Summary

REPORT BY THE STATE AUDITOR OF CALIFORNIA A REVIEW OF THE STATE'S BOND SALES FOR 1991 AND 1992 93016 January 1994 A Review of the State's Bond Sales for 1991 and 1992 93016, January 1994 California State Auditor Bureau of State Audits Table of Contents Page Summary S-1 Introduction 1 Chapters 1 A Review of the State Treasurer's Process for Selecting Underwriters of Negotiated Bond Sales 7 Recommendation 15 2 Maintenance of Required Records at the State Treasurer's Office, the California Housing Finance Agency, and Three State Financing Authorities 17 Recommendation 22 3 Comparison of California Bond Issues With Other Government Bond Issues 23 Appendix Negotiated Bond Issues for Which the State Treasurer Selected the Underwriters 29 Responses State Treasurer's Office 55 to the California State Auditor's Comments 63 Audit California Educational Facilities Authority 65 California State Auditor's Comments 69 California Pollution Control Financing Authority 71 California Housing Finance Agency 73 BLANK PAGE INSERTED FOR REPRODUCTION PURPOSES ONLY Summary Results in Brief Chapter 1434, Statutes of 1990, added Section 5703 to the California Government Code, which specifies that the state treasurer is the sole authority for selecting the underwriters to negotiate state bond sales, except for California Housing Finance Agency (CHFA) bond sales. Section 5703 of the Government Code also requires the state treasurer to use a competitive process in selecting underwriters for negotiated bond offerings and requires the state treasurer, financing authorities, and the CHFA to maintain records of the costs of issuance of negotiated bond sales. In addition, this section requires the state treasurer to maintain certain records related to bond issues that are sold by competitive bid. We reviewed the cost records of negotiated bond sales that the State Treasurer's Office (STO), CHFA, and state financing authorities maintained and reviewed the records of competitive sales that the STO maintained. In addition, we determined whether Government Code, Section 5703, was being fully implemented and compared costs with similar initial bond offerings in other states. During our review, we noted the following conditions:  From January 1991 through December 1992, the STO selected the lead underwriter through a competitive process for all except one of 83 negotiated bond issues we could review. According to the Government Code, Section 5703, a competitive process may be conducted on an issue-by-issue basis or may be used to establish one or more pools of underwriters for various types of negotiated issues. For one bond issue, the STO did not competitively select the lead underwriter because staff believed the relatively small sale would provide a good opportunity for a Target Business Enterprise firm to obtain the experience of being a lead underwriter.  The Government Code, Section 5703, requires that the state treasurer establish a competitive process for the selection of underwriters. However, the law does not specifically state that the state treasurer is to establish a competitive process for co-managing underwriters (co-managers) as well as lead underwriters in a negotiated bond sale. In view of this, the state treasurer interprets the statute to allow for the selection of only lead underwriters through a competitive process. When the state S-1 treasurer selected only lead underwriters through a competitive process, a portion of the State's underwriting business was awarded without the benefit of competition. This is because the state treasurer, in certain instances, did not select co-managers through a competitive process. This primarily occurred from January 1991 until April 1992, when a new policy was implemented. Under the new policy, the state treasurer uses its competitively-established pools of firms to select co-managers for all negotiated sales, other than short-term interim notes and revenue anticipation notes (RANs). Instead, the RANs, which accounted for 9 percent of the amount that co-managers earned during the two years we reviewed, are selected based on various factors including past performance on RANs financings. However, because the state treasurer does not interpret the law to require competitive selection of co-managers, the current policy of selecting co-managers for the other negotiated sales could change. Thus, in spite of the recent policy change, we believe that more needs to be done to ensure a competitive process is used to select all underwriters. We believe the Government Code, Section 5703, is vague in one of its key provisions, which could keep the objective of attaining competition in the selection of all underwriters for state bond sales from being fully realized. While we realize that state departments are generally afforded broad latitude in interpreting those laws that they administer, we feel that excluding co-managers from the competitive process for selecting underwriters is not in the best interests of the State. If in the future a state treasurer does not select co-managers through a competitive process, the State could award millions of dollars of underwriter's discount annually without the benefit of competition. For the two years we reviewed, the total underwriter's discount earned by all co-managers the state treasurer selected was $21.4 million.  Three of the five entities we reviewed did not collect all of the cost information the Government Code, Section 5703, requires to be maintained for negotiated bond sales. During 1991 and 1992, the California Educational Facilities Authority (CEFA) did not maintain all the cost information required by the Government Code, Section 5703(e), for bond issues sold by private placement. We reviewed nine files at CEFA and found that three of the files were missing records documenting $11,000 to $22,000, or from 18 to 28 percent, of the total underwriter's discount and costs of S-2 issuance. The STO and the California Pollution Control Financing Authority maintained most of the required cost information, although 2 of 12 files reviewed at STO were missing items documenting $57,000 to $60,000, or 1 to 2 percent of the costs, and 2 of 5 files reviewed at the California Pollution Control Financing Authority did not sufficiently identify certain costs. We reviewed 29 files at the California Health Facilities Financing Authority and 8 files at the California Housing Finance Agency and found that both agencies maintained the required cost information in the files we reviewed.  During 1991 and 1992, only the STO sold bonds by competitive bid. We reviewed 18 of 24 bond issues sold by competitive bid during 1991 and 1992 and found that in all 18 bond sales, the STO maintained records of all bids and bid verifications as required by the Government Code, Section 5703(f). In addition, the STO awarded the bonds in each sale we reviewed to the underwriter who submitted the bid with the lowest true interest cost.  We compared the true interest costs of ten bond issues sold by other governmental entities with the true interest costs of nine California bond issues. We found no indication from the information obtained in this limited review that California is paying more than necessary for interest costs on state bonds. Seven of the nine California bond issues had lower true interest costs than the comparison bond issues. In the other two bond offerings, the true interest costs differed by .0184 and .226 percentage points. The differences in the true interest costs may be attributable to the different features of the bonds or to differing market conditions at the time of the bond sales.  