US Local Government General Obligation Methodology and Scorecard User Guide

Transcription

1 JULY 24, 2014 U.S. PUBLIC FINANCE SPECIAL COMMENT US Local Government General Obligation Methodology and Scorecard User Guide Table of Contents: UNDERSTANDING THE SCORECARD 1 HOW THE SCORECARD WORKS ABOVE THE LINE 3 FACTOR 1: TAX BASE/ECONOMY (30%) 4 FACTOR 2: FINANCES (30%) 6 SUBFACTOR 3: MANAGEMENT (20%) 11 FACTOR 4: DEBT/PENSIONS (20%) 13 BELOW-THE-LINE ADJUSTMENTS: CASE STUDY OF RADNOR SD 16 MOODY S RELATED RESEARCH 20 APPENDIX A 21 Analyst Contacts: NEW YORK Dan Seymour, CFA Analyst dan.seymour@moodys.com Jack Dorer Managing Director - Public Finance jack.dorer@moodys.com Julie Beglin Vice President - Senior Analyst julie.beglin@moodys.com SAN FRANCISCO Matthew A. Jones Senior Vice President matthew.jones@moodys.com In January 2014, we published our US Local Government General Obligation Debt Methodology, introducing a general obligation (GO) scorecard into our rating process. We designed the scorecard to provide greater visibility into how we evaluate various factors in our GO rating process. In this follow-up report, we describe the scorecard inputs in more detail to enable investors, issuers and other market participants to understand our scorecard and the role it plays in our analysis. Because the US local government sector is so diverse and fragmented, seemingly straightforward factors like tax base size or debt levels could be calculated and interpreted differently by different users of the scorecard. To facilitate the discussion of our scorecard, we use the Radnor Township School District, PA (Radnor SD) as a case study. We upgraded the district s rating to Aa1 from Aa2 with a positive outlook on May 28. Understanding the scorecard The scorecard is a starting baseline for the consideration of a rating. It is neither a rating calculator nor a comprehensive listing of all factors affecting the rating. The scorecard features two parts: an above-the-line score based on standardized weights, and below-the-line adjustments for any factors not sufficiently captured above the line for that particular issuer. Radnor SD s above-the-line scorecard (see Exhibit 1) expressed some of the reasons for our upgrade, but not all. We incorporated additional considerations into our below-the-line adjustments that were specific to this issuer and that we believe were important considerations. While the scorecard does reflect the main elements key to our analysis of a credit, our ratings are not bound by either the above-the-line score or the final score after below-the-line adjustments. The scorecard is a tool that helps direct our analysis and discussion, and reinforces our goal of rating consistency across widely disparate issuers. As always, the majority vote of a committee of experienced municipal credit analysts determines our ratings. That determination is made only after an in-depth discussion of the credit. The scorecard is reviewed during the rating committee but is not determinant of the outcome. Additional components include detailed comparisons with similar credits nationally, explicit consideration of how the scorecard might not fully reflect important aspects of a credit, and a discussion of the alternate paths the credit will likely take over the next few years. This publication does not announce a credit rating action. For any credit ratings referenced in this publication, please see the ratings tab on the issuer/entity page on for the most updated credit rating action information and rating history.

2 EXHBIT 1 Radnor SD s Above-the-Line Scorecard School District Very Strong (Aaa) Strong (Aa) Moderate (A) Weak (Baa) Poor (Ba) Very Poor (B & Below) Input Weight Score Weighted Score Numerical score ECONOMY/TAX BASE (30%) $60M $4.25B 10% Tax Base Size: Full Value (in 000s) Full Value Per Capita Socioeconomic Indices: MFI FINANCES (30%) Fund Balance as % of Revenues 5-Year Dollar Change in Fund Balance as % of Revenues Cash Balance as % of Revenues 5-Year Dollar Change in Cash Balance as % of Revenues MANAGEMENT (20%) Institutional Framework Operating History: 5-Year Average of Operating Revenues / Operating Expenditures DEBT/PENSIONS (20%) Net Direct Debt / Full Value Net Direct Debt / Operating Revenues 3-Year Average of Moody's Adjusted Net Pension Liability / Full Value 3-Year Average of Moody's Adjusted Net Pension Liability / Operating Revenues Source: Moody s Investors Service > $12B $12B n > $1.4B > $150,000 $150,000 n > $65,000 > 150% of US median > 30.0% > 25.0% for School Districts 150% to 90% of US median 30.0% n > 15.0% 25.0% n > 10.0% for SD > 25.0% 25.0% n > 10.0% > 25.0% > 10.0% for School Districts 25.0% n > 10.0% 10.0% n > 5.0% for SD > 25.0% 25.0% n > 10.0% Very strong legal ability to match resources with spending Strong legal ability to match resources with spending > 1.05x 1.05x n > 1.02x < 0.75% 0.75% n < 1.75% < 0.33x 0.33x n < 0.67x < 0.90% 0.90% n < 2.10% < 0.40x 0.40x n < 0.80x $1.4B n > $240M $65,000 n > $35,000 90% to 75% of US median 15.0% n > 5.0% 10.0% n > 2.5% for SD 10.0% n > 0.0% 10.0% n > 5.0% 5.0% n > 2.5% for SD 10.0% n > 0.0% Moderate legal ability to match resources with spending 1.02x n > 0.98x 1.75% n < 4.00% 0.67x n < 3.00x 2.10% n < 4.80% 0.80x n < 3.60x $240M n > $120M $35,000 n > $20,000 75% to 50% of US median 5.0% n > 0.0% 2.5% n > 0.0% for SD 0.0% n > % 5.0% n > 0.0% 2.5% n > 0.0% for SD 0.0% n > % Limited legal ability to match resources with spending 0.98x n > 0.95x $120M n > $60M $20,000 n > $10,000 50% to 40% of US median 0.0% n > - 2.5% 0.0% n > - 2.5% for SD -10.0% n > -18.0% 0.0% n > - 2.5% 0.0% n > - 2.5% for SD -10.0% n > -18.0% Poor legal ability to match resources with spending 0.95x n > 0.92x 4.00% n < 10.00% n < 10.00% 15.00% 3.00x n < 5.00x 5.00x n < 7.00x 4.80% n < 12.00% n < 12.00% 18.00% 3.60x n < 6.00x 6.00x n < 8.40x $10,000 $134,887 10% % of US median -2.5% -2.5% for SD 233% 10% % 10% % 13.3% 5% % -2.5% for SD 32.1% 10% % 15.5% 5% Very poor or no legal ability to match resources with spending A 10% x % > 15.00% 2.34% 5% > 7.00x % > 18.00% 1.8% 5% > 8.40x % Above-the-line score: JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

