time to install the new system. When key staff left during implementation, the department decentralized project management, and staff essentially abandoned the original implementation plan. Because it did not follow its implementation plan, replace a number of key staff who left in the middle of the project, or fully commit to getting the system up and running, the department must spend additional program resources, which could otherwise be used to support loans to veterans, to ensure the system satisfies the program’s needs. The department also neglected to properly test the system’s performance or maintain adequate documentation of the conversion of program files to the new system. Because of the magnitude of the incomplete testing and the lack of documenta- tion, the department may not be aware of potential problems and cannot be certain that the system properly maintains C A L I F O R N I A S T A T E A U D I T O R 43 borrowers’ records or accurately accumulates program and financial data. Neither has the department documented changes to or provided adequate security for the system. Ensuring that proper system change procedures are followed and restricting employee access to the system is necessary to prevent unauthorized modifications and to safeguard the program’s information and assets. Prolonged efforts to implement the system disrupt the Cal-Vet program’s normal operations and create an additional workload for department staff to correct these problems. As yet, the department remains unable to generate program and financial information from the system. As a result, it lacks reports it needs to adequately track and effectively manage critical program functions such as loan processing, delinquent loans, and repossessed properties. Furthermore, the system is unable to generate financial reports, which hinders department managers in assessing the Cal-Vet program’s operations. THE NEW SYSTEM STILL FAILS TO MEET PROGRAM NEEDS After more than Because management did not commit enough resources to 15 months of effort ensure the necessary preparatory work was completed in an and $2.7 million to approved manner, as of March 2000, 15 months after the install a new system for department’s original target date for taking the new system into the Cal-Vet program, it production, the system still does not meet its needs. The system still does not provide was originally scheduled for completion by January 1, 1999, the information or with an estimated initial cost of $2.8 million, which included efficiency that was hardware and software, department staff time, project oversight, initially envisioned. and other operating costs. Because the department did not discretely track its expenditures during the installation period, it is unable to tell how much has been spent to install and start up the system, or whether it exceeded its budget for implementing the system. For example, as shown in Table 5, the department has paid the vendor approximately $1.9 million to date, but since it does not differentiate between payments made for one-time acquisition and implementation costs for the hardware and software and payments made for ongoing maintenance costs, it cannot separately track these costs or compare them to budgeted costs. Similarly, since the department did not track either the time staff actually worked on the project or the operating expenses incurred during installation, it cannot determine whether budgeted costs for these one-time expense categories were exceeded. Finally, although it did track 44 C A L I F O R N I A S T A T E A U D I T O R TABLE 5 The Department Did Not Track Implementation Expenditures to Control Costs Actual Expenditures One-Time Costs Budgeted Expenditures as of March 2000 Difference Hardware/ software costs $1,609,054 $1,885,310* $276,256* Staff costs 769,824 Unknown† Unknown† Other operating and expense costs 154,475 Unknown† Unknown† Consultant costs to oversee project implementation 227,585 860,956 633,371 Subtotal 2,760,938 2,746,266 N/A Five-year ongoing maintenance costs 565,644 Unknown† Unknown† Total $3,326,582 $2,746,266 N/A *This amount may include ongoing hardware and software maintenance costs because the department does not track continuing and one-time costs separately. †The department did not track these costs for the project. payments to the consultant it hired to oversee the implementa- tion of the project, records of those payments show that the department overspent its budget by more than $630,000. In moving to an integrated management system, the depart- ment expected to save a significant amount of time that staff would otherwise spend manually performing a variety of tasks. Reducing the time it takes staff to process loans would also reduce operating costs. Some of the past problems the new system was expected to solve included inconsistencies and limitations in accounting for the program, billing inefficiencies, and the inability of the program to provide prompt and accurate information. As yet, the system has failed to provide either the information or the efficiency intended. Because of the magni- tude of incomplete or undocumented testing the system and the frequency of the problems that still plague its performance, the department may not yet be aware of all the problems with the system. Consequently, the department will need to spend a C A L I F O R N I A S T A T E A U D I T O R 45 substantial amount of additional resources to ensure the system satisfies the Cal-Vet program’s business requirements, money that could otherwise be used for loans to veterans. MANAGEMENT LACKED COMMITMENT TO THE PROJECT Department management did not adequately monitor the progress of the implementation of the system or dedicate enough resources to ensure its success. Although the department originally planned to have the new system running by January 1999, it had to delay start-up until July 1, 1999, and only met that deadline by omitting planned testing that would ensure the system worked properly. Executive management’s general lack of commitment to the project was demonstrated by its failure to replace key project staff who had resigned, give the vendor sufficient information to properly modify the system, dedicate the system consultant’s time to completing the project, ensure the system was fully tested, and attend over half the scheduled project meetings. The department’s system implementation plan called for a team, consisting of a project manager from its information services division, an outside quality assurance consultant, a team leader from the farm and home purchase division (division), staff who would eventually use the system, and a member of senior man- agement, to serve as the project’s executive sponsor. The purpose of an executive sponsor was to provide the authority that could A lack of sound project allocate sufficient resources to complete the project and who management caused would ensure the project’s benchmarks were met. When person- delays in system nel key to the project’s success, including the project manager, implementation and left the department, neither the executive sponsor nor depart- disrupted staffs’ normal ment managers ensured they were replaced. Instead, department workload. managers divided the project up among the system users who participated on the team. Not only were these user groups responsible for testing and implementing the system, they were expected to complete these tasks in addition to performing their regular duties, causing delays in the system’s implementation as well as in the staff’s normal workload. The progress of the project was further hindered because needed information about the department’s business processes were not promptly supplied to its vendor so that software could be tai- lored to the department’s needs. Minutes of the implementation meetings and project status reports document the department’s 46 C A L I F O R N I A S T A T E A U D I T O R failure to respond to the vendor’s request for business process information and its slow progress in system testing. These minutes also reveal that no members of the department’s execu- tive management were present at 16 of 28 meetings held during the period from April 20, 1998, through November 23, 1999, underscoring management’s failure to attend to the project. The department’s reassignment to other projects of the consultant hired to provide quality assurance further impeded the necessary preparatory testing designed to ensure the system was fully functional. As the progress of the project slipped, management allowed the implementation team to compress some testing and skip other testing altogether to meet the revised deadline. According to the deputy director of operations, the department was unable to devote adequate attention to the new system for the Cal-Vet program because at the same time it was also installing a new system for its veterans homes in Yountville and Barstow, as well as working on Y2K compliance projects. PLANNED SYSTEM TESTING WAS NOT PERFORMED The department had planned to perform both modular tests of specific functional areas and comprehensive integrated tests before implementing the system, but it did not properly com- plete the testing. The tests were intended to verify that the system served the needs of the program and management by processing data correctly and generating prompt and accurate reports and documents. The system includes nine separate modules intended to serve different processes of the Cal-Vet program and work together to provide comprehensive program and financial information. Accordingly, the department planned tests of each module that serves the program operations, bond finance division, and accounting unit, as well as of the system as a whole. However, due to delays in the system’s implementation, the department never completed its planned testing before taking the system on-line on July 1, 1999. What testing it did perform Over half of all planned was not thorough enough to ensure the entire system func- testing was never tioned properly. Because of the magnitude of the incomplete performed. testing and lack of documentation, the department may not be aware of potential problems with the system’s performance and cannot vouch for the accuracy of its program and financial data. Of the 358 total tests planned for system implementation, C A L I F O R N I A S T A T E A U D I T O R 47 only 20 percent were adequately completed and documented, 9 percent indicated problems with the system’s processing but lacked evidence the problems were ever resolved, and 19 percent contained handwritten evidence that the tests were performed but lacked system documentation or signatures from staff who performed the tests or approved the results. The department could not provide any evidence it performed the remaining 52 percent, or 185, of the tests. Of the total tests planned, the department intended to perform 173 on the separate modules for program operations and the bond finance division. However, the department could not provide any evidence it performed approximately 67 percent of these tests. Figure 9 shows the status of tests performed on the separate program operations and bond finance division modules. FIGURE 9 Tests Performed for Separate Modules 13% 13% 67% 7% Tested, signed, and documented Tested, but missing system documentation or signatures Tested, but no evidence of resolution of problems found Not tested Figure 10 shows that the department also failed to perform 42 percent of the 86 planned tests of the accounting process. An additional 15 percent of the accounting tests indicated problems with the system’s processing but contained no evidence the problems were fixed. The remaining tests for separate accounting modules were performed, but the tests lacked system documentation, including the names of the personnel who performed the tests or whether the results were ever approved. The department also did not complete any tests to determine if the system would perform the accounting proce- dures necessary to close out monthly, quarterly, and annual accounting periods, and produce financial reports. 