& At H3 Executive Office for Administration and Finance March 31. 2000 The Benefits of Pooled Financing for School Building nice GOVtRHWW^^ Borrowing UMASS/AMHERST ArgeoPaulCellucci ANDREWS. NATSIOS governor 2tf» SECRETARY j 7 JaneSwift . lieutenantgovernor 31EDbt, 0271 ESSfl M mnm The School Building Assistance Program Restructuring LegisfeitK byiha.CglJ$ci/ Swift Administration provides for a pooled financing program for school bdrrVswnig. 'Given the fact the Commonwealth contributes on average 69% ofthe debt service costs to communities, a pooled financing borrowing structure is the most efficient way ofaccessing the capital markets, reducing debt service costs for both the Commonwealth and the communities. This is why many other state-supported local infrastructure assistance programs, such as those in Illinois and Michigan, have adopted a pooled financing structure. A pooled structure will produce savings in many ways. Lower Costs of Issuance Many ofthe costs of issuing bonds, such as printing costs, bond counsel, and rating agency fees, do not vary significantly with the size of the issue. Most communities also must purchase bond A insurance to market their bonds. pooled issuance would need no bond insurance and could spread fixed costs over a much larger principal amount, reducing the cost to each community. To determine the savings attributable to lower costs of issuance, the Executive Office for Administration and Finance reviewed projected costs of issuance for the current School Building Assistance wait list with the costs of issuance oftwo pooled deals. The analysis assumed that localities would have costs ofissuance of 1% oftheir total issue or $200,000, whichever was less. These costs did not include insurance and averaged 0.83% for the entire waiting list or $6.4 million.1 The estimated costs of issuance of the pooled program included paying for local bond counsel for local borrowers and ratings by three rating agencies. The analysis assumed two transactions ofroughly $400 million, enough debt to finance the entire wait list ofprojects approved for fiscal year 2000. The total costs ofissuance for both transactions were projected to be $5.3 million or 0.69%. Therefore, a pooled program would save about $1.1 million in costs of issuance each year. As noted, these savings do not include the costs ofbond insurance, which can add more than 0.4% to the costs of issuance. Insurance does lower the interest expense communities would need to pay, so we did not include it as part ofthe costs ofissuance analysis. However, it should be noted that in many instances where cities and towns receive rates competitive with larger and higher rated issuers like the Commonwealth, these communities have often had to pay for the lower interest rates with significant insurance premiums. Lower Debt Service Costs Since the Commonwealth pays on average 69% of the financing costs for debt issued pursuant to the School Building Assistance Program, this debt should be rated at or close to the Commonwealth's rating. This would be particularly true under the Cellucci/ Swift legislation since it provides that the Commonwealth would be the paying agent for almost all debt service, either through the School Building assistance appropriation or a local aid offset. Unfortunately, Fleet Securities, Incorporated reviewed 12 transactions that occurred between November, 1999 and March, 2000 and determined that these transactions had costs ofissuance averaging 0.81%, not including insurance. Executive Office for Administration and Finance March 31. 2000 under the current system, borrowers do not get the benefit of the Commonwealth's Aa2/AA-/AA- rating because they must borrow on their own. For the vast majority ofthe Commonwealth communities, this means higher borrowing costs for the communities and higher subsidy costs for the Commonwealth. Below is a table indicating the different rates issuers with differing credit ratings might be expected to receive: 1 year rate Basi:5 Point Savings Aaa 5.275% -0.10% Aa1 5.375% 0.00% Aa2 5.375% 0.00% Commonwealth's Credit Aa3 5.375% 0.00% A1 5.450% 0.08% A2 5.450% 0.08% Median Mass. Community's Credit A3 5.500% 0.13% Baal 5.625% 0.25% Baa2 5.700% 0.33% Ba2 6.000% 0.63% Source: Lehman Brothers, Rates as 2/29/00 Two hundred ninety-eight communities, or 85% of those in Massachusetts, are rated at the single A level or worse. Allowing these communities to borrow at the Commonwealth's rate could save at least 8 basis points and up to 60 basis points due to improved credit ratings alone. This savings exists even though credit spreads (the interest rate benefit received by more highly rated debt) are at historically low levels. Credit Rating Interest Rate Spreads (A-Aa) 1948-1997 Spread to Aa BPS 1.15- A-Aa 0.95- 0.75- 0.55 0.35 0.15 -a05 1948 1953 1958 1963 1968 1973 1978 1983 1988 1993 Source: Moody's Municpal &Government Manual, 1998, vol. 3 Ifcredit spreads returned to the levels ofthe early and mid-1980s, the benefit ofthe improved credit rating would be three to four times its current value. The analysis ofthe 2000 waiting list Executive Office for Administration and Finance March 31. 