Every music catalog valuation starts with a model. But before you build a model, you need to know whether the data feeding it is complete, consistent, and trustworthy. The difference between a good acquisition and a bad one is rarely the discount rate -- it is almost always the data.
This checklist covers six fundamental data-quality checks that should be performed before any valuation model is built. Each check has a clear pass/caution/flag framework so that analysts, investors, and advisors can quickly assess whether a catalog's royalty statement data is ready for modeling -- or whether gaps need to be closed first.
Why Pre-Valuation Data Quality Matters
Valuation models are only as reliable as the inputs they consume. In music catalog acquisitions, the primary inputs are royalty statements from distributors, publishers, collection societies, and sub-publishers. These statements arrive in dozens of formats, cover different periods, use different track identifiers, and report in different currencies.
If you feed incomplete or inconsistent data into a valuation model, the model will still produce a number. It will look precise. It will be wrong.
The six checks below are designed to catch the most common and most impactful data-quality problems before they propagate into a valuation.
Check 1: Statement Coverage
Statement coverage is the most fundamental check. If a major income source is missing from the data set, the model will systematically understate (or occasionally overstate) earnings.
What to check:
- Are all active distribution agreements represented in the statement data?
- Are collection society statements present for all material territories?
- How many years of history are available for each source?
- Are there any unexplained gaps in the reporting timeline?
- Has the catalog changed distributors during the history period, and if so, are transition periods covered?
The goal is to confirm that the statements collectively represent the full earnings picture of the catalog -- not just the portion that was easiest to collect.
Statement Coverage Assessment
| Check | Standard | Pass | Caution | Flag |
|---|---|---|---|---|
| Distributor coverage | All active distribution agreements represented | All distributors present | 1-2 minor distributors absent | Major distributor absent or more than 2 missing |
| PRO / territory coverage | Collection society statements for all material territories | All material territories present | Minor territory absent (under 5% earnings) | Major territory absent (US / UK / EU) |
| History depth | Minimum 3 years; 5-7 years preferred | 5+ years available | 3-4 years available | under 3 years for material distributors |
| Period continuity | No unexplained gaps over 1 consecutive period | Complete history, no gaps | 1 gap period, explained | 2+ consecutive gaps, unexplained |
A missing distributor does not necessarily mean missing income -- it may mean the income is reported through a different entity. But it does mean you need to trace the money. If a distributor representing 15% of earnings is absent from the data set, your LTM figure is wrong by at least that amount.
Check 2: Period Alignment
Royalty statements from different sources cover different time periods. Distributors may report monthly or quarterly. Collection societies often report semi-annually with significant lag. If these periods are not properly aligned before aggregation, you will double-count some earnings and miss others.
What to check:
- What is the reporting frequency for each source (monthly, quarterly, semi-annual)?
- What is the typical reporting lag for each source?
- Have statement periods been normalised to a common timeline before aggregation?
- Are accounting periods (the period the royalties were earned) distinguished from payment periods (the period the statement was received)?
- For sources with long lag times, has the most recent available period been confirmed?
Period Alignment Assessment
| Check | Standard | Pass | Caution | Flag |
|---|---|---|---|---|
| Reporting frequency mapping | Frequency documented for every source | All sources mapped and documented | 1-2 sources undocumented | Frequency unknown for material source |
| Lag identification | Reporting lag quantified per source | Lag documented, consistent | Lag estimated but not confirmed | Lag unknown or highly variable |
| Period normalisation | All data aligned to common periods before aggregation | Full normalisation applied | Partial normalisation, minor sources excluded | Raw statements aggregated without normalisation |
| Earning vs. payment period | Accounting period used, not payment date | Earning periods used consistently | Mixed usage, documented | Payment dates used as earning periods |
Period misalignment is one of the most common causes of LTM distortion. A collection society statement received in Q1 2025 may cover earnings from H1 2024. If you assign those earnings to Q1 2025, you are inflating the most recent period and deflating the period where the income was actually generated. This creates artificial growth trends that do not exist.
Check 3: Track Resolution (ISRC and Identifier Matching)
Track-level analysis requires that every line item in every statement be matched to a specific recording or composition. The primary identifier for recordings is the ISRC (International Standard Recording Code). For compositions, it is the ISWC. In practice, many statements arrive with partial or missing identifiers, requiring resolution through title matching, artist matching, or other heuristics.
What to check:
- What percentage of line items have a valid ISRC or ISWC?
- For unresolved items, what is their share of total earnings?
- Has fuzzy matching been applied, and if so, what is the confidence threshold?
- Are there duplicate ISRCs caused by reissues, remasters, or compilation appearances?
- Have sound recording and composition identifiers been separated where both rights types are present?