48 C A L I F O R N I A S T A T E A U D I T O R FIGURE 10 Tests of Accounting Process 42% 43% 15% Tested, but missing system documentation or signatures Tested, but no evidence of resolution of problems found Not tested Finally, the department performed limited testing of a sample of the 32,243 loan and escrow accounts it converted from its old system, but retained very little evidence of the extent or results of these tests and relied instead on borrowers to verify that these records were correct. The department told us it tested the con- version of 50 loan files and retained a listing of the loans and their balances. However, it did not retain evidence of the specific information from the files it tested, the results of the tests, or any actions it took to correct any conversion problems. Accord- ing to the current project manager, the department sent letters to borrowers explaining the plan to convert to a new system and counted on them to verify their records. Without adequate testing of the new system, the department has no assurance that it properly maintains borrowers’ file records and accurately accumulates program and financial data. Inadequate Project Management Has Disrupted Normal Operations and Created Errors in Borrowers’ Accounts Problems with implementing the system caused an increase in complaints from borrowers. Because of the time spent resolving system issues, some department staff have fallen behind on their normal duties. In addition, some units are suffering staffing shortages. These conditions, together with the increased errors the system made in borrowers’ accounts, caused numerous complaints from program participants. For example, problems C A L I F O R N I A S T A T E A U D I T O R 49 in system processing have caused some borrowers to receive erroneous delinquency notices and late fees. These errors have increased complaints in the customer service unit and increased workloads because accounts must be manually researched and adjusted. Department managers stated that they were aware of the system problems that cause the complaints but do not know when they will be corrected. The department is also facing a potential increase in workload due to a problem in converting borrowers’ escrow accounts to the system. These accounts accrue the borrowers’ payments for property taxes and insurance premiums. System errors in the conversion of the last batch of loans have caused problems with escrow balances. The department believes that it will be able to resolve these issues by June 2000 when it completes an escrow analysis of these accounts. An escrow analysis is required annu- ally by federal law and will be performed on approximately 32,000 accounts, including the 1,500 with known problems. However, the department has not yet finalized the procedures necessary for performing the analysis, nor has it attempted to test its process to determine the extent of the problems or the time involved in correcting each account. Additionally, because the department did not retain the documentation of the tests it performed on the conversion of the escrow accounts, it cannot be certain that the analysis will uncover any new problems. Problems implementing the new system are also to blame for System errors affected at inaccuracies in over 200 borrowers’ accounts. The department least 1,500 veterans’ did not ensure that it had established payment controls, which escrow accounts and tell the system how to apply loan payments. Transactions incorrectly applied over lacking payment controls have been incorrectly posted to 200 borrowers’ payments borrowers’ accounts for the wrong month, prompting to the wrong month. erroneous late charges, or have been added to an unapplied balance account. Staff must individually research these transac- tions and ensure payments are applied to the correct escrow accounts to pay for property taxes and insurance on time. As of March 15, 2000, the department had almost $1 million in its unapplied balance account from 238 borrowers’ accounts. Of this total, $388,000 had been in the account for more than three months. Until they are applied to the appropriate loan balances, unapplied payments will also result in interest being incorrectly charged to borrowers. 50 C A L I F O R N I A S T A T E A U D I T O R The Department Has Not Been Able to Produce Critical Program and Financial Information The department remains unable to generate the program and financial information it needs to operate the Cal-Vet program effectively and efficiently. The system is programmed by the vendor to produce one standard report and has the capability to format customized reports suited to the department’s operations. Because department staff and management lack sufficient train- ing and knowledge of the system to properly customize reports, they are unable to develop or produce any of the following types of reports: performance reports, such as semiannual status reports; delinquency status reports; reports that track the status of foreclosed properties; insurance premium reports; and reports that track the time it takes to process loans. Without such reports, the department’s ability to monitor and control its Cal-Vet program is severely limited. For example, the collection/foreclosure unit does not have access to activity reports to evaluate its performance for managing delinquent accounts or repossessed properties. Similarly, operations manag- ers do not currently have access to “pipeline” reports that display loan processing activities, including the time to process applications. As a result, information critical to effectively track and manage the loan processing workload throughout the headquarters and field offices is unavailable. Because the system is not yet capable of completely processing Lack of expenditure all expenditures and other financial transactions and the reports has hindered the department has not customized its reporting capability, department’s ability to the department cannot produce reliable expenditure reports manage the Cal-Vet or generate financial statements. From July 1, 1999, to program and control April 2000, expenditure reports were not available to the costs. program’s managers. Without access to periodic expenditure reports, the department is deprived of one of its most important means of controlling costs—the comparison of actual operating expenditures to budgeted expenditures. In April 2000, the department was able to produce its first monthly financial statement for fiscal year 1999-2000, for the month of July 1999, by performing manual reconciliations of general ledger accounts and downloading the account balances to software unrelated to the system. C A L I F O R N I A S T A T E A U D I T O R 51 The Department Has Not Adequately Safeguarded Cal-Vet Program Data and Assets The department failed to safeguard Cal-Vet program data and assets by requiring that all changes to the system be authorized, documented, and approved, nor did it ensure security over system access. Adequate controls over changes to the system are necessary to ensure that management has authorized any alterations and that those alterations produce the desired result on all affected system modules and accounts. To prevent unauthorized changes and to safeguard the program’s informa- tion and assets, restricting employee access to the system is necessary as well. Adequate controls over system changes were covered in the Inadequate control over project’s initial implementation plan; however, staff responsible system changes and for making the changes did not follow those procedures. The unauthorized access may plan required that each requested change to the system be expose crucial program authorized and documented. It also required that any documen- data to risk of loss or tation indicate who was responsible for testing and approving misuse. the change, include all vendor information, and receive final approval from the appropriate knowledgeable staff. The department chose to abandon this rigorous process in favor of a much less formal approach so it could meet its revised July 1, 1999, deadline for bringing the system on-line. This less formal approach lacks many of the controls called for in the initial plan. As just one example, no record is kept evaluating the effect changes to one part of the system will have on other related modules or accounts. Rather, the department relies on the system vendor to notify it if the changes requested affect other modules. Another drawback is that staff retain no record of the changes they have made to the system, a lack of documentation that could later undermine their ability to effectively troubleshoot system errors. The department has also failed to secure the information within the system. For example, staff responsible for providing employees access to the system could not provide us with an information security policy. Currently, lead staff in the department units or managers grant access to the system by phone or e-mail. They do not document these requests or evalu- ate whether the access granted could result in incompatible activities for department employees. For