There is a phrase used by almost every Brazilian rightsholder I work with, and it is the single most expensive phrase in the entire music rights business.
The phrase is “os royalties de fora.”
Sometimes it is “international royalties.” Sometimes it is “the money from abroad.” The translation does not matter because the damage is identical in every language. The phrase treats as a single thing something that is, in reality, at least one hundred and ninety-five separate operations, with different timing, different legal regimes, different data requirements, different political risks, and different failure modes. A Brazilian label or publisher signing away a percentage of “international” to some administrator has rarely asked which of those one hundred and ninety-five operations the administrator actually does well, which ones they do badly, and which ones they have never touched. They sign, they wait, and they receive whatever arrives. And then, eventually, they convince themselves that whatever arrived is what there was.
You might reasonably ask whether this really matters. If the money arrives, and it matches what the administrator reports is due, where is the problem? The problem, as I will try to show, is that the administrator is reporting only what came in, not what was owed. Those are two very different numbers. One of them is verifiable from a royalty statement. The other is only visible to someone who has gone out to look at the machinery upstream. Most rightsholders have never seen the machinery. Most administrators, frankly, have not seen all of it either.
This is a convenient belief for a lot of people. It is not convenient for the rightsholder.
In 2024, CISAC’s member societies collected 13.97 billion euros on behalf of creators, across 228 societies in 111 countries. Music alone accounted for 12.59 billion euros of that figure, a 7.2% increase over the year before. These are the public numbers, and they are reported in a way that quietly does its own damage. One report. One currency. One growth percentage. The reader ends up with a mental model of a single global pool of money, slowly filling up year on year, from which creators are paid in proportion to how much their songs get played. That mental model is a fiction. There is no pool. There is a network of national operations, most of them opaque, several of them dysfunctional, and a few of them actively hostile to the foreign rightsholders they claim to serve through reciprocal agreements.
What follows is a partial tour of that network, grouped by the kind of failure a Brazilian rightsholder should expect to encounter in each type. It is not exhaustive. It is meant to be enough to end, permanently, the habit of treating “o exterior” as one thing.
Timing pathologies: the money exists, but not yet, and not when you think
Performance royalties in most major territories operate on a quarterly distribution cycle, and mechanical royalties are not far behind. The theoretical minimum lag between a song being played and the composer receiving payment is usually six months. The practical average, for most foreign rightsholders, is twelve to eighteen months. This is not fraud. It is architecture.
Let us start with our neighbours. SADAIC, Argentina’s performance rights society, distributes on a schedule tied to its own internal accounting calendar, not to the convenience of any foreign rightsholder. For a Brazilian composer whose song was played on Argentine radio in January, the pathway is: the radio station pays SADAIC, SADAIC processes and distributes to its members and to reciprocal societies, a Brazilian society such as UBC or ABRAMUS receives the portion belonging to its affiliated rightsholders, the Brazilian society processes and distributes to the publisher, and the publisher finally distributes to the composer. Each step adds a lag. Each step applies a fee. Each step introduces a point at which data can be misread, a work unmatched, a share disputed. The January play can easily become a December payment, and that is when everything in the chain is working as designed. Which, in Argentina, it frequently is not.
And that is only the performance side. If the same Brazilian song is streamed on Spotify in Argentina rather than played on the radio, the mechanical royalty takes a completely different path. It does not go to SADAIC at all. It goes through Latin Autor, the regional mechanical collection body for Latin America excluding Brazil, which acts in that role much as the MLC does in the United States. Most Brazilian publishers I work with have never heard of Latin Autor. They assume “Argentina” means “SADAIC” and stop there. They are missing half the picture by definition.
When the chain is not working as designed, the lag stops being measured in months and starts being measured in years. Payments arriving three, four, or five years after the usage that generated them are not hypothetical. They are a regular feature of collections from several mid-tier territories, and they carry a second, less obvious problem. By the time the payment arrives, nobody is auditing it. The statement is filed, the money is booked, and the specific usage period that generated it is so old that verifying the number would require archaeological work nobody will pay for. This is one of the quieter ways money gets lost: not stolen, not withheld, just aged past the point where anyone still cares to check.
For a catalogue being valued for acquisition, this matters enormously. If someone offers you a valuation multiple based on “international revenue” as a line item, you should be asking which territories contribute to that line, what the average age of those payments is, and whether the recent growth reflects real growth or simply the arrival of backlogged distributions from three years ago.
