Municipal credit analysis you talk to: every ACFR spread, every covenant tested, every figure on its page, every view signed by two people
Every held obligor spread every year and every covenant tested, each figure on its page.
See one case, screen by screen ↓Municipal credit analysis, or municipal bond credit analysis, is how a muni research team judges whether an issuer can pay its bonds: spreading its ACFR, testing debt service coverage against the rate covenant in the bond resolution, reading pension and retiree health liabilities, and watching its continuing disclosure. Municipal Credit Analysis gives that team seven agents that read every filing, spread it with every value pinned to its page, map the numbers to the resolution and draft the view, for an analyst to question and submit and a senior analyst to approve.
AI in asset management: a buyer's guide to agentic AI for investment managers →
Updated
Municipal credit analysis starts with paper nobody standardized. A muni research team covers thousands of obligors: cities, school districts, water and sewer systems, ports, airports, hospitals, special tax districts. Each files its annual comprehensive financial report (ACFR) to the municipal disclosure feed on its own deadline, in its own layout, often 200 to 400 pages long. For each one an analyst copies the fund statements, debt schedules and pension notes into the sector template, finds the bond resolution to see which revenues are pledged and in what order they are paid, rebuilds debt service coverage the way the resolution defines it, reads the GASB 68 and GASB 75 notes, and only then writes a credit view.
So the largest holdings get a full spread and most obligors get a glance. The risk sits in the glance: a water system whose coverage slipped from 2.06x to 1.84x while operating costs rose, a sanitation district below its rate covenant, a port drawing on its debt service reserve, and a new official statement that says the issuer complied with its continuing disclosure agreement when last year's report was filed 41 days late. Under Rule 15c2-12 a listed event notice is due within 10 business days, and someone still has to read it before the next deal prices.
Estimated split for one revenue bond obligor, read in full by hand and with the solution; with the agents, the analyst's hours go to questioning, reviewing and deciding.
Seven agents read, spread, map, tie and explain every filing; an analyst submits, a senior analyst signs.
The City of Harlow Bend, Ohio is bringing $92.5M of Series 2026 water and sewer bonds, pricing Tuesday, October 27, 2026. Quillfield Capital holds $38.6M par of its Series 2019 bonds across two funds and 412 SMAs. It is Tuesday, October 20, and the agents opened a new-issue review the day the preliminary official statement posted. Here is how the analyst works with them, in the working solution.
Elena covers 486 obligors; 61 annual reports arrived this month, all spread. Today shows four items: Harlow Bend's new-issue review, Quarry Hill Sanitation at 1.19x against a 1.20x covenant, Calder Point Ports' $2.1M draw on its debt service reserve, due to be read today, and Emberfield Transit's annual report, 20 days overdue. Beside them, what the agents did overnight, every line cited, and a box to ask them anything.
Disclosure Watcher, 6:48 a.m.: “Read 1,284 annual reports in the last 30 days and 146 event notices this week on held obligors; flagged 2 official statements that leave out a late filing.”
The credit file opens on the money as Ordinance No. 2016-41 §5.03 says it flows: Gross Revenues of $112.4M, less Operation and Maintenance of $63.8M, with $21.4M of depreciation excluded as §1.01 defines it, leaves Net Revenues of $48.6M. Then the Senior Bond Fund, the reserve with no deposit due, subordinate state revolving fund loans, renewal and replacement, the payment in lieu of taxes and the balance. On the right, each agent says what it did and how sure it is, with a proposed view waiting for her decision.
Rate covenant test, §7.01: “1.84x · Passes · Net Revenues could fall 32% ($15.6M) before coverage reaches 1.25x.”
Every step carries a page pin. Net Revenues opens page 112 of the FY2025 ACFR, Table 12, Pledged-Revenue Coverage, Water and Sewer System, with the 2025 row highlighted and what the agents read from it: net available revenue $48.6M, total debt service $26.4M, coverage 1.84x.
“Read by the Spreading agent · tied by the Checker.”
The FY2025 report was filed September 22, five days before its deadline. The FY2024 report was not: due September 27, 2025, it was filed November 7, 41 days late, and no failure-to-file notice was posted. The Series 2026 preliminary official statement says the city “has complied in all material respects” in the past five years. The Watcher put both pages in front of Elena a week before the bonds price.
The Watcher cites the page for every flag and never contacts an issuer or underwriter; materiality is a proposal, and the analyst decides.
The Spreading agent read 214 values from 236 pages into the water and sewer template, FY2021 to FY2025, each with its page. The Checker tied operating revenues to the statement total, Net Revenues to Table 12, debt service to Note 7 and the reserve requirement to Note 6, and re-derived coverage independently at 1.84x. Two differences are explained: investment income, where the ordinance excludes a $0.2M unrealized gain on reserve investments, and the water account count. No blocks.
The Checker can block a spread; only a person can release it, with a reason code.