example, employees in the cashiering unit receive checks for loan payments and also enter the checks into the accounting system, a circumvention of standard internal controls designed to avoid incompatible 52 C A L I F O R N I A S T A T E A U D I T O R duties. To minimize the risk of theft, the person responsible for receiving cash should not also enter the amount into the system. Finally, although employee passwords are disabled when they separate from the department, no one reviews or modifies access to the system when employees transfer within the depart- ment. This practice may allow unauthorized access to the system and could unnecessarily expose crucial data to risk of loss or misuse. These security shortcomings are particularly disturbing because they coincide with a period when the department is unable to generate financial statements that could alert manage- ment to any anomalous transactions. THE SYSTEM’S INADEQUACIES MAY BE TRACED TO THE FAILURE TO FOLLOW IMPLEMENTATION PLANS The failure of the department to follow adequate project management procedures outlined in its implementation plan has fragmented the system’s operations and impeded the system’s ability to provide essential program and financial information. Management did not remain committed to the original plan; instead it decentralized control over the system’s implementation and failed to allocate enough staff to ensure that project benchmarks were met. Without centralized control and a commitment from management, the department’s approach lacked the unity necessary for ensuring the success of a system intended to integrate different program functions throughout the department. When key staff on the implementation team left their jobs, the department shifted control of the project from a central project manager to the three main divisions that would be using the system and sidelined the implementation plan. As a result, the implementation project lost its systemwide focus and failed to accomplish all of its objectives. Under the initial implementa- tion plan, the project manager was responsible for oversight, such as reviewing the team’s overall plan, monitoring team Once project and vendor activities to ensure deadlines were met, ensuring management was that deliverables were promptly produced and tested, and decentralized, staff maintaining the final project schedule. In addition, the project abandoned the original manager was responsible for scheduling periodic status meet- implementation plan ings, tracking outstanding issues, and ensuring staff followed and systemwide focus accepted procedures when making modifications to the system. was lost. Without a project manager, no central authority ensured the above duties were performed either properly or at all. C A L I F O R N I A S T A T E A U D I T O R 53 Once the department decentralized project management, staff essentially abandoned the implementation plan. The implementation project leader from the division stated that the team did not use the initial plan to track progress and the plan was not an integral part of monitoring the project. Information in the initial plan confirmed his statement. The initial plan was intended to be used to list project tasks and track their progress. The part of the plan that displays project progress shows that 100 percent of systemwide tests were complete by June 17, 1999. However, we found this was not correct. We determined that none of the systemwide tests for accounting processes were ever completed. In May 1999, when it became apparent that the department would not meet its revised July 1, 1999, deadline, the depart- ment attempted to regain centralized control of the project. However, its new approach still did not take into consideration the objectives of the original project plan nor fully incorporate the vendor’s project plan. Rather, staff involved in the project attempted to identify the status by surveying each user unit and compiling a list of tasks to be completed. Additional tasks were added to the plan as staff encountered problems in the system’s performance. This approach did not ensure that staff completed the tasks from the original implementation plan, including ensuring that system testing was carried out. Furthermore, because the department’s revised plan did not contain all of the necessary tasks, it provided an insufficient basis for measuring and monitoring the overall progress toward satisfactory project completion. RECOMMENDATIONS To ensure that its integrated information system functions reliably, the department should convene a centralized imple- mentation team, including a project manager and a sponsor from the department’s executive management team with the authority to allocate the necessary resources. Additionally, the department should contract with an outside consultant with experience in project management to oversee the team. The team should gather all data from prior implementation efforts, assess what tasks still remain incomplete, and identify the steps needed to properly test the modules and system. In addition, the team should obtain the training necessary to design the program 54 C A L I F O R N I A S T A T E A U D I T O R performance reports and financial reports the department needs to efficiently and effectively operate the Cal-Vet program. The team’s activities should include the following actions: • Adequately identify current business processes. • Take steps necessary to ensure that the system’s capabilities serve the current business processes. • Develop appropriate tests for the system to ensure that it accurately and completely processes transactions and accu- mulates program and financial data. • Identify existing implementation issues to ensure current efforts solve the problems. • Maintain a central file for documentation of the efforts to complete the implementation, including planning, tests performed, and resolution of identified issues. • Develop and maintain a method for tracking system issues from discovery to resolution. • Ensure that enough staff is available and properly trained to complete the implementation. • Adequately monitor the progress of efforts to ensure timely completion of the project. To adequately safeguard program data and assets, the depart- ment should implement a security policy to limit system access to employees who are properly authorized and ensure these employees’ access is not incompatible with their other duties. Further, the department should follow proper procedures for making changes to the system. C A L I F O R N I A S T A T E A U D I T O R 55 We conducted this review under the authority vested in the California State Auditor by Section 8543 et seq. of the California Government Code and according to generally accepted government auditing standards. We limited our review to those areas specified in the audit scope section of this report. Respectfully submitted, MARY P. NOBLE Acting State Auditor Date: May 25, 2000 Staff: Doug Cordiner, Audit Principal Norm Calloway, CPA Celina M. Knippling Laura B. Ronneberg Nathan Checketts Michelle J. Tabarracci, CISA 56 C A L I F O R N I A S T A T E A U D I T O R APPENDIX Cal-Vet Program Benefits T o determine whether the Department of Veterans Affairs California Veterans Farm and Home Purchase Program (Cal-Vet program) provides unique benefits to California veterans, we compared key features of the Cal-Vet program with features of four other government loan or loan-guarantee programs and with conventional lenders’ offerings. We reviewed the following programs: the U.S. Department of Veterans Affairs (VA) Home Loan Guaranty Program; the California Housing Finance Agency (CHFA) loan program; the U.S. Department of Housing and Urban Development, Federal Housing Adminis- tration (HUD/FHA) Mortgage Insurance Program; and the U.S. Department of Agriculture (USDA), Rural Development loan and loan-guarantee programs. In contrast to loan programs that provide borrowers with funds, loan-guarantee programs provide insurance for loans with low down payments in case the borrowers cannot repay the loans. Currently, the greatest benefit of the Cal-Vet program is its low interest rates. A rate of 6.65 percent for Qualified Veterans Mortgage Bonds (QVMBs) and unrestricted funds is well below conventional market rates, which averaged 8.23 percent as of the week ending March 31, 2000, according to a survey by the Federal Home Loan Mortgage Corporation. The CHFA loan program, which offers loans to any borrower meeting its qualifi- cations, is funded through revenue bonds and is similar to the Cal-Vet programs Qualified Mortgage Bonds (QMBs). The Cal-Vet program’s QMB-backed loans currently offer the significantly lower interest rate of 5.95 percent to borrowers, compared with CHFA’s rate of 7.25 percent as of February 2000. In addition to low interest rates, the Cal-Vet program offers several other benefits. Some of the other government programs we examined target home buyers with low or moderate incomes and require that participants’ incomes fall below certain limits. While the QMB-backed loans require that veterans meet certain income limits, wartime veterans who do not qualify for this program can apply for loans backed by QVMBs or unrestricted funds without meeting these income requirements. Also, when C A L I F O R N I A S T A T E A U D I T O R 57 compared to most other programs, the Cal-Vet program funds the highest loan amounts. Finally, the Cal-Vet program offers veterans low premiums on fire, hazard, and disaster insurance. Table 6 compares key features and requirements of the Cal-Vet program to those of other government programs and conven- tional lenders. TABLE 6 Comparison of the Cal-Vet Program With Other Government Programs and Conventional Lenders Cal-Vet Loans Program Requirements QVMB-Backed Loans CHFA and Features and Unrestricted Loans QMB-Backed Loans Loans Type of military service Wartime or required Wartime peacetime None Length of service required 90 days 90 days None Income restrictions No Yes Yes First-time home buyer requirement No Yes Yes 3 percent, unless insured Minimum down payment 2 or 3 percent 2 or 3 percent by VA or USDA Maximum loan amount/ purchase price $250,000 $88,266 to $250,000* $88,267 to $256,000* Type of mortgage Adjustable with a cap of 7.5 percent Fixed rate 7.25 percent as of Interest rates 6.65 percent 5.95 percent February 2000 Loan origination fee 1 percent 1 percent 1 percent 0.55 to 0.80 percent 1.25 to 2 percent, which may be financed for initial annually unless guaranteed Mortgage insurance costs loans; 1.25 to 3 percent for subsequent loans by VA, HUD/FHA, or USDA Offers low-cost home and loan protection plans Yes Yes No Other *Maximums for existing homes not located in target areas. Maximums are somewhat higher for homes located in target areas and for newly constructed homes. 