3 How the scorecard works above the line Each of the above-the-line factors is assigned a weight and a value. The factor weights are the same for all GO credits. The values are issuer-specific. The values entered for each Subfactor determine the Subfactor score. The value corresponds to a number in the numerical score line. The numerical scores are valued on a linear sliding scale, meaning a value that is at the midpoint within a rating band will get a numerical score at the midpoint of the range associated with that band. For example, the Aa band for full value scores from 2.49 (weakest point of the range) to 1.5 (strongest point) for full values ranging from $1.4 billion to $12 billion. A full value at the midpoint of this full value range will get a numerical score at the midpoint of the 1.5 to 2.49 range. Radnor SD's full value input of $4.25 billion yields a score of 2.23 for that Subfactor given where it sits within the Aa band. The institutional framework is scored a 3. With the exception of the institutional framework, each Subfactor is scored on a continuum; the institutional framework Subfactor is a whole number. The strongest possible score in every category, except the institutional framework, is 0.5, and the weakest is 6.5 (see Appendix A for the specific metric inputs that would achieve maximum and minimum scores). For the institutional framework, the strongest possible score is 1 and the weakest possible score is 6. The composite weighted above-the-line score sums to an "above-the-line" score (see Exhibit 2), which is then adjusted based on below-the-line factors. EXHIBIT 2 Rating Score Rating Upper Level Lower Level Aaa Aa Aa Aa A A A Baa Baa Baa Ba Ba Ba B B B3 and below Source: Moody s Investors Service 3 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

4 Below, we walk through Radnor SD s scores in each of the above-the-line factors, as well as the belowthe-line considerations, to explain the role the scorecard played in our upgrade of the district. Factor 1: Tax base/economy (30%) Subfactor 1a: Full value (10%) What it is: The aggregate value of taxable property. What it shows: Size. Larger tax bases are generally more diverse, flexible and resilient against shocks than smaller ones. How we measure it: Our inputs for full value (FV) usually come from the issuer s most recent official statement, the required supplemental information in the back of the issuer s comprehensive audited financial statement, or directly from state or local tax assessors offices. We usually utilize the official assessed valuation (AV) of property for taxation purposes, often with some adjustment. When making adjustments, our analysts are aware of assessment practices in each state and utilize standardized approaches to ascertain full value. Obtaining the aggregate value of taxable property in a municipality can be imprecise. Different jurisdictions have different assessment practices, and the true market value of all property in a municipality is seldom observable each year because most properties do not sell each year. A few examples of assessment practices in select states:» Georgia: AV is assessed by counties, with AV generally set at 40% of full value.» Texas: Each of the state s 254 counties has a Central Appraisal District responsible for determining AV, which is set to be equal to FV.» New Mexico: Counties assess most property, with some properties such as utilities assessed by the state. AV is set at one-third of full value.» Maryland: Counties assess one-third of all properties each year, set to be equal to full value. The effect is for the entire tax base to be effectively revalued on a smoothed three-year cycle.» Ohio: Counties assess AV at 35% of full value. Counties conduct a full revaluation every six years.» California: Counties determine AV subject to the constraints of Proposition 13, under which a property s assessed value can only grow by the rate of inflation or 2% annually, whichever is less, except for when a property is sold, at which point the sale price becomes the new assessed valuation. In Pennsylvania, each of the commonwealth's 67 counties is responsible for deciding how often and when to reassess properties. As a result, some counties have reassessed very recently, and the AV is close to true market value. Others have gone many years without reassessing, and the AV is a fraction of the true market value. Every year, the State Tax Equalization Board estimates the relationship between each Pennsylvania county's AV and its true market value, known as a common level ratio (CLR). CLRs vary widely. Three counties have CLRs of more than 100%, meaning AV is higher than market value. Five counties AVs are precisely equal to full value. Most counties AVs are lower than full value, in many cases dramatically so. Delaware County, in which Radnor Township wholly lies, assesses to a base year of 1998, meaning any properties assessed since that year are assessed at what the estimated value would have been in 4 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