Political and regulatory risk: the rules change under you
In March 2024, the Milei government in Argentina issued a deregulation decree that fundamentally altered how authorship royalties can be collected in the country. The decree removed the historical obligation for artists to channel collections through state-authorised societies such as SADAIC, SAGAI, DAC, and AADI. SADAIC filed an administrative challenge and a federal lawsuit arguing the decree was unconstitutional. At the time of writing, the legal and operational fallout continues, and nobody, including the Argentine government, has a clear picture of how the reciprocal collection pathway behaves while the litigation is pending. I say this not as a throwaway observation but as an honest admission: at this moment in time, I do not know with certainty how money earned by a Brazilian song on an Argentine platform is flowing. Nobody I have spoken to in the reciprocal chain does either. That should concern every Brazilian publisher reading this.
I do not care, for this piece, whether the decree is good or bad policy. The relevant question for a Brazilian publisher is simpler and more uncomfortable. Do you know what the status of the reciprocal agreement is between SADAIC and your Brazilian society right now, this quarter? Do you know whether a decision issued in Buenos Aires in the last six months has altered the payment pathway your catalogue depended on last year? If the answer to either of those questions is no, you are not doing international royalty administration. You are receiving whatever arrives, and calling it administration.
The United States is the more consequential case, because every Brazilian catalogue with any material streaming income has real exposure there, and almost none of that exposure is understood by the rightsholders themselves. Most countries collect performance and mechanical rights through a single combined society, or through two closely coordinated ones. The United States does not. It treats them as fundamentally different rights, collected by different entities, under different rules, and with different problems. A foreign rightsholder is exposed to all of them at once, whether they know it or not.
To keep this readable, let me separate the two rights properly, because the rest of the piece depends on the distinction. The first is the performance right, which pays the songwriter when the song is publicly performed, broadcast, or streamed. In the US, this is collected by ASCAP, BMI, SESAC, or GMR, four competing PROs. A Brazilian songwriter reaches one of these four indirectly, through the reciprocal agreement between their US PRO counterpart and their Brazilian society. The second is the mechanical right, which pays the songwriter when the song is reproduced, including by being streamed on demand. For interactive streaming in the US, this is collected by the Mechanical Licensing Collective, a body created by the Music Modernisation Act in 2018. These are two different pipelines, with two different registrations, two different sets of data requirements, and two different paths home to a Brazilian account. A single play of a Brazilian song on Spotify in Chicago generates both rights at the same time. The performance royalty travels through ASCAP or BMI back to the Brazilian PRO. The mechanical royalty travels through the MLC. Two separate payments, two separate timelines, two separate registrations, from one stream. If the registration is correct in one pipeline and wrong in the other, income leaks from the broken pipeline without anyone ever noticing, because the statement you eventually receive will only ever show you what arrived.
Importantly, the two pipelines do not perform equally well, and the rest of this section is specifically about the mechanical side. The performance side in the US has its own dysfunctions, its own black-box practices, and its own opacity issues which deserve a separate piece. What follows is about mechanicals only.
The MLC deserves a specific note, because it is the part of the US mechanical system that actually works, and the comparison it sets up for anyone operating in Brazil is instructive.
The MLC was designed, among other things, to recover the mechanical royalties that US DSPs had accumulated but never paid out during the streaming era before 2021. When the MLC launched, DSPs transferred roughly 427 million dollars in historical unmatched royalties, covering streaming usage all the way back to 2007, into the MLC’s custody. Since May 2022, the MLC has been matching those historical royalties to rightsholders and distributing them, with interest accrued from the time the funds were received. Rightsholders registered with the MLC have received payments for streams that happened ten and fifteen years ago. This is the opposite of what most royalty systems do with old money. Most systems let it disappear. The MLC went and got it back.
Alongside the historical recovery, the MLC publishes detailed figures on what is still unmatched and unclaimed, broken down by category. In its 2023 annual recap, the MLC reported approximately 209.7 million dollars in unmatched royalties, another 164.2 million dollars in matched-but-unclaimed royalties where the rightsholder has been identified but has not yet claimed their share, and another 23.1 million dollars in held royalties subject to disputes or legal review. Each of these numbers is publicly reported, broken down, and attached to a searchable database where rightsholders can hunt for their own works. Independent creator coalitions have used this published data to write to the US Congress challenging the MLC’s match rates, holding period, and distribution methodology. That public back-and-forth resulted in the MLC extending its initial holding period from the statutory minimum of three years to six, and delaying its first market share distribution until early 2027. As of today, no market share redistribution of unclaimed US mechanicals has ever happened.
Stop and read that again. Within the US mechanical system, which not long ago was widely described as a byword for industry dysfunction, the body now responsible for those royalties has clawed back fifteen years of historical money, publishes its black box figures in detail, lets rightsholders search for their own unclaimed shares, has been pressured by independent advocacy into holding unclaimed money for twice the legally required period, and has never redistributed a dollar by market share. The US PRO side has not done any of these things. The two halves of the US system are operating in completely different worlds.