The water and sewer fund's share of net pension liability rose from $38.4M to $46.9M. The Pension agent splits it: $5.9M from the plan's discount rate cut from 7.25% to 6.75%, $1.7M from investment experience, $0.9M from a larger proportionate share. The funded ratio fell from 71.8% to 66.2%. At Quillfield's own 5.75% rate, labeled as such, the liability is $68.9M, and fixed costs take 28.9% of operating revenues against a peer median of 21.4%.
“Most of the rise is the plan's discount rate cut from 7.25% to 6.75%, not a change in how Harlow Bend runs its system.”
Elena types the stress in the conversation. The Credit Writer answers with its working: with Net Revenues 15% lower, $41.3M, coverage is 1.56x, still above 1.25x; adding Series 2026's first full year of debt service, the same stress gives 1.27x. It draws the stress on the coverage scale. Below, the additional bonds test passes thinly, 1.39x on audited numbers against 1.30x; the official statement's 1.52x adds the July 2026 rate increase.
“The covenant breaks below $33.0M of Net Revenues, 32% under FY2025.”
The Credit Writer proposed keeping the score at QMS 3 and moving the outlook from Stable to Negative, every sentence cited. Elena keeps it and adds a line: raise the disclosure statement with the underwriter before pricing on October 27. She submits to Marcus Delgado, due Thursday, October 22, with 14 page pins and 212 of 214 values tied. No second signature is needed: $38.6M held is under the $50M threshold.
“Signed as Elena Marsh · maker · you will not be able to approve this view.”
Marcus opens the same credit file, with the flow of funds, the stress and the conversation, and approves with the reason code “Supported by evidence”. His name, role and the time are recorded on the audit trail with the evidence shown. On approval the note is published to the research management system.
Maker-checker on every credit view change: nobody approves their own.
The approved view moves new purchases allowed firm-wide from $40.0M at QMS 3 Stable to $25.0M at QMS 3 Negative, sent to the order management system. Hannah's $25.0M indication, $15.0M for the Municipal Bond Fund and $10.0M for the Intermediate Tax-Free Fund, was waiting for an approved view. Now it is within the limit and ready; she releases it, and the order management system checks the limit again.
The Holdings agent never changes a limit or an order; an indication waits for an approved view.
The same Harlow Bend credit, seen by five of the people who carry it: what their month looked like, and what it looks like now.
Seven agents read, spread, map, tie and explain every filing; an analyst submits, a senior analyst signs.
Picks up every annual report, official statement and event notice on the municipal disclosure feed for 4,012 obligors, matches CUSIPs to obligors, classifies notices into the 16 Rule 15c2-12 events with materiality, compares each filing date with the obligor's own continuing disclosure agreement, and checks a new official statement's compliance statement against the record.
Reads the fund statements, debt notes, GASB 68 pension and GASB 75 retiree health notes and the statistical section into the sector template, every value with its page and confidence, from tables, scanned pages and documents in Spanish.
Reads the bond resolution's definitions, flow of funds, rate covenant, additional bonds test and reserve requirement, maps the audited numbers onto each step, and tests coverage by the resolution's words.
Ties every figure to audited totals, from fund statements to the government-wide statements, notes and schedules, reads key values a second time independently, and explains every difference over the tolerance.
Reads pension and retiree health notes and ten-year schedules, splits each change in net pension liability by cause, restates at Quillfield's common discount rate and computes the fixed-cost burden.
Drafts the proposed score, outlook and research note from the spread and the other agents' findings, every sentence cited; answers the analyst's questions in the credit file and runs the stresses she asks for, showing the working.
Joins each morning's positions from portfolio accounting to obligors by CUSIP, shows exposure by fund and SMA, and computes the new-purchase limit from the limit policy once a view is approved.
Maker, checker, indications, evidence. The agents propose; a named person decides.
The credit file and every main screen carry an ask box. Ask about any obligor, filing or covenant in plain words, or tell the agents what to stress or change; every answer points to its page.
Why did Harlow Bend's coverage fall?
Coverage fell to 1.84x from 2.06x. Operation and maintenance rose 9.4% to $63.8M while the council deferred the 5.5% rate increase to July 2026; debt service rose $1.3M with Series 2022 principal. It still clears the 1.25x covenant with 32% headroom.
Stress Net Revenues −15%
With Net Revenues 15% lower, $41.3M, coverage is 1.56x, still above 1.25x. The covenant breaks below $33.0M of Net Revenues, 32% under FY2025. Adding Series 2026's first full year of debt service ($6.1M), the same stress gives 1.27x.
Is the additional bonds test right?
Yes, 1.39x on audited FY2025 Net Revenues of $48.6M over maximum annual debt service of $35.0M including the new bonds. The official statement's 1.52x adds $4.6M from the July 2026 increase; §7.03(b) allows that only with a consultant's certificate, which Appendix C provides. The house view uses the audited figure: headroom 0.09x.
