Music royalty investment research.
Why music royalties, why now, and why independent songs. The figures behind SONOR Capital, with their sources, including Goldman Sachs' 2026 outlook.
All music revenue Goldman expects in 2035, from $112.5bn in 2025. 5.8% a year.
Recorded music paid to rightsholders in 2035, from $31.7bn. 5.9% a year.
Paying streaming subscribers by 2035, from 837 million today.
Yearly growth in subscription income paid to rightsholders, to 2035. The fastest line in the report.
Five insights, and one warning
The research in brief: what Goldman Sachs says, then what it means for an investor in short royalty advances.
Growth is back, and revised up.
Goldman expects music to reach $198.1bn by 2035, from $112.5bn. Its first upward revision since 2022.
We do not need it. Every advance is sized on trailing royalties, never the peak. Growth is upside we do not pay for.
Subscriptions are the engine.
Paid subscribers go from 837M to 1.525bn. Prices rise every 12 to 24 months with little churn.
Our advances are recouped from monthly streaming income. Each price rise lands inside our 12-month window, free.
Goldman prefers the pipes to the catalogue.
It favours platforms, distributors and live. It is selective on companies that own music rights.
We do not buy catalogues, in general. We buy receivables over back and future catalogues. In general, we buy contingent or proven receivable rights. The platforms pay us at source; an advance recoups in about 8 months, then 30% stays with us until year 7. The artist keeps the song.
Fans pay more than anyone thought.
About half of fans pay above the minimum price when they can. Direct-to-fan sales do not cut streaming income.
That is why our advance comes in cash and in growth services, such as marketing. A bigger fan base lifts the royalties that recoup it.
Making music got easier.
Goldman flags a structural risk of market-share dilution for major labels, as tools lower the barrier to release.
More independent songs earning EUR 5,000 to 250,000 a year. That is the middle market we serve.
What cuts against us.
Millions of AI-made tracks could divert part of the royalty pool away from human music. And forecasts miss: 2024 came in below Goldman's own.
We check every song is real, size every advance on trailing royalties and reprice every cycle.
Four engines behind the income.
Durability is the foundation: good songs keep being played and the income holds up year after year. The 2026 Goldman Sachs report adds how many separate engines now sit behind that income, each with its own lever.
Price
Goldman expects the major platforms to raise prices every 12 to 24 months. Netflix's Standard plan costs almost 1.5x a music subscription (Bolero's comparison). In Goldman's 2024 edition, Netflix's US standard price was up 94% since 2013, about 15% every two years, while music prices had stood still for about a decade.
Emerging markets
In 2025, Asia, Latin America, Africa and the Middle East brought nearly 67% of new paid subscribers. Only about 10% of their internet users pay, against 42% in developed markets. Goldman expects up to 75% of net additions from these markets by 2035.
AI
Goldman now treats AI creation tools as net upside for rights holders: $3.3bn a year by 2035 in its base case, about 10% on top of subscription streaming. Bear case $0.9bn, bull case $6.6bn. It still warns that AI tracks could dilute the royalty pool.
What rights holders keep
Of every $100 listeners spend, about $57 reaches rights holders, and Goldman expects that ratio to hold. Subscription revenue to rights holders goes from $17bn in 2025 to $33bn in 2035, about 7.1% a year.
Goldman Sachs Research, Music in the Air 2026, central case unless marked. AI bars are 2035 licensing from AI creation tools, added on top of the $33bn. Figures as relayed by Bolero Music, AI Musicpreneur and Hypebot, not checked against the full report. The $57 is measured on what listeners spend, a different base from the roughly 70% of platform revenue shown earlier on this page. The Netflix comparison is Bolero's, the Netflix price history is from Goldman's 2024 edition.
None of this is our return. We size every advance on trailing royalties. These engines are upside we do not pay for, and a reason the pool the royalties come from is growing.
Beyond songs: catalogues, licences and other contingent revenue
The figures above describe recorded music. Our wider thesis applies the same method to catalogues, signed licences, ticket sales, creator payouts and game revenue, where income is real but not yet paid. Goldman favours platforms, distributors and live, which fits this direction. We do not yet have sourced research on live, creator or gaming receivables. We will add it here when we do, and we will not publish a figure we cannot source.
In one page
Independent music royalties are a growing, cash-paying asset that big money has not priced at the song level. SONOR advances against them, in 12-month cycles.
- The market keeps growing. Recorded music reached $31.7bn in 2025, up 6.4%, the eleventh year of growth in a row.
- The next ten years look better still. Goldman Sachs expects all music revenue to reach $198.1bn by 2035, from $112.5bn, its first upward revision since 2022.
