Appendix g — Data Card — MSRB EMMA (Municipal securities disclosures and trades)

Provider & what it is

The MSRB (Municipal Securities Rulemaking Board) runs EMMA — the Electronic Municipal Market Access system, the official public repository for the U.S. municipal bond market. When a state, city, county, school district, water authority, or other government body borrows by issuing a “muni,” its disclosure and trading record lives on EMMA. It is the muni market’s EDGAR.

EMMA publishes two things you will use:

Disclosures — the official statement (the muni’s prospectus: the bonds, issuer, security, use of proceeds), plus ongoing continuing-disclosure filings: audited financials, material-event notices (rating changes, defaults, calls). The accounting side of a muni.
Trades — the secondary-market trade record: every reported municipal transaction with price, par amount, and time, via the MSRB’s Real-Time Transaction Reporting System (RTRS). This is what a muni actually traded at — and, against a benchmark, what yield the market demanded of that borrower.

The reveal that motivates the muni thread: a corporate bond’s yield reflects one firm’s credit risk, but a muni yield is a window onto how expensively a government can borrow — a “cost of credit” that varies across issuers and time in ways that are live research (Prof. Gao’s Rainbow of Credits). EMMA is where that price lives, in public, for free.

Coverage

Trades: real-time municipal trade reporting from the mid-2000s (RTRS launched 2005; earlier trade data exists in less granular form). Coverage is the reported secondary market, which for munis is thin and episodic — many bonds trade rarely. [CHECK] exact RTRS start date and pre-2005 trade availability.
Disclosures: EMMA became the centralized disclosure portal in 2009; official statements and continuing disclosures are broadly available from then, with some earlier documents. [CHECK] earliest reliable disclosure year.
Universe: essentially the whole U.S. muni market — roughly a million-plus distinct outstanding CUSIPs across tens of thousands of issuers (states, localities, authorities, school and water districts). Far more issuers than the corporate market, and far less liquid per issue.

Key identifiers

CUSIP (9-character) identifies a specific muni issue/maturity. A single bond deal is sold as a serial — many CUSIPs, one per maturity — so one borrowing event spans dozens of CUSIPs.
Issuer — the governmental borrower; the link from CUSIP to issuer (and from issuer to state/sector) is the join you need for any cross-issuer study, and it is not always clean.
Trade timestamp, price, par traded, and a trade-type flag (customer-buy, customer-sell, inter-dealer) on the trade side.
There is no single tidy primary key spanning disclosures and trades except the CUSIP, and disclosures often attach at the deal/official-statement level rather than per-CUSIP — reconciling the two is real work.

Access path

EMMA website (https://emma.msrb.org) is the primary public access point: search by issuer or CUSIP, read official statements, view per-bond trade history. Free, no account for browsing.
Bulk access is the catch. MSRB offers no simple open public API for bulk trade history the way Treasury or FDIC do. Bulk historical trade and reference datasets come through MSRB subscription/data products (some mirrored in academic services); the free site is built for one-bond lookups, not large downloads. A real cross-issuer panel uses a licensed muni dataset (MSRB historical files or a vendor feed) on GMU infrastructure (CONVENTIONS §5). [CHECK] current MSRB API/bulk-product names and terms, and whether the camp container mirrors a muni trade file.

# EMMA has no clean free bulk API; the disciplined pattern for the camp is to treat
# the public site as a per-CUSIP lookup and use a LICENSED muni trade file (on Hopper /
# the container's read-only mount) for any panel. Pseudocode for the per-bond route:
#
#   1. Search emma.msrb.org by issuer  ->  list of CUSIPs for the deal
#   2. For a CUSIP, pull its trade history and official statement (manually or via the
#      documented endpoint for your pinned vintage)  [CHECK: endpoint + terms]
#   3. Aggregate trades to an issue-level or issuer-level yield/price series
#
# Never scrape the public site for a large panel; use the licensed file instead.

