The content on this website is intended for investment professionals and institutional asset owners. Individual retail investors should consult with their financial advisers before using any of the content contained on this website. Breckinridge uses cookies to improve user experience. By using our website, you consent to our cookies in accordance with our cookie policy. By clicking “I Agree” and accessing this website, you represent and warrant that you are agreeing to the above statements. In addition, you have read, understood and agree to the terms and conditions of this website.

Municipal Perspective published on April 13, 2020

The Fed’s New Municipal Liquidity Facility: Modest Help or Market Game-Changer?

The following Perspective was published on April 13, 2020. On April 27, the Federal Reserve expanded and amended certain aspects of the Municipal Liquidity Facility. Significant changes include extending the termination date for the program until December 31, 2020, expanding the group of governments and entities eligible for direct loans, and extending the maximum length of loans to three years from 24 months.

 

On Thursday, April 9, the Federal Reserve announced the creation of the Municipal Liquidity Facility (MLF). The MLF will extend up to $500 billion in short-term funds to state and local governments using $35 billion in seed capital appropriated in the Coronavirus Aid, Relief, and Economic Security Act (CARES Act).

The MLF’s size and mandate differ meaningfully from the Fed’s other recently launched facilities for municipals: the Money Market Mutual Fund Liquidity Facility (MMFLF) and the Commercial Paper Funding Facility (CPFF).1 The former improves liquidity in the secondary market by permitting banks to post short-term municipals as collateral in exchange for Fed loans.2 The latter buys three-month municipal paper, which is a limited charge given the tiny market for three-month munis (Appendix A).

By contrast, the MLF is intended to “help state and local governments manage cash flow stresses caused by the coronavirus pandemic.”3 The program authorizes a Fed Special Purpose Vehicle (SPV) to buy 24-month notes directly from municipal issuers, and permits eligible munis to borrow up to 20 percent of their own-source general government and utility revenue in 2017.4 That’s equal to $515 billion, nearly identical to the $500 billion authorized under the MLF and a reasonable proxy for state and local governments’ possible cashflow needs in 2020 (see Appendix B). The facility’s $500 billion in lending authority exceeds outstanding short-term municipal debt by 11 times (Figure 1). The program likely exceeds the Fed’s legal authority. The Federal Reserve Act and CARES Act are explicit that the MLF, and programs like it, are intended to promote market liquidity, only.5 However, it will plainly be an important player in the municipal market in 2020.

In the near-term, the MLF is likely to help stabilize the market for short-term municipals without purchasing a significant amount of short-term debt. Weaker, but solvent, municipal borrowers with routine cash flow needs should benefit from improved overall liquidity and, consequently, access to short-term capital on better terms. Likewise, few municipal issuers are likely to make use of the MLF, immediately. Most lack urgent cash flow challenges despite the COVID-19 crisis, and for some, launching a cash-flow borrowing program, through the MLF or on the open market, may present cumbersome administrative hurdles.

However, if the COVID-19 pandemic persists, state and local deficits could become more acute. In that event, the MLF could become a more popular municipal deficit-financing option later this year. The short-term municipal market is small, and state and local governments would likely happily reach out for the Fed’s extended hand. This would mark a significant development in the municipal market, with positive near-term, but mostly negative long-term, credit implications. As a reminder, until just weeks ago, the Fed had never intervened in the municipal bond market.

Municipal Liquidity Facility (MLF) Basics

The MLF joins two recently created Fed programs that benefit municipals: (1) the Money Market Mutual Fund Liquidity Facility and (2) the Commercial Paper Funding facility. Whereas those programs were relatively limited in scope and designed to promote secondary market liquidity, the MLF is authorized to buy up to $500 billion in short-term munis and will purchase the securities, directly.

Per the Federal Reserve’s April 9 term sheet, the MLF will6:

  1. Purchase only short-term notes with maturities no longer than 24 months (e.g., BANs, TANs, TRANs, and like securities).
  2. Lend only to: (a) the 50 states, (b) the District of Columbia, (c) cities with more than one million residents, or (d) counties with more than two million residents. Only one issuer per state, city, or county is eligible to receive funds, but that issuer can issue debt multiple times.
  3. Lend, in an aggregate amount, up to 20 percent of the issuer’s general own-source and utility revenue.7 States can request an exception to this rule to enable lending to otherwise ineligible political subdivisions and instrumentalities.
  4. Permit borrowers to call their notes prior their stated maturity.
  5. Permit borrowers to lend note proceeds to other political subdivisions and public entities within their state.
  6. Purchase notes at a price consistent with a borrower’s bond rating. (Note that this mandate requires no actual rating for the notes that the Fed purchases. It just requires pricing consistent with the issuer’s existing rating).
  7. Purchase notes issued to finance tax deferrals, deficits stemming from the COVID-19 pandemic, or meet principal and interest obligations.