58 C A L I F O R N I A S T A T E A U D I T O R TABLE 6 USDA Rural HUD/FHA Development Loans VA Guarantee Mortgage Insurance and Loan Guarantee Conventional Wartime or peacetime None None None 90 days to 2 years None None None No No Yes No No No No No None, unless mortgage is graduated 3 percent None Generally 3 to 5 percent $121,000 for most regions; $219,849 for high-cost $203,000 regions $80,000 to $180,000 N/A Fixed rate or graduated Fixed rate or adjustable Fixed Fixed rate or adjustable 7.37 percent as of April 2000 for the Loan No set rate, rate is negoti- Program; Guarantee ated between borrower and Program has no set rate, No set rate; 8.23 percent lender; 8.07 percent negotiated between average for fixed rate No set rate average as of March 2000 borrower and lender as of March 2000 No fee for the Loan Program; varies for the 1 percent 1 percent Guarantee Program 1 percent average 1.25 to 2 percent for 2.25 percent (or 2 percent Varies with down payment; initial loans; 1.25 to if the buyer attends a on average 0.78 percent for 3 percent for subsequent class), plus 0.50 percent 1 percent for the a loan with a 5 percent loans annually Guarantee Program down payment No No No No Home must be in a rural area C A L I F O R N I A S T A T E A U D I T O R 59 Blank page inserted for reproduction purposes only. 60 C A L I F O R N I A S T A T E A U D I T O R Agency’s comments provided as text only. Department of Veterans Affairs P.O. Box 942895 Sacramento, CA 94295 May 22, 2000 Mary Noble Acting State Auditor 555 Capitol Mall, Suite 300 Sacramento, California 95814 Dear Mrs. Noble: Attached is a copy of the Department’s response to the Bureau of State Audits audit report of the Department of Veterans Affairs. Sincerely, (Signed by: Gerald Rucker) GERALD RUCKER Undersecretary C A L I F O R N I A S T A T E A U D I T O R 61 TABLE OF CONTENTS Transmittal Letter Summary 1 Introduction 3 Chapter 1 4 The Department is looking at ways to expand the eligibility pool and address funding alternatives in order to continue to provide benefits to the maximum number of veterans. Chapter 2 7 The Department disagrees that poor cost controls and inefficient administration erode funds and delay applications. Chapter 3 12 The Department disagrees that inadequate management of the new information system increases costs and creates doubt about the reliability of program data. Cal-Vet Program Benefits 16 Appendix 17 62 C A L I F O R N I A S T A T E A U D I T O R SUMMARY* The California Department of Veterans Affairs (the department) appreciates the Bureau of State Audit team’s recognition in their report of the department’s improvement in efficiency over the last four years. The department has completed a number of significant changes, and its principal goals have been to lower the interest rate on existing contracts of Purchase, provide an interest rate on new Contracts of Purchase which more closely reflects its borrowing costs. Amendments made in 1998 to the Veterans Code eliminated the uniform interest rate requirement for Contracts of Purchase executed on and after January 1, 1999. The interest rates established can be different for different Contracts of Purchase. However, the Veterans Code still requires that all Contracts of Purchase originated prior to January 1, 1999 bear a uniform interest rate that is subject to periodic adjustment. The current interest rates for newly originated Contracts of Purchase are 5.95% per annum for Con- tracts of Purchase financed with Qualified Mortgage Bond (QMB) proceeds, and 6.65% per annum for all other Contracts of Purchase. These rates are in the process of being changed by the department. The department has completed a series of steps to reengineer the loan program and restructured a large part of the outstanding indebtedness that financed the Program. The restructuring has enabled the department to lower the interest rate on existing Contracts of Purchase. Legislation effective January 1, 1998 expanded veteran eligibility under State law to include Early Vietnam Veterans and Peacetime Veterans. Recent revisions to the Veterans Code allow the department to require a 1% loan origination fee. As a result, the department has begun to allow loan origination through mortgage brokers. This eligibility expansion, together with the lowered interest rates on Contracts of Pur- chase, lowered down-payment requirements, and expanded marketing efforts through the use of Mortgage Brokers, has resulted in a 300% increase in the number of new loan applications over the past fiscal year (2,995 new loans funded this fiscal year to date of $500,352,346). The department has aggressively managed the real estate losses and loss expo- sure by obtaining primary mortgage insurance (with a deductible) for a large portion of the existing Contracts of Purchase with high loan-to-value ratios, and obtained loan guaranties from the United States Department of Veterans Affairs (USDVA) for most new Contracts of Purchase with high loan-to-value ratios. Not all planned reengineering efforts have been accomplished as yet; the depart- ment is still in the process of centralizing loan underwriting, implementing workload stan- dards, and updating its Operations Manuals. These items are expected to be implemented during the 2000-2001 fiscal year. The department currently is in the process of securing approval from the Veterans Board and Finance Committee of a new methodology to allow it to react quickly to the changing market and allow the department to control application rates. *In this summary, the department takes exception with numerous issues contained in our report. These issues are developed in the department’s detailed response following this summary. Therefore, we provide our perspective on the department’s response on those later pages. C A L I F O R N I A S T A T E A U D I T O R 63 The department agrees with many (but not all) of the recommendations shown in the report, since most of these arise from current goals and objectives shared with the audit team, and documentation/information was provided to verify the department’s inten- tions. However, the department disagrees with certain conclusions of the audit team regarding its assertions of mismanagement of its new mortgage information system, pro- gram cost/budget allocations, loan funding controls, and program operation. The depart- ment also disagrees that high program costs erode current funding and that its new mort- gage finance system is not operating “properly”. These findings will be discussed in their appropriate chapters. For instance, the report cites the department’s lack of documented workload standards as a drain on its funds; however, as discussed with the audit team, each employee’s production is monitored continually by the manager or supervisor and there has been a dramatic increase in the efficiency of the staff responsible for the loan process, which will be further discussed in the response to Chapter 2. In addition, the audit results found “that the department follows its lending guidelines and the guidelines or its loan guarantors when granting loans to veterans. The lack of documented workload standards has not resulted in the increase in processing times, nor has it resulted in a drain of the available funds. A recent framework for comparable costs with order of magnitude was furnished by the department’s bond consultant, David Gressel of CFX. The comparable costs indicated that it would cost approximately $21,150,421 if the administration of the home loan pro- gram was contracted out to private industry. Therefore, the department’s costs are not excessive and are in line with other comparable housing finance agencies. Conditions have been created over the past few years to position the Cal-Vet Home Loan Program (the Program) for continued growth of the portfolio, new programs have been developed to benefit veterans, technology has been upgraded for efficiencies and Y2K compliance, and the Program has become much more aggressive in its management of its bonds and financial issues. The Program is currently rated AA by the nationally recognized rating agencies as a result of the program changes. There is a demand for the program that exceeds audit report projections, and CDVA expects that the Program will be a viable program well into the future able to meet its mission of providing below-market rates to deserving veterans to ensure that the State’s “promise” to honor those who served is met. 64 C A L I F O R N I A S T A T E A U D I T O R INTRODUCTION CDVA agrees with the facts presented by the audit team in the introduction regarding the portfolio, organizational structure, Veterans Board, program eligibility requirements, funding limitations, LAO report (although CDVA disagrees with certain findings of said report), and is aware of the scope and methodology the audit team used to prepare the report. C A L I F O R N I A S T A T E A U D I T O R 65 CHAPTER 1: The Department is looking at ways to expand the eligibility pool and address funding alternatives in order to continue to provide benefits to the maximum number of veterans. The department agrees with certain recommendations in this Chapter (and, in fact, began developing methodologies prior to the audit team’s arrival), but disagrees with certain other audit findings regarding the interpretation of demographic data as it relates to the future pool of eligible veterans. Audit Report Conclusions/Responses: Cal-Vet Loans will Benefit few Veterans after 2010 • The department agrees that without modifications in Federal regulations, the pool of veterans able to use Qualified Veterans Mortgage Bonds (QVMB’s) will dwindle. However, the department does not agree with the extent, and does not agree that Cal-Vet loans will benefit “few” veterans after 2010. The report projects that only 100 of the estimated 161,000 California veterans who are still eligible for QVMB’s 1 * will apply for loans in 2010. This projection appears to be based on the assumption that Vietnam Era veterans will not be interested in the program due to their age. Based on USDVA information provided to the audit team, the department projects that Cal-Vet loans will be attractive to a much higher percentage group. An analy- sis of our current loan originations shows that the average veteran purchaser is 45 years old, verifying that older veterans remain in the housing market. Also, approxi- mately 22,000 of the 30,000+ veterans released nationally muster into California each year, a percentage of which will fall into this eligible group. The audit team also bases this 100 on current market goals that may not apply in 2010. Market goals are established based on current market conditions and other factors which are constantly changing. • The department agrees that QMB and Unrestricted funds are limited; however, 2 there is little doubt that additional allocations of QMB are available for Cal-Vet’s veteran first-time homebuyers. Although the bulk of the unrestricted funds may be used by 2006, a certain amount will continue to regenerate. If unrestricted money tax-exempt bonds are blended with taxable bonds, an additional source of funds will be created, although also limited. If legislation is successful to modify federal restrictions on use of tax-exempt QVMB’s, these funds could continue as long as voters continue to authorize them. Considering that the Gulf War era is still “open”, this could continue many years into *California State Auditor’s comments begin on page 101. 