2000 indicates that savings fromjust improving local credit ratings to that ofthe Commonwealth would be another $450,000 per year or $9 million over the life ofthe bonds, even with today's narrow credit spreads. This analysis includes communities with higher ratings than the Commonwealth and those that benefit from the preferential treatment accorded small or "bank-qualified" issuers.2 Local issues do benefit from a strong market of local buyers, who receive both a federal and a state tax-exemption for municipal debt. The retail market in which local issues are usually sold is also less price-sensitive than the larger institutional market. Yet a pooled issuance has benefits that might offset these advantages. By utilizing a statewide issuer, creating a diversified pool of Massachusetts credits, and augmenting the state's School Building Assistance appropriation with a local aid intercept for the balance ofthe debt service, a pooled issue would present a very strong package to the capital markets that could achieve interest rates even lower than those usually received by Aa2 credits. This is because large investors in Massachusetts debt, such as mutual funds and insurance companies, have limits on the amount ofdebt they can hold in their portfolio from any one issuer. The pooled program would be a new, recognizable investment vehicle that A these large investors could purchase outside those limits, lowering the interest rate. pooled program would also create what investors call "liquidity," or the ability to buy and sell the bonds regularly on the secondary market. Liquidity is essential to large institutional investors, and few local bonds have such liquidity. The lack ofliquidity for a debt issue often results in a "liquidity premium." The cost ofthe liquidity premium may compensate the pooled program for the benefits of selling to less price sensitive retail buyers. The benefits of having a higher credit rating and more liquidity produce savings today, but become much more valuable during periods ofrising interest rates and falling bond prices. In such an environment, smaller issues will have difficulty selling at all. As noted, most small issues require bond insurance. This is because the retail market, while being somewhat price insensitive, is very wary ofcredit risks. In difficult markets, such insurance will be more A expensive or not available at all. larger, more established issue will still be able to access the markets at a reasonable interest rate in almost any market conditions. Ability to Take Advantage of Refunding Opportunities and Sophisticated Financing Techniques Another benefit ofthe pooled program is its ability to take advantage ofrefunding opportunities and innovative financing techniques. A refunding occurs when interest rates drop and a borrower issues new debt at a lower rate to retire the higher rate old debt, thereby generating savings. Refunding debt involves transaction costs, so it only makes sense when the savings will be in excess ofthose transactions costs. As in an initial sale, many of those transaction costs are fixed, so a larger issue increases the likelihood that the issuer can take advantage of interest rate declines to refund debt. Moreover, smaller issues tend to have higher call premiums and more restrictive call provisions than larger transactions. The typical savings threshold for performing a refunding is 4%. Assuming there will be one opportunity to refund debt issued by the pool over the life of the bonds, and assuming about halfofeach issue will be outstanding when the opportunity occurs, the ability to refund debt efficiently can be worth as much as $16 million. Prorating this across the life ofeach issue, these potential savings would equate to about $800,000 per year. A large pooled issue could also take advantage ofcommonly used sophisticated financing techniques like floating-to-fixed swaps, forward contracts for construction funds, and Qualified 2 Bank qualified issuers can only issue $10,000,000 in debt annually. Executive Office for Administration and Finance March 31. 2000 Zone Academy Bonds or QZABs. The cost of structuring these deals often includes a large fixed component as well, so once again, significant issuance size is required to make these techniques worthwhile. Under different market conditions, these techniques can render different savings figures. For illustrative purposes, we can assume one issue each year utilizes a fixed-to- floating swap, where the pooled program would issue floating rate debt and then contract with a financial institution or "counter-party" for that entity to pay the variable interest payments. In return, the pooled program will pay a fixed interest payment to the counter-party. These transactions often save 20 basis points, or an additional $500,000 annually if applied to only half ofthe pooled issuances. More Efficient Short-term Borrowing Currently, cities and towns financing construction with short-term financing must issue notes with maturities ofapproximately one year. This is because they do not have access to commercial paper financing vehicles that can access lower rates at the shorter- end ofthe yield curve. Commercial paper programs provide short-term financing in the 1-day to 270-day range, which as the graph below indicates, costs less in interest than the one-year note rate. Yield Curve - 1 Day to lYear % Yield 3.50 - & A ^ J? <* .S^ V $ *• V* <£ Source:BearSteams,RatesasofMarch23,2000 Yet commercial paper programs require large volume and a sure source ofpermanent financing. With a pooled financing program for the permanent School Building Assistance