We recommend the Legislature consider amending the Government Code, Section 5703, to specifically state that co-managers as well as lead underwriters be selected through a competitive process. S-3 Agency The State Treasurer's Office generally concurs with the report but Comments believes that the law clearly was written to require the state treasurer to use the competitive process solely for selecting the lead underwriter on a bond sale. Our comments follow the response from the State Treasurer's Office. The California Educational Facilities Authority disagreed with our statement in the report that it no longer had in its files a list of underwriters on the pool that was in effect from January 1991 through April 1992. Our comments follow the response from the California Educational Facilities Authority. The California Pollution Control Financing Authority and the California Housing Finance Agency agreed with the sections of the report pertaining to them. The California Health Facilities Financing Authority chose not to respond to the report. S-4 Left intentionally blank S-5 Introduction The California Government Code, Sections 5700 and 5702, states that the state treasurer shall be the sole agent for offering and selling bonds issued by any state department, board, agency, or authority. Chapter 1434, Statutes of 1990, added Section 5703 to the Government Code, which specifies that, in fulfilling the duties of agent for offering and selling bonds, the state treasurer is the sole authority for selecting the underwriters to negotiate state bond sales, except for California Housing Finance Agency bond sales. Section 5703 of the Government Code also requires the state treasurer to use a competitive process in selecting underwriters for negotiated bond offerings and requires the state treasurer, financing authorities, and the California Housing Finance Agency to maintain records of the costs of issuance of negotiated bond sales. In addition, this section requires the state treasurer to maintain certain records related to bond issues that are sold by competitive bid. The Government Code, Section 5703(d), and the California Health and Safety Code, Sections 51050(f)(1) and 51358, exempt the California Housing Finance Agency from the requirement to select underwriters for negotiated bond sales through a competitive process. A wide variety of state agencies, departments, and authorities issue bonds. For example, during 1991 and 1992, bonds were issued by the Department of Water Resources, the Department of Veterans Affairs, the Regents of the University of California, the California State University, the State Public Works Board, the California Housing Finance Agency, the California Health Facilities Financing Authority, the California Educational Facilities Financing Authority, and the California Pollution Control Financing Authority, and others. Financing authorities are governmental entities that issue bonds on behalf of private nonprofit or other public organizations. The financing authority issues bonds and uses the proceeds to provide financing for the other organization. The principal and interest of the bonds are repaid by the other organization. The interest on the bonds is tax-exempt; therefore, the borrowing organization pays a lower interest rate than if it had issued taxable bonds. This type of financing is referred to as conduit financing, and is used to finance projects that serve the public interest. For example, the California Health Facilities Financing Authority issues bonds on behalf of private nonprofit or public health facilities, such as hospitals, skilled nursing facilities, adult day health centers, community clinics, and child care facilities. 1 Competitive Bids Bonds may be sold using one of two types of sale procedures: a and Negotiated competitive bid or a negotiated sale. In a competitive bid, the issuer determines the date and principal amount of the bond sale and invites Sales Offer underwriters to submit sealed bids for the bonds. The issuer sells the Different bonds to the underwriter who offers to buy the bonds at the lowest Advantages interest cost. The underwriter then offers the bonds for sale to the public. In a negotiated sale, the issuer selects the underwriter before the bond sale, and the underwriter works closely with the issuer in activities related to the bond sale, such as developing financing alternatives and preparing legal documents. In a negotiated sale, bonds are usually sold to an underwriter who then offers the bonds for sale to the public. The underwriter's compensation lies in the difference between the price it pays for the bonds and the price at which it sells the bonds to the public and is referred to as underwriter's discount. In certain instances, negotiated bonds are sold through a private placement. A private placement is another type of negotiated sale in which the issuer sells the bonds to a single investor, who usually does not offer them for sale to the public. Each type of bond sale offers advantages to the issuer depending upon the type of bonds the issuer is selling. When bond issues are not complex in structure and the issuer is well known to the investment community, competitive bids offer the advantage of creating competition among underwriters in an open market. For bond issues that are complex in structure or in which the investment community is not familiar with the issuer, a negotiated sale allows the underwriter more time to develop a successful marketing plan. Negotiated sales also offer the issuer flexibility in timing the bond sale to take advantage of beneficial market conditions. A private placement is generally used when the issuer anticipates difficulty in marketing the bonds or, for smaller bond issues, when use of a private placement lowers the costs related to issuing the bonds. The underwriters of a bond issue are liable for purchasing the bonds from the issuer whether or not they are able to sell all of the bonds to the public. For large bond issues, underwriter syndicates may be formed to share in the risk of purchasing the bonds. One underwriter usually assumes the role of lead underwriter and, on behalf of the syndicate, negotiates with the bond issuer, executes the bond purchase contract, and manages the syndicate accounts. The lead underwriter is compensated for these activities by being paid a portion of the underwriter's discount referred to as the management fee. A syndicate may include one or more co-managing underwriters (co-managers), who assume some of the management duties related to the bond sale 2 and share in the management fee as compensation for these duties. The lead underwriter and co-managers usually assume liability for purchasing a larger portion of the bond issue than the other members of the syndicate. In addition to the syndicate members, a group of firms called a selling group may be used to assist in the sale of the bonds. Members of the selling group do not assume liability for purchasing any of the bonds. They receive compensation in the form of a commission on the bonds they sell, usually referred to as takedown. Bonds Sold in During 1991 and 1992, the State Treasurer's Office (STO), three 1991 and 1992 financing authorities, and the California Housing Finance Agency (CHFA) sold a total of 137 bond issues, with a total principal value of Amounted to $31.17 billion: the California Educational Facilities Authority sold 20 $31.17 Billion bond issues with a total principal value of $431 million; the California Health Facilities Financing Authority sold 41 bond issues with a total principal value of $1.357 billion; CHFA sold 14 bond issues with a total principal value of $1.162 billion; and the California Pollution Control Financing Authority sold 20 bond issues with a total principal value of $782 million. The financing authorities and CHFA sold all of their bond issues by negotiated sale or private placement. The STO sold 42 bond issues on behalf of various state agencies. It sold 24 of the 42 bond issues, with a total principal value of $8.519 billion, by competitive bid. Of this amount, $7.113 billion were general obligation bonds. The STO sold the remaining 18 bond issues, with a total principal value of $18.919 billion, by negotiated sale. Of this amount, $16.05 billion were short-term revenue anticipation notes. Scope and Section 5703(g) of the Government Code required the Office of the Methodology Auditor General to audit the cost records of negotiated bond sales that the STO, the CHFA, and the state financing authorities maintained and to review the records of competitive sales that the STO maintained. In addition, Section 5703(h) of the Government Code required the Office of the Auditor General to report whether Section 5703 was being fully implemented, to make cost and interest rate comparisons with similar initial bond offerings in other states, where possible, and to submit a report to the Legislature for bonds sold during 1991 and 1992. Section 5703 requires a second report on bonds sold during 1993 and 1994. 