The Brazilian mechanical system produces retention data at the individual catalogue level, and Backoffice publishes that data, which in some ways is more granular than the MLC’s aggregated black-box figures. What the Brazilian system does not produce is the aggregated public reporting, the regulator-facing accountability, the searchable claiming portals, and the organised advocacy infrastructure that in the US has produced outcomes like a six-year hold and a 2027 distribution date. The information exists here. The political structure around that information does not.
The structural risk for a Brazilian rightsholder collecting mechanicals from the US is, therefore, in the short term, smaller than it is in most other territories. If your Brazilian catalogue is earning mechanicals in the US and the registration is wrong, the money is not going to be redistributed to someone else tomorrow. It is sitting in a pool with your name on the registered work, waiting for you to claim it, with interest accruing. What you need is for someone to know how to look, how to claim, and how to reconcile. The data is there. The US mechanical system, more than most, rewards the rightsholder who actually shows up.
Structural oddities: the rules were not written for you
Japan’s JASRAC distributed 142.3 billion yen in fiscal 2024, a record, with royalties going to over 430,000 creators and 52,653 music publishers via 112 foreign collective management organisations. These numbers sound welcoming. The welcoming ends the moment you read the affiliation requirements for foreign rightsholders. To become a direct JASRAC affiliate, a music publisher must be a legal person that can submit a corporate registry certificate, must maintain a bank account within Japan to receive royalties, and must be able to communicate in Japanese or appoint a contact who can.
These are not hostile requirements. They are domestic requirements that happen to make direct affiliation impractical for almost any Brazilian publisher. What this means in practice is that Brazilian rightsholders reach JASRAC only through the reciprocal chain: Brazilian society, to its counterpart society in a country where direct affiliation is easier, to JASRAC. Every step in that chain is a place where data can degrade, where a work registration can fail to match, and where a retention or administrative fee can be applied.
The black box, globally
The industry term for royalties that have been collected but cannot be matched to a specific rightsholder is the “black box.” It is the most important concept in international royalty administration, and it is the concept least often explained to the independent rightsholders whose income is at stake.
Industry estimates have long put the proportion of global streaming song royalties that fails to reach the intended rightsholder at more than 20%, whether because the work was never matched, because the match was contested, or because the payment was delayed long enough to fall out of the normal accounting cycle. One out of every five dollars a song earns on streaming globally does not reach the person who wrote the song. It sits in escrow somewhere, waiting to be matched to a name. In most territories, when the holding period expires, the money is redistributed on a market share basis to the largest publishers in that territory, and this rule applies to both performance and mechanical pools. This is the standard rule across Europe, across most of Asia, and across most of Latin America. The US mechanical pool, as described above, is a partial exception because advocacy has pushed the MLC to hold longer and because market share redistribution there has not yet occurred. The US performance pool, run by the PROs, is not exempt and never has been. Everywhere else, redistribution is actively occurring, on a continuous schedule, year after year. Music Business Worldwide has described this pattern as a “reverse Robin Hood” system, and the description is exact: the smaller the rightsholder, the higher the probability their money ends up funding someone larger. Read that again. The money your song earned, which was never paid to you because the paperwork did not match, is being handed over every quarter to the publishers who already have the most money in each territory. This is not a bug. This is written into the distribution rules. And the further down the long tail you are, as an independent Brazilian publisher, the more mathematically certain it is that the system is engineered to transfer your income upward.
The Brazilian situation adds a further twist, and here the separation between royalty types becomes critical. In Brazil, public performance rights are collected by ECAD on behalf of the seven associations that make it up. Domestic mechanical rights are handled mainly by UBEM and Abramus Digital, with some smaller players such as ONErpm and SBACEM also operating in the space.
For international income, the two tracks diverge sharply, and the divergence is where most Brazilian publishers lose the plot. On the performance side, a reciprocal network exists on paper. Almost every major society has signed reciprocal agreements with almost every other. The pipes are drawn on the map. Whether water actually moves through any given pipe is a separate question, and frequently the answer is no. I have seen Brazilian catalogues with clear US performance income receive a payment from a major US PRO once in five years. Not delayed. Not reduced. Once. The network exists, the agreement exists, the catalogue is generating income, and the money is not arriving. Nobody picks up the phone. Nobody flags the gap. The Brazilian publisher looks at their statement, sees a small line for the territory, and assumes that small line is what was earned. It is not. It is what was successfully extracted from a system that quietly fails to send most of it. On the mechanical side, the picture is different in a particular way, and whether a Brazilian publisher has any pathway at all depends on which Brazilian entity administers their mechanical rights. Of the two main players, one has built out reciprocal arrangements with foreign mechanical collection bodies, and its affiliated publishers can, at least in principle, see foreign mechanicals flow back through that structure. The other has not. Its affiliated publishers, collectively, are the largest group of Brazilian music publishers with no default international mechanical pathway at all.