5 1998. The county's CLR is 74%. To approximate Radnor Township's full value, we divide the AV provided by Delaware County by the CLR provided by the commonwealth (see Exhibit 3). EXHIBIT 3 District s FV Derived From the CLR Year Assessed valuation Common Level Ratio Full value ,063,987, % $5,014,709, ,084,742, % $5,291,154, ,099,811, % $5,056,788, ,131,318, % $4,877,443, ,131,330, % $4,639,008, ,124,938, % $4,340,192, ,147,305, % $4,253,115,114 Source: Delaware County, Pennsylvania State Tax Equalization Board The increasing CLR over this time period implies that property values in Delaware County have been decreasing, which is consistent with the rest of the Philadelphia metropolitan region. Input: $4.25 billion Score: 2.23 Frequently asked questions Why use full value and not AV? Doesn t AV show the actual properties accessible for taxation? Full value offers two advantages over AV. First, because AV is not standardized, it would be inappropriate to compare an issuer s AV in Ohio (in which AV is 35% of full value) with an issuer s AV in Maryland (in which AV is 100% of full value). Second, full value offers a clear glimpse into the base s capacity to pay more property taxes. The true market value of a property aligns more closely with its ability to bear a greater tax burden. Subfactor 1b: Full value per capita What it is: Full value divided by population. What it shows: Per capita property wealth. The amount of taxable real estate per person indicates the relative capacity of the tax base to pay property taxes and fund the issuer s spending needs. How we measure it: Full value per capita is a simple function of the full value input into Subfactor 1a divided by population. The US Census Bureau s American Community Survey (ACS) provides population figures for most issuers. We may use a proxy population figure, derived from a county or some other entity, if the ACS does not report population for a rated entity. The 2012 ACS reported a population of 31,531 for Radnor Township. Full value ($4,253,115,114) / Population (31,531) = Full value per capita ($134,886.78) Input: $134,887 Score: JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

6 Subfactor 1c: Median family income (10%) What it is: Median annual income for the families living within a municipality, relative to the US median. What it shows: The relative wealth of the taxpayers, suggesting their capacity to pay more taxes, and potential needs for social service spending. How we measure it: The American Community Survey reports median family income (MFI) for most local governments in the US. As of 2012, the MFI for Radnor Township was $150,167 and the MFI for the US was $64,585. Hence, the input for MFI is 150,167/64,585, or 233%. Input: 233% Score: 0.5 Frequently asked questions Why use MFI and not per capita income? MFI is a better measure of households capacity to pay taxes and social service needs. The shortcoming of per capita income is that universities, prisons, and military bases depress the numbers. This skewing affects PCI more than MFI. Factor 2: Finances (30%) Subfactor 2a: Fund balance (10%) What it is: Operating fund balance as a percentage of operating revenues, adjusted in cases where the reported fund balance is incomplete or misleading. What it shows: The assets, net of liabilities, that will be available over the short term to satisfy obligations. How we measure it: The approach to measuring fund balance differs depending on the issuer and the way it reports its assets and liabilities. Most frequently, we start with the total general fund balance and subtract certain items, such as any receivables that we determine to be uncollectible. Sometimes, we may start with the unassigned general fund balance and add other reserve balances that we believe are usable. In either case, we would also look outside the general fund for any potentially accessible resources that should count as available fund balance. The goal is to identify any resources convertible into operating cash in the near term, generally within one year, net of all short-term operating outflows. We are likely to include some items even if they are classified as restricted, if we believe they are actually usable. The majority of Radnor SD s fund balance is assigned into the following year s budget. We decided to count the entire fund balance, both unassigned and assigned, because our analysis concluded that all of the assigned fund balances were actually operating reserves. How did we know the district wouldn t spend its assigned fund balances in 2014? Because a review of the district s budget-to-actual variances during the past five years reveals that every year the district has assigned more and more fund balance to be spent, and every year it hasn t spent any of it (see Exhibit 4). 6 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