3 The Office of the Auditor General had completed some of the audit work when the office was closed in December 1992. In accordance with the Government Code, Section 8546.8, the Bureau of State Audits, which began operations in May 1993, completed the remaining audit work and prepared this report. To determine how the state treasurer selected underwriters during 1991 and 1992, we interviewed STO and financing authority officials and examined documents on file at both the STO and financing authorities. These documents included Requests for Qualifications (RFQ) issued by the state treasurer and financing authorities for the various competitive processes conducted during 1991 and 1992, underwriter responses to the RFQ for the most recent competitive process, lists of selected underwriters, and the STO's reports on the underwriters used on each bond sale during 1991 and 1992. We determined whether the state treasurer selected the lead underwriters during 1991 and 1992 through the competitive process in effect for each bond issue. After it established new underwriter pools in March 1992, the STO began using the pools for selecting underwriters for all bond sales. Additionally, at that time, the STO implemented a new policy of selecting co-managers, as well as lead underwriters, from the pools. Thus, for bond issues sold from April 1992 to December 1992, we also determined whether the state treasurer selected the co-managers from the underwriter pools in effect during that time. We did not review the propriety of the state treasurer's selection or rejection of specific underwriters who responded to any of the RFQs. To determine if the state treasurer maintained the required records relating to bond issues sold by competitive bid, we reviewed the STO's records for a sample of bond issues the STO sold by competitive bid during 1991 and 1992. To determine if the state treasurer, the CHFA, and the three financing authorities maintained the required cost of issuance information for negotiated bond sales, we selected a sample of bond issues sold during 1991 and 1992 at each entity and reviewed the detailed summary of costs of issuance contained in its files. We compared the total of the cost records contained in the files with the total costs of issuance set forth in the official closing documents of the bond issue. 4 We determined whether the cost of issuance summaries identified all the costs of issuance paid from bond proceeds. We did not determine the propriety of the identified expenditures. When possible, we examined invoices related to individual cost of issuance items. However, the financing authorities do not make cost of issuance disbursements for most of their bond issues, and the STO does not make cost of issuance disbursements for the bond issues it sells on behalf of other state departments or the state universities. For those bond issues that the STO or financing authorities do not themselves make the cost of issuance disbursements, the STO or financing authority obtains the cost of issuance information from the underwriter, the issuing state agency, or the issuing private organization. This information is not supported in the files by copies of invoices or other supporting documentation; therefore, we were not able to audit this information. To obtain information about similar bonds issued by other governmental entities, we selected a sample of California state revenue bonds sold through negotiation during 1991 and 1992, defined what constituted a similar bond issue, and obtained information about similar bonds from Securities Data Corporation, a firm that collects and maintains information about government bond issues. 5 BLANK PAGE INSERTED FOR REPRODUCTION PURPOSES 6 Chapter 1 A Review of the State Treasurer's Process for Selecting Underwriters of Negotiated Bond Sales Chapter The California Government Code, Section 5703, identifies the state Summary treasurer as the sole authority for selecting underwriters for state bonds, except bonds sold by the California Housing Finance Agency, and requires the state treasurer to develop and implement a competitive process for selecting the underwriters of negotiated bond offerings. According to the general counsel for the State Treasurer's Office (STO), the Government Code, Section 5703, requires the state treasurer to select only the lead underwriter of negotiated bond sales through a competitive process. From January 1991 through December 1992, of 83 negotiated bond sales that we were able to review, the state treasurer selected all the lead underwriters through a competitive process except for one negotiated bond sale. In April 1992, the state treasurer began implementing a new policy of selecting co-managing underwriters (co-managers), as well as lead underwriters, through a competitive process. From April 1992, when the current underwriter pools became effective, to December 1992, the state treasurer selected both lead underwriters and co-managers from the competitively-established underwriter pools, except co-managers for the September 1992 revenue anticipation notes financing. The Government Code, Section 5703, requires that the state treasurer establish a competitive process for the selection of underwriters. However, the law does not specifically state that the state treasurer is to establish a competitive process for co-managing underwriters (co-managers) as well as lead underwriters in a negotiated bond sale. In view of this, the state treasurer interprets the statute to allow for the selection of only lead underwriters through a competitive process. When the state treasurer selected only lead underwriters through a competitive process, a portion of the State's underwriting business was awarded without the benefit of competition. This is because the state treasurer, in certain instances, did not select co-managers through a competitive process. This primarily occurred from January 1991 until April 1992, when a new policy was implemented. Under the new policy, the state treasurer uses its competitively-established pools of firms to select co-managers for all negotiated sales, other than short-term interim notes and revenue 7 anticipation notes (RANs). Instead, the RANs, which accounted for 9 percent of the amount that co-managers earned during the two years we reviewed, are selected based on various factors including past performance on RANs financings. However, because the state treasurer does not interpret the law to require competitive selection of co-managers, the current policy of selecting co-managers for the other negotiated sales could change. Thus, in spite of the recent policy change, we believe that more needs to be done to ensure a competitive process is used to select all underwriters. We believe the Government Code, Section 5703, is vague in one of its key provisions, which could keep the objective of attaining competition in the selection of all underwriters for state bond sales from being fully realized. While we realize that state departments are generally afforded broad latitude in interpreting those laws that they administer, we feel that excluding co-managers from the competitive process for selecting underwriters is not in the best interests of the State. If in the future a state treasurer does not select co-managers through a competitive process, the State could award millions of dollars of underwriter's discount annually without the benefit of competition. For the two years we reviewed, the total underwriter's discount earned by all co-managers the state treasurer selected was $21.4 million. We believe the Legislature should consider amending the