1 Source: U.S. Department of Defense Manpower Data Center 2 Source: DD214 database maintained by CVDA, Veterans Services Division (veterans released into CA) 66 C A L I F O R N I A S T A T E A U D I T O R the future. The population of Gulf War era veterans in California is sizable (3185,000 as of July 1, 1999). • The audit report states that the department has not done a detailed analysis to 3 project how quickly veterans in the State are losing QVMB eligibility. The depart- ment in fact performs formal and informal analysis of veteran demographics as part of its marketing efforts, to determine field office boundaries and staffing needs, and for cursory funding projections. Information obtained from the USDVA and Depart- ment of Defense is primarily used for these informal analyses. The Department’s Current Loan Strategy will serve Fewer Veterans than Alternative Strategies would • The department agrees that a new source of program funding is a means of serving a greater number of veterans for a longer period of time while still providing a below-market rate. The department anticipates blending taxable bonds with unre- stricted funds or tax-exempt funds when the time is appropriate. There is not an immediate need for this new program. • The department disagrees with the statement in the report that “the department has some options in the way it uses QMB and unrestricted funds but it has thus far not determined how best to serve eligible veterans with the limited available funding.” 4 The department has shared its plans for managing funds with the audit team, and the audit team’s recommendations/alternative strategies are based on those plans. The department agrees that its short-term strategy has been to boost its overall loan portfolio, which is critical to program stability and growth of the fund, but CDVA is also planning long-term strategies (interest rate adjustments, limiting unrestricted fund allocations, etc.) to decrease new application intake and preserve available funds. • The audit report concludes that as Cal-Vet rates approach market rates (from alternatives such as blending taxable and non-taxable bonds), the program would no longer offer a unique benefit to veterans that would not be available from com- 5 mercial lenders and other housing programs. The department found that marketing its other benefit features (such as its low-cost insurance programs and life-of-loan service) actually increased demand for its product, even when the department’s interest rate was higher than conventional rates. For example, in fiscal year 1994- 95 when the department’s rate was 8% as compared to lower conventional rates, the department funded more loans than the over the previous 3 Source: USDVA report on veterans population in California C A L I F O R N I A S T A T E A U D I T O R 67 year (from 1,180 loans funded in fiscal year 1993-94 to 2,931 in 1994-95). See Exhibits 1 and 2 in the Appendix. Audit Recommendations: The department concurs with the audit team’s recommendations for managing its funds; i.e. maintaining demographic data, seeking approval for an interest rate change methodology, exploring methods of additional funding for loans, and adapting the program to providing loans to the greatest number of qualifying veterans for as long as possible. 4 These recommendations mirror the department’s planned goals, and are currently being implemented. 68 C A L I F O R N I A S T A T E A U D I T O R CHAPTER 2: The Department disagrees that poor cost controls and inefficient administration erode funds and delay applications. Audit conclusions/responses: Poor Cost Controls and Inefficient Administration Erode Funds and Delay Loan Applications. The department disagrees with these conclusions. Program administration expenses 6 declined 4.1% in fiscal year 1999. After taking into account a $1,935,000 increase in operating revenue, net administrative operating expenses decreased 15.0% over fiscal year 1998 results. The Program has been operating under budget for the past two years. • The audit report asserts that excessive costs erode funds available for veterans, and that while the number of loans in the portfolio declined between fiscal year 1996-97 to fiscal year 1998-99, the overall cost of administering the program re- mained constant, with an average cost of loan increasing from $508 to $599 re- spectively. The department believes the cost to be overstated, and the cost of processing a loan did not actually rise. During this period, there were substantial 7 one-time costs associated with MITAS and Y2K, such as increased personnel costs (see Exhibit 3 in the Appendix). These costs were beyond the control of the de- partment, as a new system was required to provide improved service to customers and bring the department into Y2K compliance, and should have been removed from the cost calculation. Also included in recent operating expenses was a one- time cost to purchase private mortgage insurance, which is now paid for by the borrowers at the time their loans are funded. Also during this period, program changes occurred that added to the cost of pro- cessing, such as use of VA guarantees. This added additional time required to process a loan; however, the guarantee offsets the risk of loss, and this should be factored into the calculation. • The audit report asserts that the loan processing times for the department exceed 8 loan processing times reported by commercial lenders. The report states that Program staff currently take 44.7 hours to process a loan. However, per the BCP report that contained this information, the actual number of hours to process a loan is 29.766, not 44.7. The 44.7 hours used by the audit staff included 4.23 hours for loan origination/public relations; 6.343 hours for prequalifications and 4.319 hours for home improvement loans. These tasks are not related to the loan process and are not included in the processing time for the department’s loans or conventional loans. C A L I F O R N I A S T A T E A U D I T O R 69 The increase in loan processing time since 1995-96 is due to the changes in the 8 loan program that have occurred over the last two years, such as the inception of the VA loan guarantee program. The conclusion that the department spend $800,000 more to process loans this 9 year than if the staff “worked as quickly as they did in fiscal year 1995-96” is there- fore not correct. Per the attached Exhibit 4 (see Appendix), there has been a dramatic increase in the number of loans processed per employee since 1995. During the period July 1, 1999 through May 18, 2000, 54 employees processed 2,995 loans compared to 1995/96 wherein 78.5 employees processed 984 loans. Considering that private sector lenders require 10 to 26 hours to process a loan 0 (according to the audit report), and Cal-Vet has additional steps in its processing that other lenders do not have (eligibility review and obtaining documents; funding allocation), 29.77 hours is reasonable and in no way “excessive”. In addition, the hours required by outside lenders does not include the amount of time required to handle all of the public information calls handled by the department’s loan process- ing staff. The outside lenders have other personnel assigned this responsibility, whose time is not included in the loan processing time. The Department has Inappropriately Charged Program Funds for Administrative Costs • The department acknowledges that all administrative costs should be charged to the appropriate program, and has contracted with an external consultant to perform a functional review of administrative services and to develop a cost allocation model. The Department has not Completed Implementation of Its Reengineering Plan • The department agrees with the audit report description of the business reengineering process that began in 1996 and continues to the present. Of the 42 approved recommendations, the department has successfully implemented 32 of the recommendations that resulted from the process, including implementing MITAS, centralizing loan contract servicing, adopting new underwriting standards, instituting mortgage insurance and becoming an approved lender for the U.S. Department of Veterans Affairs Loan Guaranty Program, improving management of delinquent and foreclosed contracts, and reducing losses from insurance programs. Of the 10 remaining recommendations, all are currently in process. 70 C A L I F O R N I A S T A T E A U D I T O R Lengthy Processing Times Increase Costs and Engender Dissatisfaction with the Program. • The department agrees that loan processing time has increased due to heavy loan volume and is a cause of some customer dissatisfaction, but many customers are satisfied with the Program’s product and service. An analysis was done of cus- tomer satisfaction surveys received since February 1, 2000 on recently funded loans. The results are shown below in Table 1. Excellent Good Fair Poor Yes No Satisfied Dissatisfied Service: 35% 37% 20% 8% Application easy to understand? 89% 11% Processing steps clear? 74% 26% Satisfied with processing speed? 51% 49% Satisfied with phone information? 70% 30% Recommend the Program to others? 