Track Resolution Assessment
| Check | Standard | Pass | Caution | Flag |
|---|---|---|---|---|
| ISRC / ISWC coverage | over 95% of earnings resolved to a valid identifier | over 95% resolved | 85-95% resolved | under 85% resolved by earnings |
| Unresolved earnings share | under 5% of total earnings unresolved | under 2% unresolved | 2-5% unresolved | over 5% unresolved |
| Duplicate detection | Reissues and remasters identified and grouped | Duplicates flagged and grouped | Partial detection, manual review needed | No duplicate detection applied |
| Rights type separation | Sound recording and composition IDs distinct | Cleanly separated | Mostly separated, some ambiguity | Commingled or indistinguishable |
A catalog where 20% of earnings cannot be attributed to a specific track is a catalog you cannot model at the track level. You may still be able to model it at the portfolio level, but you lose the ability to identify which tracks are driving value, which are declining, and which represent concentration risk.
Check 4: Source Classification
Royalty income arrives from many sources, but for modeling purposes, it must be classified into standard categories: streaming, downloads, physical, sync, performance (broadcast), mechanical, and other. Each category has different growth dynamics, decay profiles, and risk characteristics. Misclassification distorts every downstream analysis.
What to check:
- Has every income line been classified into a standard royalty type taxonomy?
- Are streaming sub-types distinguished (ad-supported vs. premium, audio vs. video)?
- Is sync income separated from master use fees?
- Are performance royalties distinguished from mechanical royalties?
- What percentage of income is classified as "other" or "unclassified"?
Source Classification Assessment
| Check | Standard | Pass | Caution | Flag |
|---|---|---|---|---|
| Taxonomy applied | Consistent taxonomy across all sources | Standard taxonomy applied, all sources | Taxonomy applied, minor inconsistencies | No standard taxonomy or major inconsistencies |
| Streaming sub-types | Ad-supported, premium, and video distinguished | All sub-types separated | Partial separation | All streaming income aggregated |
| Sync separation | Sync income separately identified | Sync cleanly separated | Sync partially identified | Sync commingled with other income |
| Unclassified share | under 5% of earnings unclassified | under 2% unclassified | 2-10% unclassified | over 10% unclassified |
Source classification matters because a catalog earning 80% from streaming has a fundamentally different risk profile than one earning 80% from sync. Streaming income is relatively predictable and platform-dependent. Sync income is lumpy, relationship-dependent, and harder to forecast. If you model them as a single revenue stream, your confidence intervals will be meaningless.
Check 5: Currency Handling
Music catalogs generate income in multiple currencies. Statements from GEMA arrive in euros, from JASRAC in yen, from MCPS in pounds. To build a valuation model, all income must be converted to a single reporting currency. How that conversion is done -- and when the exchange rate is applied -- materially affects the result.
What to check:
- What is the reporting currency for each source?
- Has currency conversion been applied, and if so, at what exchange rate (period average, period end, payment date)?
- Are exchange rate gains/losses separated from underlying royalty trends?
- For catalogs with significant non-USD income, has currency sensitivity been quantified?
- Are any sources already converted at the distributor level, and if so, at what rate?
Currency Handling Assessment
| Check | Standard | Pass | Caution | Flag |
|---|---|---|---|---|
| Source currency identification | Native currency documented for every source | All source currencies documented | Most documented, 1-2 assumed | Source currencies unknown for material sources |
| Conversion methodology | Consistent, documented conversion approach | Period-average rates, consistently applied | Mixed methodology, documented | No documented methodology or inconsistent rates |
| FX vs. underlying trend separation | Currency effects quantified separately | FX impact isolated and quantified | FX impact estimated | No FX separation; growth figures include currency effects |
| Pre-converted statement detection | Pre-converted sources identified | All pre-conversions identified and noted | Some pre-conversions suspected | Unknown whether sources are pre-converted |
Currency handling is especially important for catalogs with significant European or Asian income. A catalog that appears to show 5% growth year-over-year may actually show flat or declining earnings in local currency terms, with the apparent growth entirely attributable to a weakening dollar. If your model projects that growth forward, you are implicitly betting on continued dollar weakness -- which may not be your intent.
Check 6: Rights Structure Verification
The final check -- and often the most complex -- is verifying that the rights structure is correctly represented in the data. Music rights are split across multiple dimensions: sound recording vs. composition, writer share vs. publisher share, territory-specific sub-publishing agreements, and co-ownership splits. If the data does not correctly reflect the seller's actual ownership share, the model will overstate or understate the acquirable income.
What to check:
- Is the seller's ownership share documented for each track and each rights type?
- Are co-writer and co-publisher splits reflected in the data?
- Are territory-specific sub-publishing agreements accounted for?
- Does the data distinguish between gross royalties and the seller's net share?
- Are there any reversion clauses or term limits that affect future earnings?
- Has the chain of title been verified against registered data at PROs?