Government Code, Section 5703, to specifically state that co-managers as well as lead underwriters be selected through a competitive process. We could not review the state treasurer's selection of lead underwriters at the California Educational Facilities Authority (CEFA) for bond sales occurring from January 1991 to April 1992 because the CEFA did not have in its files a list of the underwriters that were on the CEFA pool in effect during this time. Background The Government Code, Section 5703(a), identifies the state treasurer as the sole authority for selecting underwriters for state bonds and requires the state treasurer to develop and implement a competitive process for selecting the underwriters of negotiated bond offerings. This "competitive process" is different than the sale of bonds through competitive bid. We describe the differences between selling bonds through competitive bidding and selling bonds through a negotiated sale in the introduction of this report. The competitive process that the state treasurer is to develop and implement pursuant to the Government 8 Code, Section 5703, is a process for selecting underwriters for negotiated bond sales. This competitive process may be conducted on an issue-by-issue basis or may be used to establish one or more pools of underwriters for various types of negotiated issues. For the remainder of this report, any reference that we make to the competitive process for selecting underwriters refers to one of these two methods of selecting underwriters. The competitive process must have the following features:  It shall solicit written qualifications from at least 20 underwriting firms;  It shall consider the goals for minority and women business enterprise participation in professional bond service contracts;  The written submissions shall be available for inspection at the STO for at least 6 months; and  If a pool of underwriters is established, the competitive process shall be repeated at least every 24 months to reestablish the pool of underwriters. The Government Code, Section 5703, states that a competitive process may be implemented by establishing a pool but does not state that the underwriters in the pool must participate in an additional competitive process to be selected for a particular bond sale. Thus, the state treasurer may select any underwriter from the pool to participate in a bond sale. We did not review the process by which the state treasurer assigned underwriters from the pools to a particular bond sale. Section 5703(b) of the Government Code further specifies that for negotiated bond offerings by state financing authorities that act as conduits to provide financing to other public, nonprofit, or private organizations, the state treasurer shall use the competitive process to establish one or more pools of underwriters for each financing authority. The state treasurer may make additions to a financing authority pool without competitive solicitation, on a case-by-case determination upon the recommendation of a project applicant, when the state treasurer finds that the underwriter to be added has provided significant services 9 to the project applicant with the expectation of compensation for those services from underwriting the revenue bonds that will fund the applicant's project. The only exceptions to these requirements are provided by Sections 5703(c) and 5703(d) of the Government Code. Section 5703(c) states that the state treasurer may select underwriters for a negotiated sale of bonds by means other than as described in Section 5703(a) if the state treasurer makes a written finding that extraordinary market conditions do not allow enough time to do so without risking financial detriment to the State. Section 5703(d) states that Sections 5703(a), (b), and (c) shall not apply to the issuance of state bonds for which the state treasurer is precluded by statute from selecting underwriters. This section therefore exempts the CHFA from the requirement to select underwriters for negotiated bond sales through a competitive process because Section 51050(f)(1) of the California Health and Safety Code states that the procurement of underwriters for the CHFA is not subject to the review or approval of any other division or officer of state government. During this review, we discussed specific provisions of the Government Code, Section 5703, with the general counsel for the State Treasurer's Office (STO). Of particular interest to us was whether, in his opinion, the law was intended to require the state treasurer to select both the lead underwriter and the co-managers, or only the lead underwriter, of a negotiated bond sale through a competitive process. According to the general counsel for the STO, the intent of the law was to require the state treasurer to use a competitive process solely for selecting the lead underwriters. We discuss our conclusions regarding the STO's interpretation of the law later in the chapter. Although the state treasurer's interpretation is that the law does not require the selection of co-managers through a competitive process, the current state treasurer established a policy in December 1991 of selecting most of the co-managers, as well as the lead underwriters, through a competitive process. In April 1992, after it had established new underwriter pools, the STO began implementing the policy. Current In March 1992, the state treasurer established four pools of Underwriter underwriters through a Request for Qualifications (RFQ) process. The state treasurer established one pool for state negotiated offerings (SNO Selection Process pool) and one pool each for the California Educational Facilities Authority (CEFA), the California Health Facilities Financing Authority (CHFFA), and the California Pollution Control Financing Authority (CPCFA). The state treasurer established the pools for the selection of lead underwriters and co-managers for all negotiated bond sales, except 10 the short-term interim notes and revenue anticipation notes (RANs). According to a letter from the state treasurer to underwriters, the state treasurer planned to select the lead underwriters for the RANs from the SNO pool but planned to select the co-managers based on various factors including past performance on RANs financings. The state treasurer established the following minimum qualifications that the underwriters must meet to be considered for inclusion in the pool:  Minimum net capital of $25,000;  Proper licensure by the Securities Exchange Commission, the National Association of Securities Dealers, and the California Department of Corporations, if applicable; and  A commitment to furthering the STO's Target Business Enterprise goals for participation of minority, women, and disabled veteran-owned firms in underwriting services. The state treasurer issued the RFQ in December 1991 to 166 underwriting firms that were invited to apply for inclusion on any or all of the pools. Eighty-six of the 166 firms applied for inclusion in at least one of the four pools. Most of the firms applied for all four pools. Seventy-six firms applied for the SNO pool, 75 applied for the CEFA pool, 73 applied for the CPCFA pool, and 81 applied for the CHFFA pool. The state treasurer established an evaluation committee to rate the firms based on their responses in certain areas. The committee divided the responding firms into four groups: large firms, medium firms, small firms, and Target Business Enterprise firms. The size categories were based on the firms' net capital. The committee evaluated the responses for each pool by rating the firms "high," "medium," or "low." The committee applied the rankings within the four groups outlined above to ensure that firms from each group were included in each pool. According to the STO's summary of its underwriter selection process, the evaluation committee submitted its recommendations to an executive committee. The executive committee added its comments to the evaluation committee's recommendations and submitted the recommendations to the state treasurer for approval. The state treasurer made the final selection