95% 5% Satisfaction level with D.O. service: 88% 12% Table 1: Results of Customer Satisfaction Surveys • However, increased processing time was inevitable because loan volume almost tripled (from 967 in 1997/98 to 1607 in 1998/99, to 2,995 fiscal year to date (as of May 19, 2000), while staff levels remained constant. • The department is in the process of decreasing processing times by hiring and training additional loan processing staff, and through greater efficiencies gained through centralization of the loan underwriting functions and use of an interactive Internet presence for on-line loan application. q • In the audit report analysis of processing time, loans on homes under construction were included in their calculation of days (see Figure 6 in the audit report). The department does not agree that these loans should be considered, as these are loans that cannot be processed while under construction (which could take 90 days or more). Approximately 60% of our current application volume consist of these loans, and the department has no control over their processing times. The audit report indicates that only 8% of applications are processed under 60 days, 19% are processed 60 – 90 days, 41% are processed 91 to 120 days, and 32% are pro- cessed over 120 days. The department disagrees with these figures for the rea- sons stated, and believes the following are more correct: 33% are processed in less that 60 days, 44% are processed in 60 – 90 days, 14% are processed in 90 – 120 days, and only 9% are processed over 120 days (see Exhibit 5 in the Appen- dix). Over a 9 month period, from July 1, 1999 to March 31, 2000, the average loan processing time for all funded loans was 76 days, and for “existing” homes was 72 days. C A L I F O R N I A S T A T E A U D I T O R 71 w • The audit report states that industry lenders process loans in 30 to 45 days on average. What is not taken into consideration is that these processing times are based on 4business days, while Cal-Vet’s processing time as reported in the previ- ous paragraph, is based on calendar days. • The audit report indicates that some customers have had to obtain bridge loans due to processing delays. Only two out of 2,995 loans funded this fiscal year have had to use this alternative. It is by no means a common occurrence. Exceptionally Low Interest Rates and a New Mortgage Broker Program have In- creased Loan Volume • The department agrees that exceptionally low interest rates and a new Mortgage Broker Program have increased loan volume. However, the department is taking steps to hire additional loan processing staff and also control loan volume through interest rate adjustment, as discussed in Chapter 1. Audit Recommendations: • The department should complete the steps necessary to ensure its direct and indirect costs are properly charged to those programs served by the administrative staff; and The Department should identify the amount of Cal-Vet Program funds it has used for activities outside the program and seek reimbursement from other appropriate state funds. The department concurs with the audit recommendation to complete steps neces- sary to ensure that direct and indirect administrative costs are properly charged to those programs served by administrative tasks. The department has no way to ascertain the accurate allocation of past activities. As stated by the auditors, the department does not track the activities of administrative staff. Consequently, neither the department nor the auditors have been able to accurately identify the correct appropriation to be charged for such activities in the present or the past. • Complete centralization of loan processing. A small central underwriting group was established in 1998 which has continued to expand as the planned segmented centralization of loan processing has occurred. The Loan Processing Unit is currently processing approximately 35-40% of the applications being received statewide. Space issues were barriers to full implemen- tation of the Unit in the past, but space has now been identified and centralization should be complete within the next four months. 4 Source: 1991 Arthur Young report; verified May, 2000 as a current practice by contact with industry lenders 72 C A L I F O R N I A S T A T E A U D I T O R • Track industry standards and its own operations and develop and implement workload standards it can use in staffing and budgeting decisions for its field offices and headquarters office. The Division currently tracks industry standards. We have used this information to determine informal workload, and this information was factored into the Program’s recent BCP request for positions. Temporary staff is being hired to document standards. • Develop a field office-staffing model appropriate for the Cal-Vet program’s current loan volume. The Statewide Operations Manager explained the current staffing model with the audit team (a staffing ratio of one Property Agent to one Program Technician, as has been the historical ratio). The positions requested in the BCP were based on that model. • Update its underwriting manual and training plan and provide uniform training to all Cal-Vet program staff. The Program has been in the process of updating its Operations Manuals over the past few months, with each Program Unit updating its section under the coordina- tion of the Management Analysis Unit. The Marketing Unit is currently coordinating regular monthly training sessions of staff on all Program areas, and all loan pro- cessing staff in the field offices and the Loan Processing Unit have received and continue to receive training regarding the loan process. • Train mortgage brokers who receive and process Cal-Vet loan packages; Monitor application packages submitted brokers, either internally or through the use of a contracted manager, and limit participating brokers to those who consistently pro- vide loan packages that meet the program’s requirements. The Program is currently finalizing an IFP for distribution. The selected company will train, certify, and monitor the Mortgage Brokers for the department. C A L I F O R N I A S T A T E A U D I T O R 73 CHAPTER 3: The Department disagrees that inadequate management of the new information system increases costs and creates doubt about the reliability of program data. Audit Report Conclusions/Responses: The department does not agree with the broad statement in the Audit Report that “Inad- equate Management of the new information system increases costs and creates doubt about the reliability of Program data”, and that the program is not operating “properly”. • The implementation of MITAS was a major undertaking for the department. In addition to the fact that it involved a change in the software used for virtually all facets of the Cal-Vet Loan Program, including origination, servicing, accounting, cashiering, and bond finance, the method of servicing accounts also changed. Prior to implementation the department used a capitalization system for individual loan accounting where all receipts and disbursements were posted to the loan balance. MITAS changed the system to an impound (or escrow) accounting system which is prevalent in the industry. This change by itself had a major affect on loan origination, loan servicing, accounting and cashiering functions and required significant changes in staff’s approach to their functions. • Further complicating the implementation were drastic increases in outside interest rates, which were beyond the Department’s control that resulted in drastic increases in loan applications, and the necessity for the department to complete Y2K preparations. • Despite these challenges, the implementation proceeded with conversion occurring on the July 1 target date. No testing critical to conversion was omitted, although some testing of less critical functions was curtailed. Since conversion, no data conversion e issue affecting a significant portion of the loan portfolio has been identified. MITAS staff has been thoroughly responsive in providing assistance in resolving any and all issues in the form of advice, training, and system changes as necessary. The New System Still Fails to Meet Program Needs • In response to tracking MITAS one-time expenditures, Exhibit 3 (See Appendix) is a r spreadsheet that was developed in May 1999 to track MITAS expenditures. This spreadsheet shows tracking was actually taking place on many of the expenditures listed in Chapter 3, Table 5. Not tracked on this spreadsheet are “One-Time Consultant Costs to Oversee Project Implementation” and “One-Time Staff Costs”. Prior to May 1999, tracking expenditures was manually recorded on the schedule. 74 C A L I F O R N I A S T A T E A U D I T O R t • The audit report gives the impression that the MITAS system does not work completely, but does not quantify or specifically list what is not functioning properly. Of the few problems encountered with MITAS, most were due to human error because staff has not yet been through the learning curve. As staff becomes more accustomed to doing business on the system, fewer errors are seen. • The Program has been functioning from the MITAS system since July 1, 1999. y Through the conversions and data testing, the department has determined that the data on the system is correct. The Program is able to obtain the data necessary for day-to-day operations and certain reports (such as pipeline, funding, delinquent ac- counts), but has not yet combined certain data into “trend” reports that duplicate reports prepared prior to MITAS implementation. • In many areas, the MITAS system has improved contract servicing. Many manual functions can now be performed by the system, sparing staff from many word process- ing and manual functions. Staff must enter more data on the loan processing side, but consequently more data is available to retrieve. • The MITAS system resolved the Program’s Y2K compliance issues. Management Lacked Commitment to the Project • The audit report gives the impression that MITAS project delays were strictly a result of u poor management. The report did not mention delays resulting from the actual date CDVA received project approval from the Department of Information Technology (DOIT). As indicated in the FSR, the target date for project approval was January 15, 1998. The actual approval date occurred on March 20, 1998. This consequently pushed the original production implementation date from January 1, 1999 to March 1, 1999. Given this delay and Accounting Unit’s concerns regarding year-end closing, the business decision to change the production implementation to July 1 1999 was made. Attached are copies of the FSR Project Management Schedule, DOIT Project Approval Letter, DOIT Quarterly status report and Project Status Meeting Minutes (see Exhibits 6, 7, 8 and 9 of the Appendix). Planned System Testing Was Not Performed • The audit report gives the impression that very little testing was performed. Unit i and System testing logs were used to track each test that was performed on the MITAS