License & note

Free to browse; restricted for bulk. The EMMA public website is free and open to read — official statements, disclosures, and per-bond trade history cost nothing to view, and the disclosures themselves are public records. But bulk redistribution of MSRB trade data is governed by MSRB terms, and large historical datasets sit behind MSRB subscription/data products. So: quote what you read on EMMA freely with a citation; build any panel you regress on from a licensed source that stays on GMU infrastructure, and pin its vintage. [CHECK] exact MSRB data-product license terms before redistributing anything beyond a screenshot or a cited figure.

Gotchas

Munis barely trade. Like corporate bonds (see treasury-finra.md) but worse — many CUSIPs go weeks or months without a print. A “last price” is often stale; sparse, irregular trading is the data, not a glitch. Yields must be inferred carefully, often from infrequent customer trades.
One deal, many CUSIPs. A serial issue fragments one borrowing into dozens of maturities/CUSIPs. To study the issuer’s cost of borrowing you must aggregate across CUSIPs within a deal — getting this wrong double-counts or mis-weights.
No clean free bulk API. Unlike Treasury/FDIC/HMDA, EMMA is built for one-bond lookups. A cross-issuer panel needs a licensed file; do not plan a capstone around scraping the public site.
Tax status confounds yield. Muni interest is often federally (and sometimes state) tax-exempt, so muni yields are not comparable to taxable Treasury/corporate yields without a tax adjustment. Comparing a raw muni yield to a Treasury yield is a beginner error.
Issuer mapping is messy. CUSIP-to-issuer-to-sector linkage is not always clean; the same governmental borrower can appear under several issuer descriptions.

First 10 rows (illustrative schema sketch — not real values)

Trade records (per CUSIP), shape only — note the irregular timestamps (illiquidity):

cusip	trade_datetime	trade_type	price	yield	par_traded
64971QXX4	2026-04-02 11:14:08	customer_buy	101.235	3.41	50000
64971QXX4	2026-02-19 14:52:30	inter_dealer	100.980	3.47	25000
64971QXX4	2025-11-30 10:07:55	customer_sell	100.410	3.55	100000
…	…	…	…	…	…

Issuer / disclosure reference (per deal), shape only:

issuer_name	state	sector	dated_date	par_amount	official_statement_id
Example County School District	VA	Education	2025-09-15	48500000	OS-EX-000123
Sample City Water Authority	MD	Utility	2024-06-01	22000000	OS-EX-000087
…	…	…	…	…	…

Issuer-level cost-of-borrowing series you would build, shape only:

issuer_name	state	year	avg_offering_yield	n_deals
Example County School District	VA	2025	3.42	2
Sample City Water Authority	MD	2024	3.18	1
…	…	…	…	…

Which chapter / lab / capstone uses it

Mentor Session 5 — “Anatomy of a JF paper,” tied to the Rainbow of Credits municipal-borrowing working paper (Gao, Liu & Wang, AEA 2025; target JF). The session walks students through a live paper built on muni issuance and secondary-market trade records “of the kind assembled from the MSRB,” and EMMA is the public face of exactly that data. This card is the reference for where that data lives and why it is hard.
The muni-borrowing research thread more broadly (revisited in Mentor Session 7’s project pipeline): municipal cost-of-credit as an outcome, with the institutional features above (illiquidity, serial CUSIPs, tax status) as the texture that makes it a real research setting rather than a clean toy.
It sits with the other fixed-income public sources — treasury-finra.md (the risk-free curve + corporate tape) — to give students the full credit-market picture: government risk-free, corporate credit, and municipal credit, each with its own messiness, each free to read and licensed to bulk.