The MLF would accomplish the above by capitalizing a SPV with an initial equity cushion of $35 billion, appropriated by Congress in Section 4027 of the CARES Act. It is authorized to lend per Section 13(3) of the Federal Reserve Act.8

The Fed’s term sheet was silent as to whether the MLF will purchase tax-exempt or taxable paper. Presumably, it would purchase both, given that some portion of municipal cashflow borrowing over the next several months will violate the Internal Revenue Service’s (IRS) anti-arbitrage rules.9 On the other hand, tax rules may not matter much to the Fed. It has stated that it will buy at the going market price, which is likely a tax-free one, and it pays no tax.

The term sheet also included no information as to whether a broker-dealer (e.g., J.P. Morgan) or asset manager (e.g., Blackrock) will manage the assets of the MLF.

Near-Term Support for Market Liquidity

The MLF is likely to stabilize the market through better liquidity and pricing in the coming weeks. The main beneficiaries are likely to be well known issuers with low investment grade ratings, low default risk, and meaningful short-term cash-flow borrowing needs. The New York MTA is the most obvious large issuer that stands to benefit from the program. It has significant near-term cash flow needs and pricing for its short-term paper has been weak, of late.10 However, the MLF is also likely to indirectly support short-term municipal borrowers of all kinds, including large hospital systems, airports, and states with short-term borrowing programs.

Few Issuers Likely to Utilize the MLF in the Near-Term

However, most issuers are unlikely to tap MLF or the short-term market, at all, over the next few months. This is true for several reasons.

First, many state and local issuers are well capitalized. State reserves reached a three-decade high in 2019 (Figure 2). Median days-cash-on-hand for water-sewer and electric utilities exceeded 400 and 150 days, respectively, in FY 1911 This liquidity is in addition to roughly $150 billion in funds state and local governments will receive per the recently passed CARES Act.12 It is also in addition to internal borrowing that is often available to issuers at lower cost than external financing.

Second, issuing cash flow notes imposes administrative burdens in many jurisdictions. Cash-flow borrowing will require new legislation in many states. State laws typically limit the sum and tenor of municipalities’ short-term notes.13 Often, notes must be repaid from revenues that accrue in the same year they are issued; this may conflict with the maximum 24-month note structure envisioned by the MLF. Accurate cash flow forecasts are also commonly required prior to issuance. This could prove challenging in the current environment.14

Third, the MLF is likely to be most effective through a bond bank or revolving loan fund structure. The MLF permits only one eligible borrower per state, large city, or large county–but allows borrowers to relend proceeds to other municipal issuers. This provision may limit the number of states that can quickly take advantage of the program. Authorizing statutes for bond bank or revolving loan pools can likely be repurposed for the MLF, but new legislation might be needed elsewhere. Only about half of the states have a bond bank or loan pool of substantial import.

There has been some early market chatter that the MLF’s relending provision will enable leveraging of the facility’s capital. But this, too, may prove administratively challenging. Lending against tax-free, bond-financed capital is likely to invite IRS scrutiny.15 The maximum two-year term on MLF notes may also prove problematic, as most bond banks and state revolving funds are accustomed to extending longer-term loans. In any event, municipal pooled financing programs generally operate with modest leverage. The Maine Municipal Bond Bank operates one of the market’s long-standing programs. It is leveraged only 3:1.16

The MLF Could Become More Popular If Deficits Become Unexpectedly Overwhelming

However, if the COVID-19 crisis persists and municipal deficits become more burdensome, state and local governments may want to borrow from the MLF in amounts nearing the MLF’s $500 billion in authorized lending. As a rule-of-thumb, each one-percentage point increase in the national unemployment rate decreases state and local taxes by around 3 percent (Figure 3).17 If the U.S. unemployment rate increases to 15 percent from its pre-COVID-19 crisis level of 3.5 percent in February 2020, it would imply an annual decline in state-local taxes of 34 percent, nationally. Actual revenue declines will vary widely, issuer-to-issuer, but many state and local governments might wish to borrow for operations.  

A jump in municipal short-term debt on the scale of $500 billion could overwhelm the short-term municipal market and incent significant borrowing from the MLF. Short-term notes comprise only 1.5 percent of the $3.8 trillion municipal market, and it’s unclear who might buy up to $500 billion in short-term municipals.18 The buyer base for short-term municipals is predominantly retail investors, and in recent years, they have reallocated away from low-yielding tax-free money market funds (Figure 4).