7 EXHIBIT 4 Radnor SD s Assigned Fund Balance Is an Operating Reserve $25,000 Assigned fund balance Unassigned fund balance $20,000 $ 000s $15,000 $10,000 $5,000 $ Source: Radnor Township School District The district also maintains $4.6 million of positive balance in a capital projects fund outside the general fund. We did not include this money in the operating funds balance, although it does reflect favorably on the district. We are generally less likely to include reserves held outside of an operating fund, unless they are held in a fund specifically devoted to debt service. Fund balance ($19,241,795) / Revenues ($82,151,394) = 23.4% Input: 23.4% Score: 1.61 Frequently asked questions What does operating funds mean? Operating funds are pools of money used to fund core governmental services. The analysis of whether a fund is operating usually focuses on whether its revenues are usable for general governmental operations. The general fund is always an operating fund. A debt service fund is usually an operating fund, if the revenues devoted to paying debt service flow through that fund. Certain special revenue funds dedicated to funding core governmental services, such as public safety, would also count as operating funds. A non-operating fund is a pool of money devoted to something other than what we consider core operations, often with funds restricted to a certain non-core use. Utility funds are usually nonoperating, as are capital, community development, housing, and golf course or swimming pool funds. What does operating revenues mean? We use the term operations to refer to the core, recurring resources and obligations of the local government. Given the heterogeneity of the US municipal finance universe, we reserve the flexibility to determine what counts as operating and non-operating revenues credit by credit. Examples of operating revenues are taxes whose use is unrestricted, recurring fees, and transfers from higher levels of government. Examples of non-operating revenues are taxes restricted to a narrow use, one-time asset sales, bond proceeds, cash flow borrowing, and any other money restricted for a non-operating use. 7 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

8 Non-operating revenues do not reflect the true trend of revenues and can result in misleading swings in fund balance as a percentage of revenues. Why do you have different fund balance standards for school districts compared with cities and counties? School districts tend to have lower fund balances than cities and counties, and we believe this is with good reason, even at the same rating level. School districts usually have more stable and predictable revenue sources and expenditure requirements than cities and counties do. The variability of school districts revenues and expenditures is lower, and therefore does not suggest the need for as much cushion against contingencies. Would you include capital reserves in fund balance? Not as a general rule. It depends how easily the reserves could be accessed for operating purposes. Bond proceeds restricted for capital would not be counted. A general fund transfer to a capital fund that could easily be returned to the general fund might be counted, but we are unlikely to do so unless we think such a transfer will realistically take place. We often depend on the issuer s statements, either in its Comprehensive Annual Financial Report s management and discussion section or verbally, to ascertain this. What types of dubious items would you strip out of fund balance? Anything that is likely to be non-spendable, such as receivables that are likely to be collected over a period of more than one year, or not collected at all. This is true of interfund receivables as well as taxes receivable or receivables from other governments that are unlikely to be collected in one year, and particularly for receivables that have been reported on the issuer s balance sheet for two years or more with no decrease. Subfactor 2b: Five-year dollar change in fund balance as % of revenues (5%) What it is: The change in fund balance over the past five years. What it shows: The issuer s financial results over a typical, full economic cycle. How we measure it: The formula for this input is as follows: (Most recent year s fund balance) (Fund balance six years ago) Most recent year s operating revenues For Radnor SD, the 2013 fund balance was $10.9 million higher than it was in 2008 (see Exhibit 5), representing 13.3% of revenues. 8 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

9 EXHIBIT 5 Radnor SD s 5-Year Change in Fund Balance 25% 20% Fund balance % of 2013 revenues 23.4% $25,000 $20,000 15% 10% 10.1% $15,000 $10,000 $ 000s 5% $5,000 0% $ Sources: Radnor Township School District, Moody s Investors Service Input: 13.3% Score: 2.28 Frequently asked questions Why use a five-year cycle? The five-year period is meant to represent one full economic cycle, so that the scorecard does not incorporate the volatility of ordinary macroeconomic business expansions and contractions. Of course, some economic cycles last more than five years. The most recent downturn has put significant pressure on issuers in this category because of its length, but as the five-year cycle moves beyond the worst years of the recession, that pressure will alleviate. What do you do if you don t have six years of audits to calculate a five-year change? We may, in limited cases, rely on a three- or four-year trend, though it is not preferable. Why do you rely on historic trends rather than look forward? All of our scorecard inputs are based on the best available information, and all (with the exception of the institutional framework score) come from reliable third-party sources such as audited financial statements, actuarial reports and tax assessor reports. This rule helps us achieve transparency, standardization and comparability but, as a matter of course, relies on historical data. We incorporate forward-looking analysis in our notching factors and non-scorecard-based analysis. Subfactor 2c: Cash balance (10%) What it is: Net operating cash as a percentage of revenues. What it shows: Liquid resources available to meet obligations. How we measure it: Cash and liquid investments held in operating funds as a percentage of revenues, net of cash flow notes. Cash is the paramount liquid resource. A fund balance that consists mostly of cash provides greater financial flexibility than one that consists mostly of receivables, because cash is available to be spent immediately whereas receivables are available to be spent only once they are received. 9 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