Add a rule: flag pension liability up 15% or more
Proposed. Once David Okoro, head of research, approves it, any obligor whose net pension liability rises 15% or more in a year will open a review for its analyst. Today it would flag 23 obligors, including Harlow Bend (+22.1%) and Lakemoor County (+17.0%). It is in Settings, off until approved, and on the audit trail.
13 screens from the working solution, on its sample data. Pick one to see it large.
The rules, thresholds and service levels every credit is held to belong to the head of research; an analyst can propose a rule, and every change is versioned and on the audit trail.
| Setting | Default | Choose from |
|---|---|---|
| Second signature: outlook to Negative on more than | $50M held | Set by the head of research · saved as a new version |
| Second signature: score moves by | 2 or more notches | Set by the head of research · saved as a new version |
| Extraction confidence below which a person reviews | 90% | Set by the head of research · saved as a new version |
| Tie-out tolerance | $50,000 or 0.1% | Set by the head of research · saved as a new version |
| Coverage watch | Coverage within 0.15x of the covenant opens a review | On · Off |
| Late annual report | Flag the day after the deadline in each continuing disclosure agreement; notify the covering analyst | On · Off |
| Material event notice | Categories 1–4, 7, 12, 15 and 16 to the analyst the same day, with a 1-business-day clock | On · Off |
| Official statement check | Every new official statement for a held obligor checked against its filing record before pricing | On · Off |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample data · “target” = the design goal, measured in the live solution (the per-analyst figure is estimated from 4,012 obligors and 8 analysts) · “estimated” = our estimate · “typical” = published figures (SIFMA, US Municipal Bonds Statistics, October 5, 2026: $4.5 trillion outstanding, $473.8 billion issued through September 2026; SEC Rule 15c2-12, 17 CFR 240.15c2-12: notice of a listed event within 10 business days). People, firms, issuers, bonds and CUSIPs named on this page are fictional: characters and sample data in the working solution.
Municipal credit analysis, or municipal bond credit analysis, is how a muni research team judges whether an issuer can pay its bonds: spreading its ACFR, testing debt service coverage against the rate covenant in the bond resolution, reading pension and retiree health liabilities, and watching its continuing disclosure. Municipal Credit Analysis gives that team seven agents that read every filing, spread it with every value pinned to its page, map the numbers to the resolution and draft the view, for an analyst to question and submit and a senior analyst to approve.
The Pledge agent reads the resolution's definitions of Gross Revenues, Operation and Maintenance and Net Revenues and its flow of funds, then maps the audited numbers onto each step. Coverage is tested by the resolution's words, not a standard ratio: on the sample water and sewer credit, $48.6M of Net Revenues over $26.4M of debt service is 1.84x against a 1.25x rate covenant, with the headroom shown. The Checker re-derives it independently.
An additional bonds test is the gate a resolution sets before new parity bonds can be issued. The Pledge agent tests it on audited net revenues against maximum annual debt service including the new bonds, and labels any pro forma separately. On the sample credit it passes thinly, 1.39x against 1.30x, while the official statement shows 1.52x because it adds a future rate increase under a consultant's certificate.
The Disclosure Watcher reads every annual report, official statement and event notice on held obligors as it posts to the municipal disclosure feed, classifies notices into the 16 Rule 15c2-12 events with materiality, and compares each filing date with the obligor's own continuing disclosure agreement, noting whether a failure-to-file notice was posted. It also checks a new official statement's compliance statement against the filing record and raises a conflict before pricing.
The Pension agent reads the pension and retiree health notes and ten-year schedules, splits each change in net pension liability by cause (discount rate, investment experience, proportionate share) and restates every obligor at your firm's common discount rate, labeled as yours. It also computes the fixed-cost burden of debt service, pension and retiree health.
Yes. The Spreading agent reads tables and scanned pages, and reads Spanish documents with the original beside the English. When pages are unreadable it never guesses: the spread is held, and the agent drafts a clean-copy request for the analyst to send.
Yes. The agents draft and answer; they never submit, approve or publish a view. The analyst edits or declines the draft and submits it; the senior analyst for that sector approves, sends back or rejects with a reason code and can never approve their own view; a large move also needs the head of research's second signature; and limits and indications change only after approval.
The Agentic Solution Engine builds and deploys it from your requirements and documents, such as your sector templates, score rubric, research policy and a sample of past spreads, and it goes live once every quality gate has passed. Setup connects your systems, loads the coverage universe from holdings and backfills ten years of history.
Every credit question answered in minutes, each sentence pinned to its tranche and page.
1-min demoBond Covenant AnalysisEvery basket in dollars and every leakage route found, before books close.
1-min demoProspectus ReviewEvery deal that fits a fund briefed and decided before books close, every line cited.
1-min demoWe'll run Municipal Credit Analysis on a sample of your own held obligors' annual reports and bond resolutions.