of firms for each of the four pools. As a result of this evaluation and review process, the state treasurer selected 47 firms for one or more of the underwriter pools: 40 firms for 11 the SNO pool, 19 firms for the CEFA pool, 21 firms for the CPCFA pool, and 23 firms for the CHFFA pool. The state treasurer established the pools in March 1992, and, because underwriters are selected several weeks before a negotiated bond sale, the pool became effective for bond sales occurring in or after April 1992. We reviewed the state treasurer's selection of underwriters for the 33 negotiated bond issues sold from April 1992 to December 1992 at the STO and financing authorities and found that the state treasurer selected all lead underwriters and co-managers from the applicable pool, except co-managers for the September 1992 RANs financing. As stated earlier in this chapter, the state treasurer selected the co-managers for the RANs based on various factors including past performance on RANs financings. Lead State-Negotiated and Public Works Board Bond Issues Underwriter We reviewed the state treasurer's selection of lead underwriters for the Selection Before 12 bond sales negotiated by the STO from January 1991 until April 1992, when the current SNO pool became effective, and found April 1992 that the state treasurer selected lead underwriters for state bond issues in several ways. The previous state treasurer established an underwriter pool in 1990 for selection of lead underwriters for State Public Works Board (PWB) bond issues. The current state treasurer used the pool to select lead underwriters for all four of the PWB bond issues sold from January 1991 through April 1992. The current state treasurer selected lead underwriters for three of the five other state bond issues sold from January 1991 through April 1992, excluding interim notes and RANs, through an RFQ process specific to the bond issue. For one of the five bond issues, the state treasurer selected one lead underwriter from the PWB pool. As discussed above, the PWB pool was established through a competitive process. For another of the five bond issues, the STO did not competitively select the lead underwriter because staff believed the relatively small sale would provide a good opportunity for a Target Business Enterprise firm to obtain the experience of being a lead underwriter. For the July 1991 interim notes and the August 1991 and February 1992 RANs, we found that the state treasurer selected the lead underwriters through an RFQ process specific to the July 1991 interim notes and August 1991 RANs. Although the state treasurer did not conduct a separate competitive process for the February 1992 RANs, the state treasurer selected the same lead underwriters who had successfully completed the recent competitive process for the 1991 RANs. 12 Financing Authorities The previous state treasurer established underwriter pools for CEFA and CHFFA in 1990 and for CPCFA in 1989. We reviewed the current state treasurer's selection of lead underwriters for the 38 bond issues sold at CHFFA and CPCFA from January 1991 through April 1992 and found that all lead underwriters were selected from the pertinent pool. We could not review the state treasurer's selection of lead underwriters for the seven negotiated bond sales at CEFA from January 1991 through April 1992 that were not private placements because CEFA no longer had in its files a list of underwriters on the pool that was in effect during this time. Co-Manager The purpose of Government Code, Section 5703, was to expand Participation in competition in the selection of underwriters for the State's bond sales. And, to an extent, this has occurred, as we have discussed on the Negotiated Bond previous pages of this chapter. However, because Government Code, Sales Section 5703, is vague in one of its key provisions, the objective of attaining competition in the selection of all underwriters for state bond sales may not be fully realized. The Government Code, Section 5703(a), makes the state treasurer "responsible for developing and implementing a competitive process for selection of underwriters for negotiated offerings of bonds." The vagueness of this law relates to whether the law was intended to require the state treasurer to select both the lead underwriter and the co-managing underwriters (co-managers) through a competitive selection process, or only the lead underwriter. The law simply refers to "implementing a competitive process for the selection of underwriters," but does not specify whether the competitive process should include just lead underwriters or both lead underwriters and co-managers. When we discussed the law with the STO's general counsel, he informed us that the intent of the law was to require the state treasurer to use a competitive process solely for selecting the lead underwriters, and thus, the STO's interpretation of the law is that it does not pertain to co-managers. In similar situations in the past, the Legislative Counsel has advised the Office of the Auditor General that the courts afford great weight in the interpretation of statutes to the agency charged with the administration of these statutes unless the interpretation is clearly erroneous or unauthorized. When the state treasurer selected only lead underwriters through a competitive process, a portion of the State's underwriting business was awarded without the benefit of competition. This is because the state treasurer, in certain instances, did not select co-managers through a 13 competitive process. This primarily occurred from January 1991 until April 1992, when a new policy was implemented. Under the new policy, the state treasurer uses its competitively-established pools of firms to select co-managers for all negotiated sales, other than short-term interim notes and revenue anticipation notes (RANs). Instead, the RANs, which accounted for 9 percent of the amount that co-managers earned during the two years we reviewed, are selected based on various factors including past performance on RANs financings. Co-managers account for a significant portion of the State's underwriting business. For example, the state treasurer selected the underwriters for 90 of the 113 negotiated state bond sales occurring during 1991 and 1992. The remaining 23 bond issues were sold by the CHFA or were private placements. According to the STO's 1991 and 1992 reports on underwriter assignments in state bond issues, the state treasurer appointed one or more co-managers for 74 (82 percent) of the 90 bond sales. See the appendix for a listing of bond sales for which the state treasurer selected underwriters during 1991 and 1992. The total amount of underwriter's discount that co-managers of the 74 bond issues earned for the two-year period was $21.4 million, 41.5 percent of the total amount the issuers expended for underwriter's discount. Conclusion According to the general counsel for the STO, the Government Code, Section 5703, requires the state treasurer to select only the lead underwriter of negotiated bond sales through a competitive process. From January 1991 through December 1992, of 83 negotiated bond sales that we were able to review, the state treasurer selected all the lead underwriters through a competitive process except for one negotiated bond sale. In April 1992, the state treasurer began implementing a new policy of selecting most of the co-managers, as well as lead underwriters, through a competitive process. From April 1992, when the current underwriter pools became effective, to December 1992, the state treasurer selected both lead underwriters and co-managers from the competitively-established underwriter pools, except co-managers for the September 1992 revenue anticipation notes financing. However, we feel that the question should be raised as to whether the State will benefit fully from the introduction of competition into the underwriter selection process, when the law is interpreted to require only the selection of lead underwriters through a competitive process, and not the selection of all underwriters, both lead and co-managers. When the state treasurer selected only lead underwriters though a 14 competitive process, a portion of the State's underwriting business was awarded without the benefit of competition. Additionally, because the state treasurer does not interpret the law to require competitive selection of co-managers, the current policy of selecting co-managers for the other negotiated sales could change. If in the future a state treasurer does not select co-managers through a competitive process, the State could award million of dollars of underwriter's discount annually without the benefit of competition. For the two years we reviewed, the total underwriter's discount earned by all co-managers the state treasurer selected was $21.4 million. Recommendation The Legislature should consider amending the Government Code, Section 5703, to specifically state that co-managers as well as lead underwriters be selected through a competitive process. 