10 By cash, we usually mean cash and liquid investments, such as certificates of deposit and US Treasury notes. Radnor SD reported $13.3 million of cash and an additional $13 million of investments as of the end of fiscal year We include the investments as cash because the commonwealth only authorizes school districts to invest in bank accounts, liquid asset funds, and other instruments that are very liquid. Input: 32.1% ($13,295,048 of cash and $13,052,254 of investments/ $82,151,394 of revenues) Score: 0.5 Why exclude cash flow notes from cash? The repayment of a cash flow note represents a cash outflow in the near term, and the note typically has a claim on the issuer s cash or revenues. Therefore, we do not regard that cash as usable for operating purposes. What if an issuer s fiscal year happens to occur at a point when a lot of cash flow notes are outstanding or an issuer s cash is at its low point? We rate issuers with a large variety of fiscal years. We recognize that cash balances are variable throughout the year, and an issuer s reported cash balance could be higher or lower depending on when the fiscal year ends. Our practice is to enter information in the scorecard as of the end of the fiscal year given that audited balance sheets report that point in time. We recognize in our analysis that financial statements can cast the same issuer in different lights based on the timing of the fiscal year end. Subfactor 2d: Five-year change in cash balance as a percentage of revenues (5%) What it is: The change in cash balance over the past five years. What it shows: The change in cash balance, net of cash flow notes, over a full economic cycle. How we measure it: The formula is the same as the change in fund balance, except it uses cash instead of fund balance. EXHIBIT 6 Radnor SD s Five-Year Change in Cash Balance $30,000 $25,000 Cash balance % of 2013 revenues 32.1% 35% 30% $ 000s $20,000 $15, % 25% 20% 15% $10,000 10% $5,000 5% $0 0% Sources: Radnor SD, Moody s Investors Service Input: 15.5% Score: JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

11 Subfactor 3: Management (20%) Subfactor 3a: Institutional framework (10%) What it is: The legal and practical context defining the issuer s obligations and resources. What it shows: Our assessment of the effect of the legal and operating environment on the issuer s ability to match resources with obligations. How we measure it: At least once a year, our analysts hold a formal meeting to determine the institutional framework score for every sector in every state. The determination incorporates the issuer s ability to control and predict its revenues as well as its ability to control and predict its expenditures. The meetings that set this score involve a discussion of financial, economic and legislative trends in each state and the potential impact of these trends on the credit profiles of its issuers. All institutional framework scores are reviewed at least once per year. We may change the institutional framework score at any point during the year, if warranted (e.g., if legislation is enacted that changes a sector s revenue raising flexibility up or down). Factors that influence the institutional framework score include property tax caps, the strength of organized labor and the trend of state aid. We have one score for each state/sector combination. For example, all Ohio school districts have the same institutional framework score. The rubric below guides the analysis. See here for the complete current list of institutional framework scores. EXHIBIT 7 The Institutional Framework Rubric Revenue Raising Ability Moderate ability to raise revenues Operating Revenue Flexibility Strong ability to raise revenues Weak ability to raise revenues Revenue Predictability Major revenue sources tend to be highly stable and predictable Major revenue sources tend to be moderately stable and predictable Major revenue sources tend to be somewhat unstable and unpredictable Aaa Aa A Major expenditures tend to be highly stable and predictable Aa A Baa Major expenditures tend to be moderately stable and predictable A Baa Ba or B and Below Strong ability to reduce expenditures Moderate ability to reduce expenditures Expenditure Reduction Ability Weak ability to reduce expenditures Major expenditures tend to be somewhat unstable and unpredictable Expenditure Predictability Operating Expenditure Flexibility Source: Moody s Investors Service Our score for the Pennsylvania school district sector is moderate, which maps to an A on the scorecard. Our assessment of the score incorporates the fact that school districts operate under a property tax cap, that organized labor is fairly strong in Pennsylvania (which lowers a government s flexibility to reduce costs), and that the commonwealth has shown a willingness to cut aid to school districts. 11 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

12 To highlight the importance of the tax cap to this score, we score Pennsylvania cities and counties (not bound by the cap) as strong, or Aa. The primary difference between the score for school districts and cities or counties is the tax cap. Input: A Score: 3 Frequently asked questions Doesn t mapping the institutional framework scores to integers prevent any issuer from achieving the maximum score of 0.5? Yes. These are scored at the broad rating category level, and the integer value represents the midpoint of these levels. What if an issuer operates in an unusual situation? Would you still use the statewide score? Some issuers may operate under unusual circumstances that place them outside the norm for their state/sector. For example, the Philadelphia School District (Ba2 review for downgrade) operates under a decidedly different institutional framework from Radnor SD. The Philadelphia School District cannot determine its own tax levy, while Radnor SD can. For exceptions like Philadelphia SD, we choose to still use the standard institutional framework score and address exceptions below the line. Subfactor 3b: Operating history (10%) What it is: The results of the issuer s financial operations over the past five years. What it shows: The issuer s record at achieving financial equilibrium or generating surpluses or deficits from operations, over a full economic cycle. How we measure it: The formula is shown here: Operating Revenues, Year 1/Operating Expenditures, Year 1 + Operating Revenues, Year 5/Operating Expenditures, Year 5 5 Exhibit 8 illustrates Radnor SD s operating history. 12 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