- Songs earn for decades. Music's top estates earned $34M to $600M in a single year, long after the artist died.
- SONOR buys short slices of that income. The platforms pay at source, an advance recoups in about 8 months in our model, and capital returns in 12 months.
Songs pay for decades
A good song keeps paying long after the artist stops working.
Source: Forbes, via Statista. Pretax, including sales, streams, licensing and catalogue deals in the period, so not royalties alone.
These are the stars, and the figures include one-off catalogue sales. The point is simpler: the market pays heavily for music income that lasts. SONOR works in the middle of the market, songs earning $5k to $250k a year.
How music money works, in plain words
Every play earns a small amount, and that money passes down a fixed chain once a month. SONOR sits in that chain, ahead of the artist.
Simplified. Shares vary by platform, country and contract. On a phone, swipe the diagram sideways.
- Two rights in every song. The recording (the "master") and the composition (the words and melody). Each earns on its own.
- The platform keeps about 30%. Roughly 70% of streaming revenue goes to rightsholders. Platforms pay monthly, about two months after the stream.
- The distributor collects the recording's share. Independent artists use a distributor to reach the platforms. StreamAdvance is ours.
- Splits divide each payment. An advance adds SONOR as a split, first in line, until it is recouped.
- An advance is not a loan to the artist. It is recouped only from the royalties of the songs distributed through us. If they earn less, recoupment takes longer, and the artist owes nothing else. It can be paid in cash, in growth services such as marketing, or both. That is why SONOR sizes every advance on trailing royalties, never the peak.
The next ten years: Goldman Sachs, 2026
In October 2026 Goldman Sachs raised its long-term music forecast for the first time since 2022. Its view: every part of music grows, subscriptions grow fastest, and the businesses that move music beat the businesses that own it.
Source: Goldman Sachs, Music in the Air 2026, as reported by Hypebot and Digital Music News, October 2026. Drawn by SONOR Capital from the published figures.
| Measure | 2025 | 2035 | Per year |
|---|---|---|---|
| All music, retail (recorded, publishing, live) | $112.5bn | $198.1bn | 5.8% |
| Recorded music, paid to rightsholders | $31.7bn | $56.1bn | 5.9% |
| Subscription streaming, paid to rightsholders | $17bn | $33bn | 7.1% |
| Live music | $37.4bn | $65bn+ | 5.7% |
| Music publishing | +$12.1bn over the decade | 4.8% | |
| Paid streaming subscribers | 837M | 1.525bn | |
| Generative AI, base case | +$3.3bn by 2035, about 10% on subscription streaming | ||
Source: Goldman Sachs, Music in the Air 2026, via Hypebot. Goldman also expects price rises every 12 to 24 months on individual and family plans.
- Price rises continue. Platforms have moved past $9.99 with little churn. Goldman expects increases every 12 to 24 months, and a modest per-listener fee on family plans.
- Fans spend more than modelled. Goldman now calls its old assumption, that superfans spend about twice an average listener, too conservative. About half of fans pay above the minimum when given the choice, and direct-to-fan releases do not reduce later streaming royalties.
- Ad-supported streaming recovers. After years of weak results, and with new middle tiers between free and premium.
- Live is resilient. US live spending grew in 31 of the last 35 years, and by 7.3% a year on average in recession years outside COVID.
- Labels face dilution. Cheaper creation tools mean more releases and a structural risk to major-label market share. AI-made tracks could divert part of the royalty pool from human music.
Sources: Goldman Sachs Insights, August 2025 (2023 and 2024; 2024 was the first year revenue fell short of Goldman's forecast since it began tracking); Music in the Air 2026, via Hypebot (2026).
What we take from it, and what we do not
| Goldman's view | What SONOR does with it |
|---|---|
| Music nearly doubles by 2035 | Treated as upside. Advances are sized on a bad year, not on the forecast. |
| Subscriptions grow 7.1% a year, prices rise every 12 to 24 months | Streaming income recoups our advances. Price rises inside a 12-month cycle are free. |
| Favour distributors and live, be selective on rights owners | We distribute and collect; we do not buy catalogues. The artist keeps the song. |
| Fans pay more, direct-to-fan does not cannibalise | Advances include growth services, such as marketing, to grow the fan base behind the song. |
| AI tracks dilute the royalty pool | Every song is checked as real. Every cycle is repriced. |
| 2024 missed Goldman's own forecast | We never need a ten-year forecast to be right. We need the next 12 months of an existing song. |
Why now: the capital side
The market is growing. So is the money looking for it, and the playbook already works in other industries. The figures, and what they mean.