Program debt, the financing entity could easily set up a commercial paper program to provide construction financing prior to the issuance of long-term bonds. This will help communities with any timing 3 Swaps allow issuers to access the lower variable rate credit markets without giving up the certainty of fixed rate payments. Forward contracts allow for the buying and selling offunds for delivery at a point in the future. Qualified Zone Academy Bonds are new tax credit vehicles created by the Treasury Department 0% to fund school building. Although 27 Massachusetts school districts are eligible for financing through the issuance ofQZABs, none have taken advantage ofthe program to date. Because ofthe complexity of the tax credit vehicle, issuance ofQZABs has mostly occurred where states have had significant involvement. Executive Office for Administration and Finance March 31. 2000 issues the creation ofthe pooled program might create for them, as well as provide up to 50 basis points of savings for each year that they are borrowing short-term. Assuming halfofthe projects access this commercial program for an average ofa year, these savings would equal $2 million each year. Since the legislation filed by the Administration proposes to no longer reimburse for short-term borrowing costs, all the savings ofthe commercial paper program would accrue to the localities. Savings Summary Based on two transactions of $400 million each, a pooled financing program could provide savings as follows: Source of Savings Annual Savings Savings Over the Life of the Bonds Costs of Issuance/ Insurance 1,000,000 1,000,000 Credit Improvement/Insurance Cost 450,000 9,000,000 Avoidance Retail Market Premium -500.000 -10,000,000 Liquidity Premium 500,000 10,000,000 Refunding Opportunities 800,000 16,000,000 Sophisticated Financing Techniques 500,000 10,000,000 Commercial Paper Access 2,000,000 2,000,000 Total 4,750,000 38,000,000 Putting aside savings from refundings and innovative financing techniques, which may have varying effects on different communities, savings for selected projects on the existing wait list could be as follows: Source ofSavings District Local Estimated Commercial Credit Costs of Total Total Bond Issuance Size Paper Access Improvement Issuance Savings Savings For Rating Yearl Life ofthe Bonds Athol-Royalston New A3 12,742,036 63,710 11,031 47,692 122,433 332,012 Middle Boston Latin Aa3 29,882,050 149,410 - 13,024 162,434 162,434 Brockton Belmont St. Baa2 15,072,547 75,363 34,059 56,414 165,836 812,952 Elementary Fall RiverNorth End Baa2 16,378,143 81,891 37,009 61,301 180,201 883,371 Elementary Framingham Cameron Aa3 14,430,308 72,152 - 54,011 126,162 126,162 Middle Greenfield Middle A3 11,791,939 58,960 10,208 44,136 113,304 307,256 Hanson Middle School A3 14,163,015 70,815 12,261 53,010 136,086 369,037 Holliston MillerElementary Al 30,935,963 154,680 16,052 (37,360) 133,372 438,368 Melrose Lincoln A2 10,332,002 51,660 5,361 24,046 81,068 182,930 Elementary PioneerValley High School A 14,969,409 74,847 7,768 56,028 138,643 286,225 Reading Coolidge Middle Aa3 9,419,673 47,098 - 21,923 69,021 69,021 Salem Bates Elementary Al 9,561,224 47,806 4,961 35,786 88,554 182,817 Westford Academy Al 30,155,252 150,776 15,647 (31,370) 135,053 432,352 Worcester Forest Grove Al 18,125,496 90,627 9,405 67,841 167,873 346,568 Middle Executive Office for Administration and Finance March 31. 2000 Administrative Benefits A pooled program would also provide several administrative benefits. By centralizing the issuance of School Building assistance debt, a pooled program would give the proposed School Facilities Commission a clearer sense of the amount available to support debt service grants. If the pool is able to achieve a lower than expected true interest cost in one issue, the Commission could add additional projects to the next issue without exceeding the SBA appropriation. The need to audit the communities post-issuance would also be eliminated by the new program. The pooled program could relieve communities ofthe administrative difficulties offinancing their own transactions. This would include providing on-going disclosure to the financial A community, calculating arbitrage savings rebates, and investor relations. pool would also provide communities with more flexibility in structuring their financing than they currently have under state finance law. The fact that a pooled issuance would sell in the institutional markets as well as the retail market would provide further structuring flexibility, since the retail market tends to prefer a standard product with a relatively short maturity. The longer-term bonds do not sell well in the retail market, often forcing municipalities to repay their principal relatively quickly. Conclusion A pooled financing program has many benefits. It would create a recognizable, attractive financing vehicle for investors, resulting in debt service savings for the Commonwealth and cities and towns. It would allow for a streamlined financing process, lowering costs of issuance and administrative costs. It would provide access to innovative market opportunities, including new markets of investors for school debt and new financing techniques. Finally, it would allow the Commonwealth to more effectively manage the issuance of School Building Assistance bonds and gives communities many more options in structuring their debt.