13 EXHIBIT 8 Radnor SD s Operating History $ 000s $84,000 $82,000 $80,000 $78,000 $76,000 $74,000 $72,000 $70,000 $68,000 $66,000 $64,000 $62,000 Operating revenues (left) Operating expenditures (left) Ratio (right) Sources: Radnor SD, Moody s Investors Service The average of the five years of ratios is 1.06, indicating that the district has run a 6% surplus annually on average. Input: 1.06 Score: 1.4 Frequently asked questions Would you include transfers from a utility fund to the general fund as a part of its operating history? Yes, if the transfers are recurring and we view them as likely to be stable in the future. Would you exclude non-recurring expenditures from the operating history? Yes, non-recurring expenditures do not indicate the operating trend, and we do not include them in this formula. Factor 4: Debt/Pensions (20%) Subfactor 4a: Debt to full value (5%) What it is: The issuer s net direct debt as a percentage of full value. What it shows: The burden of the issuer s long-term debt on the tax base. How we measure it: Net direct debt (tax-supported debt minus debt deemed self-supporting) divided by the full value input in Subfactor 1a. The net direct debt input captures all debt payable from the local government s taxes minus debt reliably paid from a self-supporting essential-service enterprise. Net direct debt includes the following:» direct General Obligation bonds and Bond Anticipation Notes» capital leases» GO-guaranteed utility bonds unless they are both essential-service and self-supporting» bonds secured by sales taxes, income taxes, hotel taxes, etc. 13 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

14 » guaranteed debt of a non-essential enterprise» any other debt we judge as likely to be secured by the local government s taxes Radnor SD s $99,443,889 of GO debt is equal to 2.34% of full value. Input: 2.34% Score: 2.76 How do you determine whether a GO-guaranteed essential-service utility is self-supporting? We calculate whether the utility s net revenues (operating revenues less operating expenditures, excluding depreciation) have been sufficient to cover debt service in each of the past three years. If the utility fails this test, we do not consider it self-supporting for the purposes of our scorecard. How do you determine whether a utility is essential-service? As a general rule, only GO-guaranteed utilities we consider essential are eligible to be deducted from a GO debt portfolio as self-supporting. Essential-service revenue bonds tend to have more stable revenues and are more likely to support their own debt over time. Non-essential-service revenue bonds are more vulnerable to operational swings and are less likely to consistently support their own operations. However, if there are compelling circumstances supporting a different outcome, we will deviate from this guideline. EXHIBIT 9 Essential Service Utilities Likely eligible to be considered self-supporting Water Sewer Stormwater Electric Gas Unlikely to be eligible to be considered self-supporting Healthcare (hospitals, nursing homes) Recreation (golf courses, ski resorts, skating rinks) Entertainment (sports arenas, concert halls, museums) Electrical generation (waste-to-energy, or merchant plants) Telecommunications (fiber-optic cable, phone service) Commercial/residential real estate development, revenue generated from sale or lease Parking system Hotel/Convention center Source: Moody s Investors Service What if a utility is funded by taxes? We include all tax-supported debt in the direct debt burden. Would you include cash flow notes in the debt burden? No, the debt burden measures long-term debt, not notes to be repaid from short-term cash flows. We do, however, include bond anticipation notes as they will eventually become long-term debt. 14 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

15 Subfactor 4b: Debt to revenues (5%) What it is: The issuer s net direct debt taken as a multiple of revenues. What it shows: The burden of the issuer s long-term debt on its budget. How we measure it: Net direct debt (always the same as the input in Subfactor 4a) divided by revenues (as defined in operating revenues on p. 7). Debt to revenues is virtually the same as debt to full value, except we are measuring its budgetary impact instead of the impact on the tax base. We use the same operating revenue denominator in all of our scorecard metrics with revenues. Radnor Township s $99,483,889 of direct debt is equal to 1.21 times revenues. Input: 1.21 Score: 2.73 Subfactor 4c: Three-year average of Moody's adjusted net pension liability to full value (5%) What it is: The unfunded pension liability as a percentage of full value. What it shows: The burden the issuer s unfunded pension obligations place on the tax base. How we calculate it: Because state and local government pension systems use a variety of different assumptions in calculating their liabilities, we adjust and standardize all local government pension data for comparability, arriving at a Moody s adjusted net pension liability (ANPL). The main adjustment we make is to use the Citibank Pension Liability Index (which includes Aa and higher corporate bonds) as the discount rate instead of the various discount rates used by the plans. For cost-sharing plans, such as the Pennsylvania Public School Employees Retirement System that Radnor SD participates in, we apportion the total liabilities of the plan to each participant based on its respective share of contributions to the plan. We use a three-year average of the local government s ANPL. The district s ANPL has averaged 1.82% of full value the past three years. Input: 1.82% Score: 2.27 Frequently asked questions Why not use the discount rates assumed by the pension plan? We believe the bond-like nature of future pension payments warrants discounting them based on current market bond rates, for purposes of producing a balance sheet liability for credit analysis. Also, because different pension plans use different discount rates, we use a common rate to facilitate a comparison between different issuers reporting as of the same date. Don t year-to-year changes in bond market rates change the present value of pension liabilities? Yes, and the reason we take the three-year average is to smooth out some of that volatility. Is ANPL the only pension measure you use? No, we also look at the issuer s reported pension liability, particularly the unfunded actuarial accrued liability (UAAL), and the required annual pension contributions as part of a fixed-cost analysis of their 15 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