| Shift | The data | What it means |
|---|---|---|
| Music keeps growing, and independents lead it. | Recorded music reached $31.7bn in 2025, up 6.4%, its eleventh year of growth (IFPI). Goldman Sachs expects all music revenue to reach $198.1bn by 2035. Independent music holds 46.7% of the market by ownership (MIDiA). | The growth is in independent music, the part the big funds are too large to serve one song at a time. Italy is off their map. That is where we originate. |
| Italy is outgrowing Europe, on Italian music. | The Italian market reached €513.4M in 2025, up 10.7%, against 5.6% for Europe. Streaming is about €340M of it. 85% of the top 100 albums are Italian. 245 albums passed 30 million streams. | A deep, local pool of songs, much of it from the urban and trap wave of the last decade, owned by independent labels and artists. Off the map of US buyers. |
| Institutions still bid for royalty income. | KBRA has rated more than $12.9B of music-royalty bonds since 2020, bonds paid from royalty income. It expects 2026 issuance to fall about 25%, to just over $2.5B, mainly because issuers are consolidating, not because buyers are leaving. | Fewer, bigger issuers need fresh, granular collateral. Seasoned pools of recouping songs are exactly that. Whoever produces them has pricing power, and an exit. |
| Family offices want short, collateralised yield. | Apollo calls asset-based finance an estimated $20 trillion market, bought for self-amortising payments, collateral and short duration: most of what it owns runs three to five years, "less time for things to go wrong." KKR raised $6.5B for the strategy in 2025 and expects private ABF to exceed $9 trillion by 2029. The Swiss National Bank rate has been 0% since June 2025. | A royalty advance pays back in 6 to 14 months. Shorter, more granular and more self-liquidating than what the big managers sell, in a niche far too small for their tickets. Against 0% on Swiss cash, the case is simple. |
| Model | Where the profit is | SONOR equivalent |
|---|---|---|
| Restaurants · McDonald's | Franchised restaurants bring about 90% of McDonald's restaurant margin: about 84% margin on franchise revenue, against about 15% on the restaurants it runs itself. | The collection account. Every royalty flows through our distribution first, advanced or not, and we earn on all of it. |
| Payments · Stripe, YouLend, Värde | Stripe calls its merchant cash advance "a purchase of a merchant's future receivables, not a loan." A share of each payout is withheld before the merchant is paid. YouLend, which provides Stripe's US cash advances, agreed in March 2026 to sell up to $225M of its receivables to Värde, a private credit fund. | Advances recouped at source through our distribution, with family offices funding the receivables. |
| Insurance · Lloyd's of London | About 39% of Lloyd's premium is written through delegated arrangements: specialist underwriters writing to a mandate, on capital others supply, paid through commission and often a profit share. | SONOR underwrites to a written mandate. Family offices supply the capital. SONOR Capital earns fees and owns the return engines. |
The market: large, growing, going independent
| Measure | Figure | Source |
|---|---|---|
| Recorded music revenue, 2025 | $31.7bn, up 6.4% | IFPI 2026 |
| Streaming share | $22bn, 69.6% | IFPI 2026 |
| Fastest-growing regions | Latin America +17.1%, MENA and Sub-Saharan Africa +15.2% | IFPI 2026 |
| Recorded music, 2035 forecast | $56.1bn, from $31.7bn in 2025 | Goldman Sachs 2026 |
| Subscription streaming growth to 2035 | 7.1% a year | Goldman Sachs 2026 |
| Independent share, by ownership | 46.7%, $14.3bn in 2023 | MIDiA |
SONOR does not price on these forecasts. Expected streaming price rises, every 12 to 24 months, are a tailwind it does not pay for.
How it compares with other assets
Stocks earned more on average from 2021 to 2025, but every listed asset class lost between 17% and 37% in a single year. The median music royalty asset never had a down year in the study period.
Illustrative. Indices 2021 to 2025 (Nasdaq-100 and MSCI World in USD, euro bonds and EU real estate in EUR). Swing = average year minus worst year, in percentage points. *Music royalties: median life-of-rights asset, 2017 to 2021, Cornell study of Royalty Exchange deals, a different and earlier window; no index exists. In the same study the bottom tenth of single assets lost 42.8% in 2018, which is why we spread every ticket over about 50 artists. **EU private credit: an indicative gross yield range of about 7 to 9% from manager and bank commentary, not a realised-return index, so no worst-year bar. SONOR: target range, not realised returns, no track record.
In twelve months, one of these puts cash back in your account.
CHF 100,000 in. At month 12 our approach returns your capital in cash, then pays the preferred return. Listed assets show a paper value somewhere in a range.