system. The Unit testing logs indicate those tests that were performed to verify system requirements established in the RFP. System test logs show that there was actually a total of 150 Unit Test Scripts of which 58 were completed and the remaining 92 scripts were tested but pending resolution from users and MITAS. Realizing the importance of system testing for business needs, the Implementation Team did in fact decide to proceed with system testing without C A L I F O R N I A S T A T E A U D I T O R 75 actually completing Unit testing. The Business Function Log (system test logs) shows a total of 101 system test scripts of which 83 (82%) were completed and 18 (18%) were not completed. The 18 uncompleted scripts were identified as non- critical and were given lower testing priority. • By prioritizing the testing and through concentrated training in the accounting area, the Department did not miss a single day in funding new loans or posting payments received to the veterans’ accounts. Although experiencing some problems over the past 10-11 months, as is normal with the implementation of any new system, espe- o cially one of this magnitude, department staff expediently resolved each issue and in most cases improved the process along the way. • Although system testing may not have been completed by the July 1, 1999 conver- sion date, testing has continued in a separate “test database” before any new (untested) process or transaction is run in the live system. This has been very successful and problems encountered, for example, in the accounting area are communicated to department staff at on-going weekly meetings and/or the Mitas Group, as necessary, and are fully documented from issue to resolution. A large number of reconciliations are performed on a monthly basis to attest to the integrity of the system. The Department has not been able to Produce Critical Program and Financial Infor- mation • The system generates financial reports and through detailed reconciliations at both the account and transaction level, the department feels confident that the system accu- rately accumulates and reports data. The Department Has Not Adequately Safeguarded Cal-Vet Program Data and Assets • The audit report indicates that users are free to change any information they wish and p this is not the case. Each user is placed in a group by their work unit. This group has a specific menu and security menu and security rights assigned. Some users can also be assigned extra menu items and rights as needed to perform their job. If a user does not have any of the above set up, they will receive a menu with “read only” rights. • The Accounting Office thoroughly reviewed and established criteria for access to the a accounting/cashiers related modules for strict compliance with standard internal con- trols. Procedures for the handling of receipts are in compliance with SAM 8080. • The audit report indicates that users are free to change any information they wish; this is not the case. Each user is placed in a group by their work unit. This group has a specific menu and security menu and security rights assigned. Some users can also 76 C A L I F O R N I A S T A T E A U D I T O R be assigned extra menu items and rights as needed to perform their job. If a user does not have any of the above setup, then they get a menu with “read only” rights. • ISD agrees that current procedures for tracking changes are faulty. ISD is currently researching various automated tools, such as TRACK IT and REMEDY to assist in improving our current change control process. Audit Recommendations: The department disagrees with most of the Chapter 3 recommendations: • To ensure that its integrated information system functions reliably, the department should convene a centralized implementation team . . . . The department does not believe it would be productive to redirect staff and resources s to “cover old ground”. This would disrupt processing and exacerbate those problems that the audit is criticizing in its report. The system is tested on an ongoing basis. Every billing cycle is a “test” of data. Through the TQM process and MITAS implementation process, the department has already adequately identified business processes; taken steps necessary to ensure that the system’s capabilities serve the current business processes; developed and per- formed tests for the system to ensure that it accurately and completely processes transactions and accumulates program and financial data; identified existing implemen- tation issues to ensure current efforts solve the problems (key users meet weekly to discuss/resolve issues); developed and maintained a method for tracking system issues from discovery to resolution (implementation logs were kept to track issue status). • To adequately safeguard program data and assets, the department should implement a security policy to limit system access to employees who are properly authorized, and ensure these employees’ access is not incompatible with their other duties. A security policy has been implemented, although the department agrees that it should be monitored closely to ensure that staff who transfer maintain the appropriate access. C A L I F O R N I A S T A T E A U D I T O R 77 Cal-Vet Program Benefits The department provides unique benefits to California’s veterans that no other lender offers. A Cal-Vet Loan is truly the veterans “loan of choice”. • Interest rates are the lowest in the nation • High maximum loan limit ($250,000) • Low downpayment (Minimum of 2% or 3%, depending on loan guarantee source) • Low-cost construction loans • Earthquake and flood coverage that is more comprehensive and less expensive than any available on the market • Home Improvement Loans at below market rates • Life-of-loan service by the department (the loans are not sold) 78 C A L I F O R N I A S T A T E A U D I T O R APPENDIX Exhibit 1: Historical Funding Data (New Contracts of Purchase) Exhibit 2: Interest Rate Comparison Exhibit 3: MITAS Cost Spreadsheet Exhibit 4: Loan Applications Funded by Staffing Levels Exhibit 5: Loan Processing Times for All Loans Funded from 7/1/1999 –3/31/ 2000 Exhibit 6: FSR Project Management Schedule Exhibit 7: DOIT Project Approval Letter Exhibit 8: DOIT Quarterly Status Report Exhibit 9: Project Status Meeting Minutes C A L I F O R N I A S T A T E A U D I T O R 79 100 C A L I F O R N I A S T A T E A U D I T O R COMMENTS California State Auditor’s Comments on the Response From the California Department of Veterans Affairs T o provide clarity and perspective, we are commenting on the response to our audit report from the California Department of Veterans Affairs (department). The numbers below correspond to the numbers we have placed in the response. 1 The department errs in suggesting that our projections appear to be based on the assumption that Vietnam era veterans will not be interested in the program due to their age. As we state on pages 1, 17, 18, and 19 of our report, our estimates are based on the California Veterans Farm and Home Purchase Program’s (Cal-Vet program) current share of the mortgage market and federal eligibility restrictions for loans backed by Qualified Veterans Mortgage Bonds (QVMBs). We determined the Cal-Vet program’s 0.17 percent current market share by annualizing the QVMB-backed loans made in the first half of fiscal year 1999- 2000 divided by the current estimate of QVMB-eligible veterans residing in California. By multiplying the Cal-Vet program’s current market share by the projected number of veterans still eligible in 2010, we estimate that as few as 117 California veter- ans will apply for QVMB-backed loans in the last year of the decade. Although we repeatedly asked, the department never provided us with an estimate of the number of veterans who might apply for QVMB-backed loans in 2010 that it believes is more accurate than ours. 2 The department may be overly optimistic when it states there is little doubt that additional allocations of Qualified Mortgage Bonds (QMBs) are available for Cal-Vet veteran first-time home buyers and that unrestricted funds will continue to regenerate. As we state on page 21, any allocation the department would receive from the California Debt Limit Allocation Committee (CDLAC) to sell additional QMBs will reduce the allocations given to competing lenders, such as the California Housing Finance Agency, who serve a broader market of buyers, includ- ing veterans. We also state on page 21 the annual number of C A L I F O R N I A S T A T E A U D I T O R 101 loans funded with recycled funds in 2007 will not exceed 700, and that number will continue to decline in each successive year. 3 At the time of our fieldwork, the department could not provide a detailed analysis of veteran eligibility that included critical eligibility factors such as veterans’ discharge dates, but instead provided us with a projection of the number of veterans in California as of July 1998, based on 1990 census data produced by the U.S. Department of Veterans Affairs (VA). While such data may be suitable for a point-in-time population estimate of California veterans, without further refinement using draft or military service enlistment and discharge dates as we did, such data is useless in estimating the future demand for its QVMB- backed loans. 4 The department has misrepresented the extent to which they have shared ideas with us during the course of the audit. According to the documents we reviewed, the department’s short-term strategy has been, and continues to be, to boost its overall loan portfolio. After repeated requests, the department has been unwilling to share any long-term plans it may have developed, stating that such plans would first have to be approved by its board. On several occasions we invited the department to provide us with any long-term plans for the Cal-Vet program so that we could include them in the body of our report, but our requests were ignored. The department finally did provide us with a draft of its plan to adjust interest rates as of May 16, 2000. However, the draft had not been submitted to the board for its approval as of that date. Without the approval of the board that sets its policy, we could not assess the viability of the department’s draft. 