16 operations (along with debt service and other post-employment benefits. As with our other major rating factors, we assess information that is neither above nor below the line in the scorecard. Radnor SD has moderate fixed costs (see Exhibit 10). EXHIBIT 10 Radnor SD s Modest Fixed Costs Fixed Costs Pension Contributions % of operating expenditures OPEB % of operating expenditures Debt Service % of operating expenditures Total 2011 $2,214,845 3% $614,758 1% $7,167,000 9% 13% 2012 $3,727,638 5% $614,878 1% $7,564,000 10% 15% 2013 $4,701,096 6% $370,182 0% $7,218,000 9% 15% Source: Moody s Investors Service, Radnor SD Subfactor 4d: Three-year average of Moody's adjusted net pension liability to revenues (5%) What it is: The pension liability as a multiple of revenues. What it shows: The burden the issuer s pension obligations places on the budget. How we calculate it: The same adjustments that determine the ANPL for Subfactor 4c are used for 4d. However, for this Subfactor, we take the ANPL to revenues to assess the relative budgetary impact. Input: 0.94x Score: 2.55 The district s above-the-line score of 1.83 places it right on the cusp of Aa1 and Aa2. The scorecard is not complete until we have made below-the-line adjustments. Below-the-line adjustments: Case study of Radnor SD Once we calculate the above-the-line score, we determine whether the scorecard has failed to capture any important credit factors. The above-the-line scorecard is based on universally measurable factors, and often misses some of the idiosyncratic characteristics that drive an issuer s credit quality. Following is a discussion of some of our most commonly used below-the-line adjustments, and how we considered them in the context of Radnor SD s rating. Economy/Tax Base Institutional presence: Up 0.5 notches. Radnor Township is home to Villanova University (A1 stable), a private university with enrollment of more than 10,000 undergraduate and graduate students. We are likely to employ an upward adjustment for a tax base with a stable institution representing a sizeable percentage of the population. Institutions like Villanova help to drive employment and economic activity. Universities also tend to suppress socioeconomic indicators; for example, Radnor Township s full value per capita of $134,887 would be higher without the many students and largely tax-exempt property of the university. We considered a full-notch adjustment for Villanova, but ultimately decided on a half-notch adjustment since the strength of the tax base was already wellreflected in the economy/tax base factors. We do not have a formal rule for when to notch up and by how much, although the rating committee chairs take responsibility for assuring consistent application 16 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

17 of the scorecard variables. We will usually notch up for state capitals or stable institutions representing more than 10% of the population. We usually will not notch up for county seats, smaller institutions, or institutions that are in decline. Regional economic center: Not applicable. We reserve this upward notch for tax bases, such as Philadelphia (A2 stable) or Pittsburgh (A1 stable), with substantial employment opportunities drawing strength to the tax base that is not captured through socioeconomic metrics such as resident median family income. Economic concentration: Not applicable. We might notch down if the top 10 taxpayers represent more than 15% or 20% of AV, or if a large portion of the tax base relies on a single industry, particularly if that industry is volatile or in decline. We may use this notch for coal mining towns, or areas where most residents work at a paper mill. The 10 largest taxpayers in Radnor SD are only 9.5% of AV, a concentration that will rarely warrant an adjustment. Most of the district s largest taxpayers are office complexes, implying a diversity of industries within the district. Outsized unemployment or poverty levels: Not applicable. Radnor Township s poverty rate as of 2012 was only 5.6%, well below the national rate. We might utilize this downward adjustment for tax bases such as the Philadelphia School District (poverty rate: 26.2%); the Reading School District, PA (Baa2 negative; poverty rate: 37.9%); or the York City School District, PA (Baa2 negative; poverty rate: 36.4%). We do not have a threshold for how high poverty or unemployment needs to be in order to employ this notch, though rates at double the state rate would most likely cause us to at least consider an adjustment. An adjustment hinges on our judgment of whether the tax base is able to bear the costs of paying more taxes, whether the issuer has outsized social service spending needs, and whether the scorecard already captures the credit impact of this factor. Other analyst adjustment to economy/tax base: Not applicable. All other socioeconomic indicators confirm that Radnor Township is a well-above-average tax base. The median home value as of 2012 was more than $600,000, and per capita income was far above average at $52,934. Overall, with the half-notch adjustment for institutional presence, the strength of the tax base is sufficiently captured. Finances Outsized contingent liability risk: Not applicable. We use this downward notch for issuers that have guaranteed debt or financial exposure to a non-essential or faltering enterprise, such as Harrisburg. We would use this adjustment for issuers whose enterprise exposures are large relative to financial resources. See Contingent Liabilities and Enterprise Risk Continue to Weigh on US Local Governments for how we approach this factor. Unusually volatile revenue structure: Not applicable. The district s revenue structure is less volatile than most (see Exhibit 11). We generally consider property taxes to be the most stable and predictable revenue source. We might notch down if a significant share perhaps more than 50% of an issuer s revenues come from income taxes, state aid, or other less predictable sources. We also may notch down if an important revenue source has declined by more than 10% any year in the past five. We assess whether the volatility of the issuer s revenues is interfering with its ability to balance its budget and/or is likely to do so going forward. 17 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