Illustrative. SONOR classes: targets, paid only out of portfolio profit, capital can be lost, no track record. Growth shows the 12% preferred return plus a target share of profits toward a 20%+ IRR. Listed assets: range from the worst to the average year of 2021 to 2025, paper value. Months 12 to 18: your first-cycle capital and preferred return are already back in cash, and anything you re-commit is six months into a second cycle. **EU private credit: indicative gross yield, capital still outstanding.
Markets fell in 2022. Music revenue kept growing.
Inflation, rate rises and war in Europe hit stocks, bonds and listed real estate together. Recorded music revenue grew 9.0%.
Illustrative. Asset bars: worst single year in 2021 to 2025. Music: global recorded music revenue growth (IFPI), which is the market, not a SONOR return. Residual risks: platform payout rules, USD royalties, AI-generated tracks and our short track record.
The gap we fill
Who we advance to. Independent artists, managers and small labels, with songs that already earn about EUR 5,000 to 250,000 a year, and an owner who controls their share of the income.
Why this is not a major label. Majors sign artists, own the recordings and recover costs across a whole career. We buy a share of the songs' income for about a year, and the artist keeps the songs. Buying major-label shares means owning a company and its stock-market swings; SONOR gives you the royalty cash flow itself.
| Major label share | Catalogue fund | SONOR | |
|---|---|---|---|
| What you own | A listed company | Decades of a few catalogues | One year of many songs' income |
| Spread | Tied to top artists | A handful of hits | ~50 artists per EUR 100,000 |
| Moves with stocks | Yes | If listed | No |
| Ticket | Any | Often USD 10M+ | From CHF 100,000 |
Why the money reaches us
SONOR holds a written right to the song's income and sits in the payment line before the artist. What it does not control is how much the song earns.
| Secured | Not secured |
|---|---|
| Our right to the song's income for the agreed period | The amount the song will earn |
| Our place first in the payment line | Platform rule changes and delays |
| Automatic collection, without chasing anyone | Dollar exchange-rate moves |
Why not a catalogue fund
Catalogue funds buy decades of income up front. When valuers raised the Hipgnosis Songs Fund discount rate from 8.5% to 9.63%, its catalogue value was cut by 26%. SONOR's advances run on 12-month cycles, so a rate move means a new price next cycle, not a write-down.
Sources
Goldman Sachs and institutional sources only. Goldman Sachs 2026 figures are taken from press coverage of the report.
- Forbes, highest-paid dead celebrities 2024, via Statista
- IFPI Global Music Report 2026, via Music Week
- Goldman Sachs, Music in the Air 2026, via Hypebot
- Goldman Sachs, Music in the Air 2026, via Digital Music News
- Goldman Sachs bullish on the next decade, HITS
- Goldman Sachs Insights, music revenues forecast to double by 2035 (August 2025)
- MIDiA Research, independent market share by ownership, 2023
- Music as an Asset Class, Cornell, arXiv 2602.05007
- Hipgnosis revaluation, Billboard
- Index returns: Nasdaq, MSCI, Bloomberg Euro Aggregate (via SSGA), FTSE EPRA Nareit Developed Europe (LSEG)
- FIMI, Italian market 2025
- FIMI, top albums and streams 2025
- KBRA, Playback: music royalty ABS, May 2026
- Apollo, The asset-backed finance market is having a moment, Aug 2025
- KKR completes $6.5B asset-based finance fundraise, July 2025
- Swiss National Bank, monetary policy assessment, September 2026
- McDonald's, Form 10-K for 2025 (SEC)
- Stripe, merchant cash advance guidelines
- Värde, forward flow with YouLend, March 2026
- Lloyd's delegated authority share (Insurance Journal)
We say it first.
Spotify and the others can change how they pay. Our advances run on 12-month cycles, so a rule change means a new price next cycle, not a loss this year.
Every song earns less over time. We plan for it. You can see the plan.
You cannot take money out early. Royalties are paid in dollars, so the exchange rate matters too.
Frequently asked questions
What does Goldman Sachs forecast for music?
In October 2026 Goldman Sachs raised its long-term forecast for the first time since 2022, to $198.1bn of total music revenue by 2035.
How big is recorded music today?
Recorded music reached $31.7bn in 2025, up 6.4%, the eleventh consecutive year of growth (IFPI).
Is market growth SONOR's return?
No. Market growth is not a promise on a portfolio. SONOR's targets are separate, and they are targets, not promises.
Where do the figures come from?
Goldman Sachs Research (Music in the Air 2026), IFPI and MIDiA. Some figures are taken from press summaries and not checked against the full report.
Apply for details.
Tell us who you are. We will contact you when the full offering opens, by 10 January 2027. Family offices, private credit funds and qualified investors.
Prefer email? Write to apply@sonorcapital.com
The offering is for qualified and professional investors only, subject to eligibility checks. See our privacy policy.