Banks, insurers, and foreign buyers could increase their purchases of municipals to offset some of the new supply, but it’s doubtful their participation would be enough. Combined, these participants held $1.07 trillion in municipal bonds in Q4-19.19 They would have to increase their exposure by 46 percent to meet the increase in short-term paper ($500 billion over $1.07 trillion = 46 percent). Notably, these buyers reduced their municipal exposure by $33 billion in 2019.20

Importantly, municipal borrowers would suffer little reputational damage from tapping the MLF. In corporate market, borrowing from a Fed facility communicates financial weakness to investors, and companies seek to avoid it. In the municipal market, affordable capital from the Fed that limits cuts to essential services is likely to prove politically popular.

Direct Lending from the MLF to Municipals Has Mostly Negative Credit Implications

Investors should be clear-eyed about the MLF’s potential to weaken long-term municipal credit fundamentals, if the COVID-19 crisis forces issuers to resort to cash-flow borrowing on an unprecedented scale. In the very near-term access to cash is likely to mitigate the risk of unnecessary bond defaults. But on balance, investors should be circumspect about Fed intervention.

First, $500 billion in new borrowing would significantly increase the overall state and local debt burden. State and local government debt outstanding totaled $3.1 trillion in 2019. Another $500 billion would increase the debt load by 16 percent.21 Over the long-term, this kind of increase is probably manageable. State and local bonded debt has decreased as a percentage of gross domestic product since the Great Recession, from 20 percent to 14 percent.22 Last year, Moody’s Investors Service suggested that states should increase their borrowing to fund vital transportation needs.23 Still, an increase in debt on such a large scale, so quickly, and for operating needs, would not be ideal.

Second, the structure of the debt would be short-term. State and local governments would ideally repay it from current revenues, but the more likely outcome is that they refinance it into longer bonds. Repaying short-term maturities via refinancing could create market access risks for certain issuers. This risk could grow if issuers seek to refinance their bonds, en masse, during the same several-month period one to two years from now. Overall market issuance could conceivably increase meaningfully in late 2021 and 2022.

Third, existing bondholders might be diluted or subordinated to short-term MLF notes in certain cases. For example, if Chicago borrows from the facility, it might leverage its Sales Tax Securitization Corporation (STSC) lien to do so. In this case, an MLF loan would likely reduce the amount of pledged sales tax for existing holders, and the risk would grow if the STSC decided to pay the short-term maturities from ongoing revenue. In other instances, MLF notes might come with acceleration clauses or springing liens.

Given the risks, disclosure in the municipal market may need to improve. Investors will want to surveil their holdings more frequently to understand their cash flow needs on a real-time basis. Presently, audited financial reports are published on a lag of 180 days or more. Monthly tax revenue and cash reports are available for most governments with substantial capital market needs, but for smaller issuers, more disclosure and oversight may be required.

Investors should take some comfort in the structure of the MLF insofar as it makes funds eligible solely to the largest and most sophisticated borrowers. These borrowers will likely act as gatekeepers for financing for smaller and less sophisticated issuers. Recall that eligible borrowers can relend MLF proceeds to other political subdivisions and public corporations. Internal borrowing of this sort is safer than the alternative: allowing small issuers with more limited market access to expose themselves to refinancing risks.

It’s also true that widespread MLF deficit-financing would likely speed the economic recovery from the COVID-19 pandemic. State and local governments typically cut spending during recessions, exacerbating their depth by depressing investment and consumption. If the MLF helps avoid cuts in government spending, it may give the broader economy a Keynesian boost, notwithstanding the reality that state and local governments will need to repay MLF loans, eventually.

Nonetheless, the MLF likely increases municipal credit risk, overall. There is even a remote possibility that the MLF becomes a tool for municipal deficit-financing not just in 2020 but in future recessions. A Fed facility that has successfully financed state and local deficits during the COVID-19 “panic” could conceivably be reopened in future “exigent circumstances.” Knowledge that such a facility is possible may, over time, erode some of the long-standing rules that limit state and local governments’ ability to borrow to finance operating deficits. These rules have contributed to the muni market’s historically low default rate.

Breckinridge intends to monitor developments in the short-term municipal market and the MLF, closely, in the coming months.

[1] https://www.federalreserve.gov/newsevents/pressreleases/monetary20200323b.htm

[2] The program includes variable rate demand notes and bonds in its mandate, as “short-term” municipals.

[3] Ibid.