Rights Structure Assessment
| Check | Standard | Pass | Caution | Flag |
|---|---|---|---|---|
| Ownership share documentation | Per-track, per-rights-type ownership documented | Complete ownership schedule provided | Ownership provided at catalog level, not per-track | Ownership unclear or undocumented |
| Co-ownership splits | All co-writer / co-publisher splits reflected | All splits documented and applied | Most splits documented, minor gaps | Significant co-ownership splits missing or disputed |
| Gross vs. net distinction | Clear distinction between gross royalty and seller's net | Net share consistently derived | Gross/net distinction inconsistent across sources | No distinction; unclear what share is acquirable |
| Reversion / term limits | All reversion clauses and term limits identified | All terms documented, no near-term reversions | Some terms documented, minor reversion risk | Reversion clauses not reviewed or near-term reversion present |
| Chain of title verification | Registered ownership verified at PROs | Verified against PRO registrations | Partially verified | No verification performed |
Rights structure errors are among the most expensive mistakes in catalog acquisitions. If the data shows a track generating $100,000 per year but the seller only owns 50% of the publishing, the acquirable income is $50,000. If the model uses the gross figure, the buyer will overpay by a factor of two on that track. Across a large catalog, even small, systematic rights-share errors compound into material valuation distortions.
The Complete Pre-Valuation Checklist
The table below consolidates all six checks into a single reference. For each item, the "If failed" column describes the specific impact on the valuation model.
| Category | Check | If failed: impact on modeling |
|---|---|---|
| Coverage | All distributors present | LTM understated; missing income not projected |
| Coverage | PRO / territory coverage complete | Territory-level analysis unreliable; geographic risk masked |
| Coverage | 5+ years of history | Decay curve estimation unreliable; trend analysis weakened |
| Alignment | Periods normalised to common timeline | LTM distorted; artificial growth or decline trends |
| Alignment | Earning period used (not payment date) | Seasonality misattributed; lag creates phantom trends |
| Resolution | 95%+ earnings resolved to ISRC / ISWC | Track-level analysis impossible; concentration risk hidden |
| Resolution | Duplicates detected and grouped | Track counts inflated; per-track metrics diluted |
| Classification | Standard taxonomy applied | Source mix analysis unreliable; decay assumptions misapplied |
| Classification | Sync income separated | Forecast volatility understated; lumpy income smoothed |
| Classification | under 5% unclassified income | Material earnings in unknown category; cannot model by source |
| Currency | Consistent conversion methodology | Growth rates include FX noise; trend direction may be wrong |
| Currency | FX impact isolated | Cannot distinguish organic growth from currency movement |
| Rights | Per-track ownership documented | Acquirable income overstated or understated |
| Rights | Gross vs. net distinguished | Valuation based on non-acquirable income; systematic overpayment risk |
| Rights | Reversion clauses reviewed | Model projects income beyond ownership period; terminal value inflated |
What to Do When Checks Fail
Failed checks do not necessarily mean you should walk away from a deal. They mean you need to either fix the data or adjust the model to reflect the uncertainty.
For coverage gaps:
- Request the missing statements from the seller or their administrator
- If statements are unavailable, estimate the missing income using market benchmarks and apply a confidence discount
- Document the gap and its estimated impact in the investment memo
For alignment problems:
- Rebuild the timeline using accounting periods, not payment dates
- If accounting periods are unavailable, use the distributor's standard lag profile to back-calculate the earning period
- Flag any period where normalisation required assumptions
For resolution failures:
- Apply secondary matching (title + artist + duration) to unresolved items
- If resolution remains below 85%, model only at the portfolio level and disclose the limitation
- Quantify the unresolved earnings as a range, not a point estimate
For classification issues:
- Reclassify using source-level identifiers (distributor type, statement codes)
- If sync cannot be separated, model all income with the higher-volatility assumption
- Never model unclassified income as if it were streaming
For currency distortions:
- Restate all historical income in the reporting currency using period-average exchange rates
- Run the model with constant-currency and actual-currency scenarios
- If the delta exceeds 3%, include FX sensitivity analysis in the output
For rights structure problems:
- Request a complete ownership schedule from the seller's legal team
- Cross-reference with PRO registration data
- If co-ownership splits are unresolved, use the most conservative (lowest) ownership assumption until confirmed
Conclusion
Data quality is not a box-checking exercise -- it is the foundation of every valuation decision. A clean data set does not guarantee a good investment, but a dirty data set almost guarantees a bad model.
The six checks in this article are not exhaustive, but they cover the problems that most frequently cause material valuation errors in music catalog acquisitions. Run them before you build your model. Run them again when new data arrives. And if the data fails multiple checks, fix the data before you fix the discount rate.
The model is only as good as the data underneath it. Make sure the data is ready before you ask the model to perform.