5 The department has misrepresented the facts in asserting that marketing benefits other than its below-market interest rates increased demand for its loans in fiscal year 1994-95 over the previous fiscal year. Contrary to what it has stated in the response to our report, the department’s average interest rate in fiscal year 1994-95 was actually lower than conventional rates. According to its latest Official Bond Statement, the department’s average interest rate in fiscal year 1994-95 was 7.8 percent, almost a full percentage point lower than the 8.7 percent aver- age conventional rate. In contrast, in fiscal year 1993-94 the department’s average interest rate of 8 percent was 0.7 percent higher than the average commercial rate of 7.3 percent. While the department is correct in stating that it funded more loans in 102 C A L I F O R N I A S T A T E A U D I T O R fiscal year 1994-95 than it did the preceding fiscal year, the increase in demand is more likely a function of its lower interest rate in that fiscal year. 6 We disagree. Our calculations of the program’s annual adminis- trative costs per loan on page 28 are based on the department’s audited financial statements, which do not show a decline in administrative costs per loan in fiscal year 1998-99. Further- more, by netting operating revenue against administrative expenses to portray a decline in administrative expenses, the department diverts attention away from our key point. That is, when costs for processing and servicing loans escalate, fewer funds are available for loans to veterans. Program funds are used to provide loans to veterans and to pay for the Cal-Vet program’s operating costs. Whether these funds come from the proceeds of bond sales, interest earnings, or fees charged for loans, the better the department is able to control its administra- tive costs for the Cal-Vet program, the more funds will be available for loans to veterans. 7 The department’s statement that certain costs should not be considered as expenses for specific periods is inconsistent with the way it treated these expenses in its audited financial statements. The financial statements reflect the judgments of management, that its independent auditor agreed with, related to system costs that should be expensed in a single year or allocated over future years. Our report relies on this audited data. 8 Evidence we collected from other lenders suggests that the costs added to the department’s loan processing due to its participation in the VA loan guarantee program should not be significant. Many of the commercial lenders we spoke with stated that processing a VA guaranteed loan only added approxi- mately one-half hour to their processing times. However, we agree with the department that 4.23 hours for loan originations and public relations should be deleted from our calculation and we have therefore revised our text. The lenders we interviewed were mixed as to whether they included time for borrower pre-qualifying activities. However, most stated pre- qualifying activities took little time, from 15 to 60 minutes, in contrast to the 6.34 hours the department cites. Thus, we did not change our report for pre-qualifications. Also, no change was warranted for home improvement loans since all lenders we contacted processed these types of loans. C A L I F O R N I A S T A T E A U D I T O R 103 9 Again, the department is presenting information that simply confuses the issue. Our report discusses administrative costs for each loan, whereas the department now presents information about loans processed per employee. Discussing loans processed per employee as a measure of increased efficiency is inconsistent with the methodology it used in a 1996 study prepared by the Cal-Vet program and the same methodology the department included in its September 1999 budget change proposal request- ing more loan-processing staff. As we state on page 28, based on those documents loan processing time has increased. 0 We disagree with the department’s calculation of its processing time of 29.77 hours per loan. Time spent determining eligibility, obtaining necessary documents, and fielding calls to determine whether borrowers are eligible for various government lending programs are all loan activities common among the lenders we contacted and have been included in their computations of loan processing times. q The department is incorrect when it states we included construction loans in our calculation of the number of days the department takes to process loans. As stated on page 37, we excluded new construction loans from our calculations. Despite numerous attempts to understand the department’s concerns about our calculation of the number of days to process loans, the first time we were provided its calculation was in its response to our audit report. Consequently, we have not been able to do an in-depth analysis of the data. However, the department’s nine-month averages of 76 days for processing all types of loans and 72 days for processing loans for existing homes seems to support our conclusion that the average processing times for Cal-Vet loans far exceed the 30- to 45-day average processing time reported by commercial lenders. w The department is mistaken. We contacted a number of com- mercial lenders who stated that the average loan processing times they provided to us were based on calendar days, not business days. e The department misrepresents the condition when it states that no data conversion issue affecting a significant portion of the loan portfolio has been identified. First, as we state on page 50 of our report, the department has data conversion errors affect- ing payments that borrowers have made relating to their escrow accounts for 1,500 loans, which is a significant number. In addition, as stated on page 50, the department cannot know the 104 C A L I F O R N I A S T A T E A U D I T O R extent of problems it may find in its more than 32,000 accounts because it has not yet performed an analysis of the escrow balances for those accounts. Further, staff could not recall the extent to which they tested how well the system converted loan account information into the new database because they did not retain documentation of the tests they said they performed less than a year ago. r In its response to our report, the department acknowledges that it did not track all of the one-time costs for implementing the system. After repeated requests for all pertinent data, the department never provided to us information it now says existed in May 1999. Rather, we obtained from the department’s accounting unit and information services division the data shown in the table on page 45. When we inquired with account- ing and information services staff, we were told that one-time implementation costs were not separately tracked from ongoing costs. Accounting staff was able to provide us with separate invoice payment information for the consultant’s costs to oversee system implementation. t The department is incorrect. On page 50 we clearly list two of the more serious examples of the system failing to function properly. y We disagree that the department has performed sufficient testing to determine data accuracy and reliability. As we state on pages 47 through 49, the department failed to conduct over half of the 358 tests planned to verify that the system processed data correctly. u The department’s reference to a 60-day delay in receiving approval from the Department of Information Technology appears to be an attempt to deflect responsibility away from management. That time is unrelated to management’s inatten- tiveness to the project as discussed on pages 46 through 54. i The information the department presents in its response is not consistent with information its staff shared with us during our fieldwork. We obtained schedules of unit and system tests from the information services division staff person who was the testing coordinator during project implementation. These schedules were intended to be used as logs of testing performed by unit staff, but were not. In addition, we obtained records of tests performed and the actual test scripts from the unit C A L I F O R N I A S T A T E A U D I T O R 105 staff who were responsible for performing those tests. These documents showed that the department failed to perform 185 out of 358 total planned unit and system tests. o The department overstates the extent to which they promptly resolve implementation problems. As stated on pages 48 and 49, there were numerous instances where testing identified prob- lems but there was no evidence these problems were resolved. Further, as stated on page 50, of the nearly $1 million in unap- plied veterans payments as of March 15, 2000, over $388,000 remained unapplied after more than three months. p The department has mischaracterized our finding when it says our report indicates that users are free to change any informa- tion they wish. On page 52 we state that changes to the system are not properly authorized, documented, and approved. As a result, the department does not ensure that changes to one component of the system will not adversely affect another, nor does it maintain documentation of changes that will aid it in troubleshooting system errors. In addition, on pages 52 and 53 we describe specific deficiencies in the department’s controls over access to the system that unnecessarily expose crucial data to risk of loss or misuse. a The department is mistaken in asserting that its procedures for handling cash receipts are in compliance with the State Administrative Manual (SAM), Section 8080. As indicated on page 52, the cashiers in the accounting unit currently are responsible for receiving and depositing Cal-Vet payments, as well as entering the cash receipts data into the information system. These two duties are incompatible and violate SAM, Section 8080.1. s The department misunderstands our recommendations when it states that following them would be nonproductive and would redirect staff and resources to “cover old ground.” As we indicate throughout Chapter 3, the department abandoned a well- prepared implementation plan in order to meet its revised July 1, 1999, deadline to take the system into production. As we describe on page 54, the revised implementation plan the department adopted in an attempt to regain some centralized control over the system implementation was much less rigorous and failed to include many of the tasks required in the original plan. For those reasons, our recommendations are aimed at areas and tasks that the department only partially completed, or never completed at all. 106 C A L I F O R N I A S T A T E A U D I T O R cc: Members of the Legislature Office of the Lieutenant Governor Milton Marks Commission on California State Government Organization and Economy Department of Finance Attorney General State Controller State Treasurer Legislative Analyst Senate Office of Research California Research Bureau Capitol Press C A L I F O R N I A S T A T E A U D I T O R 107