# Data Card — MSRB EMMA (Municipal securities disclosures and trades) ## Provider & what it is The **MSRB** (Municipal Securities Rulemaking Board) runs **EMMA** — the Electronic Municipal Market Access system, the official public repository for the U.S. **municipal bond** market. When a state, city, county, school district, water authority, or other government body borrows by issuing a "muni," its disclosure and trading record lives on EMMA. It is the muni market's EDGAR. EMMA publishes two things you will use: - **Disclosures** — the **official statement** (the muni's prospectus: the bonds, issuer, security, use of proceeds), plus ongoing **continuing-disclosure** filings: audited financials, material-event notices (rating changes, defaults, calls). The accounting side of a muni. - **Trades** — the secondary-market **trade record**: every reported municipal transaction with price, par amount, and time, via the MSRB's Real-Time Transaction Reporting System (RTRS). This is what a muni actually traded at — and, against a benchmark, what *yield* the market demanded of that borrower. The reveal that motivates the muni thread: a corporate bond's yield reflects one firm's credit risk, but a muni yield is a window onto **how expensively a government can borrow** — a "cost of credit" that varies across issuers and time in ways that are live research (Prof. Gao's *Rainbow of Credits*). EMMA is where that price lives, in public, for free. ## Coverage - **Trades:** real-time municipal trade reporting from the mid-2000s (RTRS launched 2005; earlier trade data exists in less granular form). Coverage is the reported secondary market, which for munis is **thin and episodic** — many bonds trade rarely. `[CHECK]` exact RTRS start date and pre-2005 trade availability. - **Disclosures:** EMMA became the centralized disclosure portal in **2009**; official statements and continuing disclosures are broadly available from then, with some earlier documents. `[CHECK]` earliest reliable disclosure year. - **Universe:** essentially the whole U.S. muni market — roughly a million-plus distinct outstanding CUSIPs across tens of thousands of issuers (states, localities, authorities, school and water districts). Far more *issuers* than the corporate market, and far less liquid per issue. ## Key identifiers - **CUSIP** (9-character) identifies a specific muni *issue/maturity*. A single bond deal is sold as a *serial* — many CUSIPs, one per maturity — so one borrowing event spans dozens of CUSIPs. - **Issuer** — the governmental borrower; the link from CUSIP to issuer (and from issuer to state/sector) is the join you need for any cross-issuer study, and it is not always clean. - **Trade timestamp**, **price**, **par traded**, and a **trade-type flag** (customer-buy, customer-sell, inter-dealer) on the trade side. - There is no single tidy primary key spanning disclosures and trades except the CUSIP, and disclosures often attach at the *deal/official-statement* level rather than per-CUSIP — reconciling the two is real work. ## Access path - **EMMA website** (`https://emma.msrb.org`) is the primary public access point: search by issuer or CUSIP, read official statements, view per-bond trade history. **Free, no account** for browsing. - **Bulk access is the catch.** MSRB offers **no** simple open public API for bulk trade history the way Treasury or FDIC do. Bulk historical trade and reference datasets come through **MSRB subscription/data products** (some mirrored in academic services); the free site is built for one-bond lookups, not large downloads. A real cross-issuer panel uses a licensed muni dataset (MSRB historical files or a vendor feed) on GMU infrastructure (CONVENTIONS §5). `[CHECK]` current MSRB API/bulk-product names and terms, and whether the camp container mirrors a muni trade file. ```python # EMMA has no clean free bulk API; the disciplined pattern for the camp is to treat # the public site as a per-CUSIP lookup and use a LICENSED muni trade file (on Hopper / # the container's read-only mount) for any panel. Pseudocode for the per-bond route: # # 1. Search emma.msrb.org by issuer -> list of CUSIPs for the deal # 2. For a CUSIP, pull its trade history and official statement (manually or via the # documented endpoint for your pinned vintage) [CHECK: endpoint + terms] # 3. Aggregate trades to an issue-level or issuer-level yield/price series # # Never scrape the public site for a large panel; use the licensed file instead. ``` ## License & note **Free to browse; restricted for bulk.