[4] “Own-source revenue” is a public finance term that includes only revenue derived by the government’s own activities or investments. For example, a state’s federal Medicaid and highway funding would not be considered own-source revenue. By contrast, a state’s tax revenue, or fines imposed for speeding on state highways, would be considered own-source revenue.

[5] Federal Reserve Act 13(3)(B)(i): “… such policies and procedures shall be designed to ensure that any emergency lending program or facility is for the purpose of providing liquidity to the financial system…” and Sec. 4003(a)-(b)(4) of the CARES Act: “…Loans, loan guarantees, and other investments made pursuant to [this section] shall be made available as follows… Not more than the sum of $454 billion… for the purpose of providing liquidity to the financial system…” (italics added by Breckinridge).

[6] Ibid.

[7] The 20% threshold is based on Census of Governments data from 2017..

[8] This provision permits Federal Reserve bank lending in “unusual and exigent circumstances” when an eligible borrower is “unable to secure adequate credit… from other banking institutions.” Federal Reserve Act, section 13(3)(A).

[9] 26 USC §148.

[10] Per the Breckinridge Trading desk, some MTA bond anticipation notes were commonly trading at yields in excess of 4% during the first week of April 2020.

[11] Breckinridge analysis of Merritt Research data.

[12] Coronavirus Aid, Relief, and Economic Security Act, §601.

[13] For example, in Georgia, tax anticipation notes (TANs) must be paid by December 31st of the calendar year in which they were issued and cannot exceed 75% of annual tax revenue from the prior year. Overview of [Georgia] Governmental Financing, Smith, Gambrell, & Russel, LLP. Available at: https://www.sgrlaw.com/briefings/461/.

[14] For example, in the last three weeks, the State of New York has revised down its revenue estimates twice. The budget director now expects a revenue loss of up to $15 billion, compared to $7 billion Mid-March.https://www.wsj.com/articles/coronavirus-crisis-could-cut-new-york-state-revenue-by-up-to-15-billion-cuomo-aide-warns-11585056388

[15] IRC 148.

[16] Loans receivable (short- and long-term) exceeded cash and investments (unrestricted and restricted) by 2.95x in FY 18. Figures exclude advances to the State of Maine. See the Maine Bond Bank’s Comprehensive Annual Financial Report, 2018 (p. c-10).

[17] Breckinridge analysis of Bureau of Economic Analysis and U.S. Census data. Analysis begins in Q1-1990. Note that there is a wide variation around the relationship, depending on the issuer.

[18] Federal Reserve Flow of Funds, Table L. 212 (March 2020).

[19] Ibid.

[20] Federal Reserve Flow of Funds, Table F. 212 (March 2020)

[21] Federal Reserve Flow of Funds, Table L. 212 (March 2020).

[22] Breckinridge analysis of Federal Reserve and Bureau of Economic Analysis data.

[23] “US states’ net tax supported debt levels remain flat for eighth consecutive year,” Moody’s Investors Service, June 3, 2019.

 

DISCLAIMER: The opinions and views expressed are those of Breckinridge Capital Advisors, Inc. They are current as of the date(s) indicated but are subject to change without notice. Any estimates, targets, and projections are based on Breckinridge research, analysis and assumptions. No assurances can be made that any such estimate, target or projection will be accurate; actual results may differ substantially.

Nothing contained herein should be construed or relied upon as financial, legal or tax advice. All investments involve risks, including the loss of principal. Investors should consult with their financial professional before making any investment decisions.

While Breckinridge believes the assessment of ESG criteria can improve overall credit risk analysis, there is no guarantee that integrating ESG analysis will provide improved risk-adjusted returns over any specific time period.

Some information has been taken directly from unaffiliated third-party sources. Breckinridge believes such information is reliable but does not guarantee its accuracy or completeness.

Any specific securities mentioned are for illustrative and example only. They do not necessarily represent actual investments in any client portfolio.

BLOOMBERG® is a trademark and service mark of Bloomberg Finance L.P. and its affiliates (collectively “Bloomberg”). BARCLAYS® is a trademark and service mark of Barclays Bank Plc (collectively with its affiliates, “Barclays”), used under license. Bloomberg or Bloomberg’s licensors, including Barclays, own all proprietary rights in the Bloomberg Barclays Indices.

Neither Bloomberg nor Barclays approves or endorses this material, or guarantees the accuracy or completeness of any information herein, or makes any warranty, express or implied, as to the results to be obtained therefrom and, to the maximum extent allowed by law, neither shall have any liability or responsibility for injury or damages arising in connection therewith.