15 Blank page inserted for reproduction purposes only 16 Chapter 2 Maintenance of Required Records at the State Treasurer's Office, the California Housing Finance Agency, and Three State Financing Authorities Chapter Three of the five entities we reviewed did not collect all of the cost Summary information that the Government Code, Section 5703, requires to be maintained for negotiated bond sales. During 1991 and 1992, the California Educational Facilities Authority (CEFA) did not maintain all the cost information required by the Government Code, Section 5703(e), for bond issues sold by private placement. We reviewed 9 files at the CEFA and found that 3 of the files were missing records documenting from $11,000 to $22,000, or 18 to 28 percent, of the total underwriter's discount and costs of issuance. The State Treasurer's Office (STO) and California Pollution Control Financing Authority (CPCFA) maintained most of the required cost information, although 2 of 12 files reviewed at the STO were missing items of $57,000 to $60,000, or 1 to 2 percent of the costs, and 2 of 5 files reviewed at the CPCFA did not sufficiently identify certain costs. We reviewed 29 files at the California Health Facilities Financing Authority (CHFFA) and 8 files at the California Housing Finance Agency (CHFA) and found that both agencies maintained the required cost information in the files we reviewed. During 1991 and 1992, only the STO sold bonds by competitive bid. We reviewed 18 of 24 bond issues sold by competitive bid during 1991 and 1992 and found that in all 18 bond sales, the STO maintained records of all bids and bid verifications as required by the Government Code, Section 5703(f). In addition, the STO awarded the bonds in each sale we reviewed to the underwriter who submitted the bid with the lowest true interest cost. Cost Records for When state bonds are sold through negotiated offerings, Negotiated Sales Section 5703(e) of the Government Code requires the state treasurer to maintain records of all cost information pertinent to the initial bond offering, except for bonds issued by state financing authorities, in which case the issuing authority is required to maintain the pertinent cost information. The information must include, but is not limited to, the following: 17  All amounts paid out of bond proceeds to the underwriter, detailed by management fee, takedown, risk, and underwriter's expenses;  All costs paid out of bond proceeds to rating agencies for rating of the bonds;  All fees paid out of bond proceeds to bond counsel, trustees, or financial advisors relating to the initial offering of the bonds; and  The interest rate to be paid on the bonds. Fees paid to underwriters are referred to as underwriter's discount and the other costs related to issuing bonds are referred to as costs of issuance. The underwriter's discount is compensation paid to the underwriter, based on four components: management fee, takedown, risk, and expenses. It is called a discount because it is usually expressed as a deduction from the principal amount of the bonds. The management fee is compensation for underwriter activities such as developing financing alternatives, assisting in the preparation of legal documents, preparing the official statements, and managing the operations of the underwriter syndicate. Takedown is similar to a sales commission in that it is a discount from the bond's public offering price allowed to the underwriter who sells the bond. Risk is compensation for the risk an underwriter assumes that not all of the bonds will be sold. The expense component is reimbursement of the underwriter's out-of-pocket expenses related to issuing the bonds. During 1991 and 1992, the state treasurer acted as agent for sale for 113 negotiated bond offerings. Three state financing authorities issued 81 of these bond offerings: the California Educational Facilities Authority issued 20, the California Health Facilities Financing Authority issued 41, and the California Pollution Control Financing Authority issued 20. The California Housing Finance Agency (CHFA) issued 14 of the negotiated bond offerings. The financing authorities and CHFA are required to maintain the cost records for the bonds that they sell. Various state agencies sold the remaining 18 bond issues. For these bonds, the STO is required to maintain the cost records. To determine if the STO, CHFA, and the financing authorities maintained complete cost information, we compared the detailed summary of the cost information contained in the file with the total of the costs of issuance and underwriter's discount set forth in the official closing documents related to the bond issue. These documents include the trust indenture, official statement, bond purchase contract, tax certificate and agreement, receipt for purchase price, and Internal 18 Revenue Service Informational Tax Return (Form 8038). Because the financing authorities usually are not responsible for making disbursements for costs of issuance, they obtain the detailed summary for their files from the underwriter or the private organization issuing the bonds. For bonds issued by other state departments, the STO is not responsible for making cost of issuance disbursements and obtains the cost of issuance information from the issuing state department. Because the requirements of Section 5703(e) of the Government Code focus on costs paid from bond proceeds, our review also focused on costs paid from bond proceeds. California Educational Facilities Authority We selected for review 10 of the 20 bonds that the CEFA issued during 1991 and 1992 and found that in 9 of these 10 sales, the issuer used bond proceeds to pay costs of issuance, underwriter's discount, or both. For 3 of these 9 sales, the CEFA had incomplete cost information. The files were missing records documenting from $11,000 to $22,000, or 18 to 28 percent, of the total underwriter's discount and costs of issuance for the bond sale. However, all the files we reviewed had the required interest rate information. All three of the exceptions occurred in bonds that were privately placed with a single purchaser, rather than offered to the public. According to the CEFA's executive director, the authority has increased the use of private placements since it was first authorized in 1990 because private placements provide significant savings to the borrowers, including savings in costs of issuance. However, the CEFA's executive director further stated that, unlike bond issues sold through public offerings, the collection of cost of issuance information for private placements has been on a case-by-case basis. As a result of our review, the CEFA has implemented new administrative procedures designed to ensure that cost of issuance information is collected in a uniform manner for all bond sales. State Treasurer's Office We selected for review 13 of the 18 bond issues for which the STO was required to maintain the cost records and found that in 12 of the 13 bond issues, the issuer used bond proceeds to pay underwriter's discount, costs of issuance, or both. Ten of these 12 files contained all the required information: the detail