** The EMMA *public website* is free and open to read — official statements, disclosures, and per-bond trade history cost nothing to view, and the disclosures themselves are public records. But **bulk redistribution of MSRB trade data is governed by MSRB terms**, and large historical datasets sit behind MSRB subscription/data products. So: quote what you read on EMMA freely with a citation; build any *panel* you regress on from a licensed source that stays on GMU infrastructure, and pin its vintage. `[CHECK]` exact MSRB data-product license terms before redistributing anything beyond a screenshot or a cited figure. ## Gotchas - **Munis barely trade.** Like corporate bonds (see `treasury-finra.md`) but worse — many CUSIPs go weeks or months without a print. A "last price" is often stale; sparse, irregular trading is the data, not a glitch. Yields must be inferred carefully, often from infrequent customer trades. - **One deal, many CUSIPs.** A serial issue fragments one borrowing into dozens of maturities/CUSIPs. To study the *issuer's* cost of borrowing you must aggregate across CUSIPs within a deal — getting this wrong double-counts or mis-weights. - **No clean free bulk API.** Unlike Treasury/FDIC/HMDA, EMMA is built for one-bond lookups. A cross-issuer panel needs a licensed file; do not plan a capstone around scraping the public site. - **Tax status confounds yield.** Muni interest is often federally (and sometimes state) tax-exempt, so muni yields are *not* comparable to taxable Treasury/corporate yields without a tax adjustment. Comparing a raw muni yield to a Treasury yield is a beginner error. - **Issuer mapping is messy.** CUSIP-to-issuer-to-sector linkage is not always clean; the same governmental borrower can appear under several issuer descriptions. ## First 10 rows (illustrative schema sketch — not real values) *Trade records (per CUSIP), shape only — note the irregular timestamps (illiquidity):* | cusip | trade_datetime | trade_type | price | yield | par_traded | |---|---|---|---|---|---| | 64971QXX4 | 2026-04-02 11:14:08 | customer_buy | 101.235 | 3.41 | 50000 | | 64971QXX4 | 2026-02-19 14:52:30 | inter_dealer | 100.980 | 3.47 | 25000 | | 64971QXX4 | 2025-11-30 10:07:55 | customer_sell | 100.410 | 3.55 | 100000 | | … | … | … | … | … | … | *Issuer / disclosure reference (per deal), shape only:* | issuer_name | state | sector | dated_date | par_amount | official_statement_id | |---|---|---|---|---|---| | Example County School District | VA | Education | 2025-09-15 | 48500000 | OS-EX-000123 | | Sample City Water Authority | MD | Utility | 2024-06-01 | 22000000 | OS-EX-000087 | | … | … | … | … | … | … | *Issuer-level cost-of-borrowing series you would build, shape only:* | issuer_name | state | year | avg_offering_yield | n_deals | |---|---|---|---|---| | Example County School District | VA | 2025 | 3.42 | 2 | | Sample City Water Authority | MD | 2024 | 3.18 | 1 | | … | … | … | … | … | ## Which chapter / lab / capstone uses it - **Mentor Session 5 — "Anatomy of a JF paper,"** tied to the **Rainbow of Credits** municipal-borrowing working paper (Gao, Liu & Wang, AEA 2025; target *JF*). The session walks students through a live paper built on muni issuance and secondary-market trade records "of the kind assembled from the MSRB," and EMMA is the public face of exactly that data. This card is the reference for *where that data lives and why it is hard*. - **The muni-borrowing research thread** more broadly (revisited in Mentor Session 7's project pipeline): municipal cost-of-credit as an outcome, with the institutional features above (illiquidity, serial CUSIPs, tax status) as the texture that makes it a real research setting rather than a clean toy. - It sits with the other fixed-income public sources — `treasury-finra.md` (the risk-free curve + corporate tape) — to give students the full credit-market picture: government risk-free, corporate credit, and municipal credit, each with its own messiness, each free to *read* and licensed to *bulk*.