The Hidden Costs of Music Royalty Infrastructure

What’s 1 + 1?

You didn’t look it up. You didn’t call anyone or wait for a statement in the mail, and you didn’t wonder whether whoever told you the answer was big enough to be trusted. You knew it was 2, and you could check it yourself. Anyone, anywhere, gets the same answer for free.

Now think about what it takes for an artist, a songwriter, a label or a publisher to trust that they were paid correctly for a stream. They don’t check anything. They trust an institution, and trusting an institution is something the whole music business pays for, every month, whether or not anyone sees the bill. This post is about that bill.

Trust Has a Running Cost

The money from a stream or a sale has to reach two sets of owners: whoever owns the recording, and whoever owns the composition, the song itself. Today that route runs through institutions: distributors, labels’ accounting departments, publishers, administrators, and collecting societies like ASCAP, BMI, GEMA, PRS and the MLC.

We trust them mostly because they’re big and they’ve been around. That isn’t irrational. A large, long-lived institution has a lot to lose by getting things wrong, and that’s a reasonable thing to lean on when you can’t check the work yourself. But being trusted that way costs a lot, and the cost doesn’t go away. It has to be paid every year to keep the institution standing:

• Operating the institution. ASCAP’s overhead was about $186 million in 2025. GEMA’s was €198.7 million (about $224 million) in 2024. These costs come out of the money before it’s paid.
• Keeping it in place. The MLC’s budget for 2024 was $39.05 million, paid by streaming services on top of royalties and allowed to rise by up to 3 percent a year. Before it paid out a cent, it took $33.5 million to set up, plus $28.5 million for its first year of operations.
• Defending it. In 2024, ASCAP, BMI, SESAC, the NMPA, SoundExchange, Universal and Sony together reported about $6.8 million of lobbying. Not all of it is about royalties, but the institutions were set up by law, so the law is where they have to defend themselves. (We wrote more about this in The Financial Rails for Music.)
• Things nobody publishes. Rate proceedings, rate courts, government oversight, and the reciprocal agreements that connect about 230 societies around the world all cost something. We couldn’t find a number for any of them.

And there’s time. The MLC pays about 75 days after the month of the stream. BMI pays about six months after a performance on average. Money from abroad takes six to eighteen months. Waiting costs money too.

None of this means anyone is doing a bad job. It’s simply what it costs to run trust through organizations: buildings, staff, audits, lawyers and politics, all so that people have a reason to believe the numbers they can’t check.

Math Doesn’t Need a Reputation

Back to 1 + 1. Nobody needs an institution to vouch for arithmetic, because arithmetic vouches for itself. Anyone can check the result, and nobody gets a different answer by being bigger.

Cryptography stretches that property a long way. A digital signature is a math problem: only the holder of a key can produce it, and anyone can check it. A blockchain is a record that thousands of computers keep and check against each other with those same math problems, so nobody has to trust any single one of them. And a smart contract is code on that record. It runs the same way every time, anyone can read what it does, and anyone can check that it did it.

So the question “was I paid correctly?” changes. It used to be answered by trusting whoever held the books. Now it can be answered the way 1 + 1 is: by checking.

Math can’t do everything, though. It can’t tell you who wrote a song. On Miso, owners record their splits before anyone pays, and the code enforces them; deciding who the owners are still happens outside it. And code can have bugs, which is why the contracts sit on a public chain where anyone can inspect them. But the expensive part of today’s system is the routing and the record-keeping, and that’s the part math is good at.

What Moving the Money Costs Each Way

We tried to put a number on both routes. We only counted the plumbing: getting money from a stream or a sale to the owners of the recording and the composition. The sale itself, licensing for radio and venues, and the splits between labels and artists or publishers and writers are all left out.

The old route, counted as low as we could. We started with the roughly $28.5 billion a year of digital rights money in the world: about $22 billion from streaming recordings and about $6.5 billion for compositions. Then every assumption went in the old system’s favor:

• We counted the recording side’s cost as zero, though labels and distributors do charge for it.
• On the composition side we counted only digital collections and pretended every society in the world runs as cheaply as the cheapest one, the MLC, at under 4 percent. The real range is 9 to 15 percent.
• We left out publisher and administrator fees, unmatched money that never gets paid, and the cost of waiting.

Then we added back only the MLC’s $39 million budget, which is documented and sits outside those collections. That comes to about $249 million a year, or $21 million a month. It’s a floor. The real figure is much higher.

The Miso route. On Miso, paying out a sale is a single transaction. Splitting a $20 sale across a release’s four tracks costs about $0.00016, and the cost is the same whatever the amount.

To be generous to the old system, we assumed every sale takes six separate transactions to reach both sets of owners, with nothing batched, each costing as much as that payout. We also assumed the same $28.5 billion a year, moved in $20 pieces. That’s about 120 million payouts a month, for about $117,000 a month: about 0.005 percent of the money moved.

	Per month	Per $100 moved
The old route, at its floor	about $21 million	about $0.87
Miso, overstated	about $117,000	about $0.005

That’s about 180 times cheaper, comparing our most generous estimate of the old system with our least generous estimate of ours.

That comparison is tilted toward the old system on purpose. For what we think is the realistic number, use the rates the industry actually charges: distributors’ fees, collecting societies’ commissions, administrators’ cuts, and the songwriter money that never gets matched to anyone. On those rates, about $11.90 of every $100 is gone before the owners see it, around $280 million a month, and that’s before the months of waiting. On Miso, with owners collecting their shares in batches the way they would in practice, the same money costs about $30,000 a month to move. That’s closer to 10,000 times cheaper.

Some honest limits. This compares the cost of moving the money, not of running Miso as a company. Miso charges a platform fee, the same way any store does, and that’s a business cost like a store’s margin, not a cost of moving the money. Song royalties earned outside Miso still go through the societies.

The Institution That Isn’t There

The biggest difference doesn’t show up in that table. Once Miso’s contracts are published, nobody can change them, and they don’t need anyone to run them. There’s no staff, no building, no budget, and no lobbying. If Miso the company stopped operating tomorrow, the payouts would keep running. The standing cost of keeping the route open is zero, because there’s no institution to keep standing.

That’s what it means for math to be the source of truth. The trust that used to be bought every year with overhead comes free with the arithmetic, the same way it does for 1 + 1.

What Else Could That Money Pay For?

Even at our lowest estimate, music spends about $249 million a year moving its own money around, and that’s just the digital composition side. The real number is several times higher.

That money comes out of the same pool as everything else in music: studio time, session players, mixing engineers, tour support, the next record. Every dollar spent keeping the plumbing standing is a dollar that didn’t go to any of those things.

We can’t promise where saved money goes. But on Miso nobody is in the middle to keep it. What a fan pays for a release goes to the owners under the splits they agreed to, and they decide what to do with it. We’d bet a lot of it goes back into music.

Nobody needs an institution to tell them 1 + 1 = 2. We don’t think anyone who makes music should need one to tell them they’ve been paid.