18 EXHIBIT 11 Radnor SD s Stable Revenue Structure Revenue Composition 2015 Budget % of Revenues Property tax levy $65,942,577 77% Commonwealth aid $12,040,066 14% Tuition $1,895,000 2% Realty transfer taxes $1,500,000 2% Other $4,582,833 5% Total $85,960, % Source: Radnor SD Other analyst adjustment to Finances factor: None. Management State oversight or support: Not applicable. We only notch up if a state program affords a clear and demonstrable benefit to the issuer. We may notch up for state oversight programs that have proven effective, such as Philadelphia with the Pennsylvania Intergovernmental Cooperation Authority. We also reserve the option to notch down if we find that state oversight weakens credit strength for bondholders. Unusually strong or weak budgetary management and planning: Up 0.5 notches. The above-the-line scorecard tells most of the story of the district s strong financial trend, but not all. The robust cash and fund balances, and their increases over the past few years, earn the district high above-the-line scores. They do not show that the district has accomplished this performance at a time when most other school districts in Pennsylvania have been drawing down fund balance; when pension contributions have been rising every year; when commonwealth aid was cut in 2010; when property tax increases are capped; and when full value has declined (down 15% since 2007). Ultimately, we felt the district s financial performance against this backdrop was strong enough to warrant an extra half-notch. This notching demonstrates that our scorecard metrics are not deterministic. In this case, we felt that factors measured above the line in finances, and operating history still did not fully reflect the true credit strength of the issuer. Other analyst adjustment to Management factor. None. Debt/Pensions Unusually strong or weak security features. Not applicable. We typically employ an upward adjustment if the issuer has a separate tax levy devoted solely to paying debt service in its own fund. Another instance when this adjustment might make sense is a lock-box structure, in which funds to pay debt service are automatically diverted to a trustee. We might employ this notch downward if the security is something less than an unlimited GO pledge, such as a GO limited tax security that we conclude is materially weaker than a real or assumed GO unlimited tax pledge of the same issuer. Unusual risk posed by debt/pension structure. Not applicable. Nothing about the district s debt or pensions is unusual. We may use this downward notch for issuers with a large exposure to puttable variable rate debt, which can create an unexpected claim on the issuer s resources. Radnor SD does have some variable-rate debt: a $15 million loan from the Delaware Valley Regional Finance Authority. The district also has a swap associated with this variable-rate loan. The district s variablerate exposure does not present pronounced risks: variable-rate debt is only 15% of the total debt burden, and the district s liquidity exceeds total VR debt. 18 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

19 We may also notch down if an issuer s debt structure forces it to rely on market access to continue operating, or if its debt service schedule is back-loaded or steeply escalating. History of missed debt service payments. Not applicable. We are likely to use this downward adjustment, perhaps with multiple notches, for an issuer that has defaulted, in particular if the default arose from a inability or unwillingness to honor debt service obligations. This could also apply to a default on a non-parity obligation, such as a lease, and would apply regardless of whether we rated the defaulted series. An administrative default caused by, say, a clerical mistake rather than a lack of willingness or ability to meet obligations, may also result in a downward notch if it indicates weak management. Other analyst adjustment to Debt/Pensions factor. None. We might consider additional downward notching for issuers with high overall debt or pension burdens resulting from the liabilities of overlapping units of local government. Anatomy of an upgrade With the two half-notch upward adjustments, the district s scorecard score is 1.49, which places it at the cusp of Aaa and Aa1. The scorecard helped to inform our decision to upgrade the district to Aa1. It was part of a multi-pronged approach, in which we considered the district s strong tax base, excellent financial management, above-average financial position, and moderate debt and pension burden. 19 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

20 Moody s related research Rating Methodologies:» US Local Government General Obligation Debt, January 2014 (162757)» Adjustments to US State and Local Government Reported Pension Data, April 2013 (151398) Sector Comment:» Vast Majority of US Local Government Ratings Are Unchanged Upon Implementation of New GO Methodology, April 2014 (167410) To access any of these reports, click on the entry above. Note that these references are current as of the date of publication of this report and that more recent reports may be available. All research may not be available to all clients. 20 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE

21 Appendix A Maximum and Minimum Scorecard Scores Scorecard limits Upper Bound Lower Bound Tax Base 50,000,000 0 Full Value Per Capita 400,000 0 Socioeconomic Indices % 0.00% Fund Balance 60.0% -15.0% 5-yr change in Fund Balance 50.0% -25.0% Cash Balance 40.0% -15.0% 5-yr change in Cash Balance 50.0% -25.0% Operating History (5yr avg) Direct Debt / Full Value 0.00% 20.00% Direct Debt / Revenue Moody's ANPL / Full Value 0.00% 24.00% Net Pension Liability / Revenue School Dist. Cash Balance 20.0% -15.0% School Dist. Fund Balance 50.0% -15.0% 21 JULY 24, 2014 SPECIAL COMMENT: US LOCAL GOVERNMENT GENERAL OBLIGATION METHODOLOGY AND SCORECARD USER GUIDE