of the underwriter's discount, the cost of issuance information, and the interest rates. Two of the 12 files contained incomplete information on a small portion of costs. Specifically, one file was missing records of approximately $57,000 (2 percent of the total costs), and another file was missing records of 19 approximately $60,000 (1 percent of the costs) paid from bond proceeds. For bond sales in which the issuing department is responsible for cost of issuance disbursements, the STO relies upon the issuing department to provide them with the required cost information. In both instances, the issuing departments had provided incomplete information to the STO. California Housing Finance Agency, California Health Facilities Financing Authority, and California Pollution Control Financing Authority The California Housing Finance Agency (CHFA) makes the cost of issuance disbursements for the state bonds it issues; therefore, it does not need to obtain cost of issuance reports from outside entities. It maintains the required cost of issuance information in management information reports and in summarized accounting records for each bond issue. We reviewed a sample of 8 of the 14 bond issues sold by the CHFA during 1991 and 1992 and found that the CHFA maintained the required records, including interest rate information, for all 8 bond issues. The California Health Facilities Financing Authority (CHFFA) and the California Pollution Control Financing Authority (CPCFA) do not make cost of issuance disbursements for most of the state bond issues that they sell. Therefore, they obtain the required cost information from the lead underwriters. We selected for review 31 of the 41 bonds issued by the CHFFA during 1991 and 1992 and found that in 29 of these bond issues, the issuer used bond proceeds to pay costs of issuance, underwriter's discount, or both. The CHFFA maintained all the required records, including the interest rate information, for all 29 of the bond sales. We selected for review 15 of the 20 bonds issued by the CPCFA during 1991 and 1992 and found that in 5 of these sales the issuer used bond proceeds to pay costs of issuance, underwriter's discount, or both. The CPCFA maintained all the required records for 3 of these 5 bond sales. For one sale, the records were insufficient to determine how much of the underwriter's discount had been allocated to expenses, as required by Section 5703(e)(1) of the Government Code. This problem occurred because the form on which the CPCFA collected the cost information from the underwriters did not require this information. During our review, the CPCFA corrected the deficiency in this file and changed their forms and procedures to prevent such a deficiency in the 20 future. In another file, the underwriter had erroneously reported to the CPCFA that all of a financial advisor's fee had been paid from sources other than bond proceeds. However, we determined that bond proceeds had actually been used to pay about one half of the advisor's fee. The State When a bond issue is sold by competitive bid, Section 5703(f) of the Treasurer's Office Government Code requires the state treasurer to maintain records of all bids submitted and documentation of bid verifications, including the Is Maintaining terms of sale and the calculation of net interest cost or true interest cost. Required Records During 1991 and 1992, only the STO sold bonds by competitive bid; for Competitively the state financing authorities and the CHFA sold bonds only through Bid Bond Issues negotiated offerings. All of the State's general obligation bonds were sold by competitive bid. Additionally, the STO sold bonds by competitive bid on behalf of the Department of Water Resources, the University of California, and the California State University. In 1992, the STO also sold by competitive bid revenue anticipation warrants, which are short-term obligations. We reviewed the STO's records for 18 of the 24 bond issues sold by competitive bid and found that the state treasurer maintained the required records for all 18 bond sales. The STO places the terms of sale for competitively bid bonds in a document called the Notice of Sale, and all files we reviewed contained the Notice of Sale. All 18 files also contained documentation that the STO had verified each bidder's compliance with the terms of sale and calculated each bid's true interest cost. In all 18 sales, the STO awarded the bonds to the underwriter who submitted the bid with the lowest true interest cost. Conclusion During 1991 and 1992, the California Educational Facilities Authority did not always maintain complete cost information for negotiated bond issues sold by private placement, as required by Section 5703(e) of the Government Code. The State Treasurer's Office and California Pollution Control Financing Authority maintained most of the required cost information in the files we reviewed, although 2 of 12 files reviewed at the STO were missing items of $57,000 to $60,000, or 1 to 2 percent of total costs, and 2 of 5 files reviewed at the CPCFA did not sufficiently identify certain costs. The California Health Facilities Financing Authority and California Housing Finance Agency maintained all required cost information in the files we reviewed. In addition, the STO maintained records of all bids and bid verifications for those bonds sold by competitive bid that we reviewed, as required by Section 5703(f) of the Government Code. 21 Recommendation To ensure that complete cost information for negotiated bond offerings is obtained for the records, the CEFA should ensure that the newly implemented administrative procedures for obtaining the required information for private placements are working. 22 Chapter 3 Comparison of California Bond Issues With Other Government Bond Issues Chapter In addition to reporting on the state treasurer's implementation of the Summary competitive process for selecting underwriters, Section 5703(h) of the Government Code requires us to compare the costs and interest rates of California's initial bond offerings with bond offerings of other states, where possible. For our comparison, we focused on true interest cost because true interest cost best reflects the overall cost to the State of issuing bonds. We compared the true interest costs of 10 bond issues sold by other governmental entities with the true interest costs of nine California bond issues. We found no indication from the information obtained in this limited review that California is paying more than necessary for true interest cost on state bonds. Seven of the nine California bond issues had lower true interest costs than the comparison bond issues. In the other two bond offerings, the true interest costs differed by .0184 and .226 percentage points. The differences in the true interest costs may be attributable to the different features of the bonds or to differing market conditions at the time of the bond sales. Selection of True The Government Code, Section 5703(h), requires us to compare the Interest Cost costs and interest rates of California's initial bond offerings with bond offerings of other states, where possible. For our comparison, we As Basis for focused on true interest cost because true interest cost best reflects the Comparison overall cost to the State of issuing bonds. Most government bond issues consist of a series of bonds that mature in different years, usually in from one to 20 years, and each maturity of bonds may pay a different interest rate to the investor. In addition, government bond issues may have other features, such as capital appreciation bonds or term bonds; therefore, the structure of government bond issues can be complex. Evaluating complex bond issues requires the calculation of the total interest cost of the bond issue. This calculation is used to determine which structure results in the lowest interest cost, and it serves as the basis for awarding bonds in competitive bond sales. There are two methods in general use for calculating the total interest cost of complex government bond issues: the net interest cost method and the true interest cost method. 