If you are tracking the contrast carefully, you should notice that the failures are different in kind. The performance side fails by appearing to work and not working. The mechanical side, for half of Brazilian publishers, fails by not even pretending. Both produce the same bottom-line result, which is that the rightful Brazilian rightsholder does not see their international income. But the diagnosis is different, and the remedy is different. You cannot fix a performance-side gap by setting up a new pathway, because the pathway already exists. You have to go in and force it to actually carry water. That is a different kind of work, and a different kind of audit.
This is not a failure of the entity without the reciprocal arrangements. Its mandate is domestic. The failure is the failure of any publisher affiliated with it to understand that “os royalties de fora” on the mechanical side is, for them, not a line that is underperforming. It is a line that does not exist, and cannot exist, unless someone goes out and constructs the international collection pathway manually, catalogue by catalogue, territory by territory. Setting up direct accounts with foreign mechanical collectors, registering works under the correct identifiers, validating the incoming statements, and disputing the errors, none of this happens by default. For a substantial portion of the Brazilian publishing industry, it does not happen at all. The statement arrives, the mechanical international line is empty or close to it, and nobody investigates why, because nobody is looking for what was never there.
Beyond that first gap, a second problem applies to whatever international income a Brazilian rightsholder does collect. Once their money leaves the foreign society, they depend on the accuracy and diligence of at least one intermediary society, and often one or more private administrators in between. To be fair to the Brazilian system, the raw data is not entirely hidden. On the mechanical side, Backoffice, the company that sits between the DSPs and the domestic mechanical collectors, publishes a list of retained amounts. It is not a well-designed file, it is not easy to work with, but it exists. On the performance side, each of the seven ECAD associations offers some way to interrogate retained amounts through their own systems, with varying degrees of difficulty and completeness. The tools are there. Nobody would call them friendly.
The problem is not that the information is hidden. The problem is that using it requires specialist work most rightsholders do not know how to do, do not have the time to do, and cannot tell whether their administrator is doing. If a work registration is wrong, a share is disputed, a CAE number is mismatched, or a foreign statement is misread, the money will show up in one of these retention files. Whether anyone notices, reconciles it, and claims it back before the holding period expires is a different question. For most independent Brazilian publishers, the honest answer is no. The file exists, the data exists, the money is listed by name, and nobody is looking at it. Then the clock runs out, and the money is redistributed to someone else.
What this means for valuation
Let me return to where I started. I opened this piece by claiming that “os royalties de fora” is the single most expensive phrase in the Brazilian music rights business. I meant that literally, and the valuation argument is where the expense becomes visible.
The practical consequence of all of the above is that catalogue valuations built on “international revenue” as a single line item are, almost without exception, wrong. Not slightly wrong. Structurally wrong. And the people doing the valuations often know this, and quote the number anyway, because a single number is easier to multiply than a map.
A catalogue earning R$ 100,000 per year from “international royalties” is not the same asset as another catalogue earning R$ 100,000 per year from “international royalties.” The first might derive most of its income from Germany, France, and the Nordic territories, where collection infrastructure is strong, match rates are high, and payments are reasonably current. The second might derive the same total from the United States and a scattering of mid-tier Latin American and Southern European markets, where the US number obscures leakage across separate performance and mechanical pipelines, Argentina’s regulatory ground is actively moving, and the reciprocal chain introduces several points of loss before a single real sits in a Brazilian account. These two catalogues have very different values, very different risk profiles, and very different futures. Nothing in the way of “international royalties” gets reported on a statement that allows a buyer, a seller, or an auditor to tell them apart.
The audit I run on any catalogue before a valuation is not, in the first instance, about finding missing money, although it usually does. It is about breaking apart the single-bucket illusion and replacing it with a territorial map, each cell of which is scored for timing risk, political risk, structural friction, and black-box exposure. Only once that map exists does a number attached to “o exterior” start to mean anything, and only then does a valuation stop being a guess dressed up in a multiple.
Here is the test. Ask whoever administers your international rights for a territorial breakdown of the last three years of income, broken down by royalty type, with payment dates, matched against the usage periods those payments cover. If the answer for certain royalty types and certain territories is that no pathway exists, that is not the end of the conversation. That is the beginning of a different conversation, about whether the pathway should be built, what it costs to build, and what income has been accumulating abroad while nobody was collecting it. If the answer is that the pathway exists but no breakdown can be produced, the question is no longer whether something is wrong with your statement. The question is what else they have not told you, and why.
One last thought. Howard Marks, writing about financial markets rather than music rights, makes the point that the most expensive mistakes in investing come not from wrong information, but from information that is missing and assumed to be present. That is exactly the mistake embedded in the phrase “os royalties de fora.” The information is missing. It is being assumed to be present. And the assumption is costing the Brazilian music industry far more money than anyone has ever bothered to count.