23 The most important difference between the two methods is that true interest cost takes into consideration the time value of money while net interest cost does not. We chose true interest cost as our basis for comparison of the bonds because it more closely reflects the value the investment community places upon the future cash flows of a bond issue. Because it is calculated on the net proceeds for the bonds, less the underwriter's discount, it also reflects the effect of underwriter's discount on the interest cost of the bonds. The total amount paid in underwriter's discount is balanced against the results of the underwriter's efforts; therefore, if one underwriter is able to structure and sell the bonds in a way that results in a lower interest cost than another underwriter, it may be worth a larger expenditure for underwriter's discount. True interest cost is also the method used by the state treasurer in evaluating the bids for bonds sold on a competitive basis. Selection of We limited our comparison to bonds that were publicly offered and Bonds and sold through negotiation rather than by competitive bid. We did not include bonds sold by competitive bid because the issuer has little Related control over the interest cost for these bonds. They must be awarded Information for to the underwriter submitting the bid that results in the lowest interest Comparison cost. We also did not include bonds that were privately placed, that is, sold to a single investor. Issuers do not always obtain a credit rating for privately placed bonds, so we could not compare the relative risk of these bonds, which has a significant influence on the interest rates. Because the state treasurer sold all of California's general obligation bonds during our review period by competitive bid, only revenue bonds are presented in the comparison. We selected a sample of 26 California revenue bond issues and two revenue anticipation notes sold through negotiation and obtained the necessary information about these issues from the bond files at the STO or at the issuing state financing authority. Then we obtained information about similar bond issues from Securities Data Corporation. Securities Data Corporation provides information about government bond issues to the financial community, maintaining a data base of information about government bond issues. We did not audit the data provided by Securities Data Corporation. When necessary, we supplemented the data from Securities Data Corporation with data obtained from the issuers or underwriters of the selected bond issues. We defined similar bond issues based on the bond issue's credit rating, sale date, principal amount, the bond-financed project, and security for the bonds. The credit rating is a measure of the risk involved with investing in the bonds and is the only attribute that we determined must 24 be exactly the same in comparing bond interest rates. The sale date is very important because market conditions can change rapidly; therefore, we limited our comparison to bond issues that were sold within three days of each other. Three of the ten bond issues selected for comparison with California bonds were sold on the same day. If the bond issuer purchased bond insurance, which insures the payment of principal and interest in the event of issuer default, this would be reflected in the rating because the rating for insured bonds is based on the credit of the bond insurer instead of the issuer. The purpose of insuring bonds is to obtain a better credit rating which results in a lower interest rate. In some instances, bond insurance is purchased only for a portion of the bonds in the issue. In these cases, the insured bonds are rated separately from the uninsured bonds. For those bond issues that were partially insured, we present the percentage of principal insured and the rating information for both the insured and uninsured bonds. We also present information about each bond issue's average life and optional redemption feature, which affect a bond issue's interest cost. Average life is the average number of years a bond in the issue is outstanding and is a measure of how rapidly the principal is to be repaid. This measure in particular would affect true interest cost because the timing of principal payments is part of the calculation of true interest cost. An optional redemption feature allows a bond issuer to "call" a bond for redemption before the bond's maturity date. It is also referred to as the call option. A typical call option allows an issuer to begin calling bonds ten years after the bond issue date at a redemption price of 102 percent of the bond's principal value. The presence of a call option represents additional investment risks, such as reinvestment risk, which is the risk that the investor will not be able to reinvest his or her principal at the same interest rate as was paid by the bonds. The search of Securities Data Corporation's data base resulted in the identification of ten bond issues with features similar to nine California bond issues. We could not identify similar issues for the other 19 issues in our sample. Of the nine California bond issues, one each was issued by the Department of Water Resources, the California Public Works Board, the Capitol Area Development Authority, the California Housing Finance Agency, and the California Educational Facilities Authority, and two each were issued by the California Health Facilities Financing Authority and the California Pollution Control Financing Authority. 25 Table 1 below presents the information about each California bond issue, the comparison bond issue, and the Bond Buyer Revenue Bond Index for the date closest to the bond issues' sale dates. The Revenue Bond Index is a measure of market conditions for government revenue bonds on the date of its compilation. The index provides additional information about market conditions during the week the bonds were sold. We obtained this information from the Bond Buyer, which is a news publication for the government bond industry. 26 27 As shown in the table above, the true interest cost of seven of the nine California bond issues was lower than the true interest cost reported for the comparison bond issue(s). In the other two bond issues, the true interest costs were greater than the true interest costs of the comparison bond issue by .0184 and .226 percentage points. These differences may be attributable to differences between the features of the California bonds and the comparison bonds or to differing market conditions during the sale of the bonds. The data presented here is meant to be informative and to satisfy the request for a comparison of bond costs and interest rates. The attributes used to define and select similar bond issues are not intended to be a comprehensive list of all the attributes that affect the interest cost of a bond issue. Conclusion We found no indication from the information obtained in this limited review that California is paying more than necessary for interest costs on state bonds. Seven of the nine California bond issues had a lower true interest cost than the true interest cost reported for the comparison bond issue(s). In the other two bond issues, the true interest costs were greater than the true interest costs of the comparison bond by .0184 and .226 percentage points. These differences may be attributable to differences between the features of the California bonds and the comparison bonds or to differing market conditions during the sale of the bonds. We conducted this review under the authority vested in the state auditor by Section 8543 et seq. of the California Government Code and according to generally accepted governmental auditing standards. We limited our review to those areas specified in the audit scope of this report. Respectfully submitted, KURT R. SJOBERG State Auditor Date: January 21, 1994 Staff: Steven M. Hendrickson, Audit Principal Karen L. McKenna, CPA Star Castro 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Left Intentionally Blank 54 Blank Page For Response 55