💰 quantitative finance

Music as an Asset Class

This paper demonstrates that Life of Rights music royalty assets, when analyzed through discounted cashflow models and backtested over five years, exhibit risk and return characteristics comparable to the S&P 500 while offering potential diversification benefits due to their low correlation with traditional stock and bond portfolios.

Original authors: Sasha Stoikov, Aadityaa Singla, Umu Cetin, Luis Alonso Cendra Villalobos

Published 2026-02-06

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Sasha Stoikov, Aadityaa Singla, Umu Cetin, Luis Alonso Cendra Villalobos

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you have a collection of old, rare baseball cards. You know that every year, the players on those cards generate a little bit of "royalty" money (maybe from a video game or a trading card game) that gets paid to the card's owner. But unlike stocks, where you can buy and sell shares every second, trading these music rights is like trying to sell a rare card at a garage sale: it happens rarely, it's hard to know the true price, and the fees to make the trade are huge.

This paper is like a team of financial detectives (from Cornell University) trying to figure out if these music "cards" are actually a good investment, similar to buying stocks in big companies like Apple or Coca-Cola.

Here is the story of their investigation, broken down simply:

1. The Problem: Guessing the Price

In the stock market, prices move every second, so it's easy to see if an investment is good or bad. In the music world, deals are rare. Because there isn't enough data, it's hard to know if a song is "overpriced" or "underpriced."

The researchers wanted to build a calculator (a model) that could look at how much money a song made in the past and predict what it's worth today, just like a stock analyst does.

2. The Experiment: Three Different Calculators

They looked at 1,295 real music deals from a platform called "Royalty Exchange" (think of it as an eBay for music rights). They tried three different ways to build their calculator:

Calculator #1 (The Optimist): Assumes the song will make the exact same amount of money every year forever, and the risk is the same for everyone.
Calculator #2 (The Realist): Assumes that if a song's income is shaky (sometimes high, sometimes low), it's riskier, so we should demand a higher return. It also assumes the income might settle down to a lower, steady amount.
Calculator #3 (The Historian): This one added a special rule: Older songs are worth more. Just like a vintage car or a classic novel, a song that has been around for decades is seen as more stable and valuable.

The Winner: Calculator #3 was the best at matching the real prices people actually paid. It confirmed two things the market already knew:

Stability is King: Songs with steady income are priced higher.
Age is Gold: Older songs get a "premium" price tag.

3. The Test: The "Buy and Hold" Simulation

Once they had their best calculator, they ran a time-travel simulation (called backtesting). They asked: "If an investor had bought a random music song in 2017 and held it for 5 years, how much money would they have made?"

They compared two types of music assets:

The "10-Year" Contract: You own the rights for exactly 10 years, then they expire and become worthless.
The "Life of Rights" (LOR) Contract: You own the rights for as long as the copyright lasts (often 70+ years).

The Results:

The 10-Year Contract: These were like a high-interest savings account that slowly drains. You got a huge payout (dividend) every year, but the value of the asset itself dropped fast as the expiration date got closer.
The Life of Rights (LOR) Contract: These were like a steady, growing tree. You got a good yearly payout (about 12%), and the value of the tree itself actually grew a little bit over time.

4. The Big Comparison: Music vs. The Stock Market

The researchers compared their "Music Portfolio" to the S&P 500 (the standard index for US stocks).

The Similarity: Over a 5-year period, the average music investor did just as well as the average stock investor. Both made roughly 12-13% per year.
The Difference: The stock market is like a rollercoaster—it goes up and down wildly from year to year. The music market was like a cruise ship—it moved much more smoothly. One year stocks might crash, while music stays steady.

5. The Catch (Limitations)

The authors are honest about the flaws in their study:

High Fees: Trading music is expensive (about 8% in fees), which eats into profits.
In-Sample Bias: They built their calculator using all the data up to 2025. A real investor in 2018 wouldn't have known the future, so the results might look too good.
Missing the "Art": Their calculator only looks at numbers. It doesn't know if a song is "good" or if it's becoming popular again on TikTok. It treats music purely as math.

The Bottom Line

The paper concludes that music royalties (specifically "Life of Rights" assets) can be a solid part of an investment portfolio. They offer returns similar to stocks but move in a different direction (they are "uncorrelated"). This means if the stock market crashes, your music portfolio might stay calm, helping to balance out your overall money.

However, for music to truly become a mainstream asset class like stocks, we need better ways to measure the "artistic" value of a song, not just the math.

Technical Summary: Music as an Asset Class

Problem Statement
The market for music royalties is characterized by illiquidity, infrequent transactions, significant information asymmetry, and high transaction costs. While revenue data for rights holders has become more transparent in the streaming era, there is no established framework to quantify the risk and return characteristics of music royalty assets with the rigor applied to equities. Existing valuation methodologies—Discounted Cash Flow (DCF) analysis and comparable market transactions—are often used ad-hoc or are limited by sparse data, making it difficult to estimate asset value fluctuations or assess investment performance relative to traditional portfolios.

Methodology
The authors address this gap by fitting three distinct DCF models to a dataset of 1,295 transactions from the Royalty Exchange platform (2017–2025), comprising 1134 primary sales with a total face value of $97 million. The dataset includes deal-level metrics (price, Last Twelve Months revenue [LTM], Last Three Years revenue [LTY], catalog age, and contract terms) and time-series revenue data.

The authors define the annual return ( $r_i$ ) of an asset as:
$r_i = \frac{p_{i+1} + c_i - p_i - t_i}{p_i}$
where $p$ represents the model-implied price, $c$ is the collected cashflow, and $t$ represents transaction costs (a $500 buyer fee and 8% seller commission).

Three models were calibrated to minimize the Mean Squared Error (MSE) between traded multipliers (Price/LTM) and model-implied multipliers:

Model 1: Assumes flat expected revenues and a constant discount rate ( $r$ ).
Model 2: Introduces risk-adjusted discount rates based on revenue volatility (measured by the deviation of LTM/LTY from 1) and assumes cashflows stabilize at a level lower than LTM.
Model 3: Extends Model 2 by adding a premium for older songs (catalog age) to the expected cashflow.

Key Results

Model Performance: Model 3 yielded the lowest MSE (5.7) compared to Model 1 (8.8) and Model 2 (7.7), indicating that accounting for catalog age and revenue volatility significantly improves pricing accuracy. The calibrated parameters suggest a base discount rate of roughly 8.3%, a cashflow stabilization factor of 0.61, and a volatility penalty coefficient of 0.058.
Backtesting (2017–2022): Using Model 3, the authors simulated buy-and-hold strategies for one-year and five-year horizons, comparing "Life of Rights" (LOR) assets against 10-year (10Y) contracts and the S&P 500.
- LOR Assets: Over a 5-year period, LOR assets achieved a median total return of 82.64% (12.8% annualized). Returns were driven primarily by consistent dividend yields (11–13% annually) with modest capital appreciation. The 5-year median return was comparable to the S&P 500 (12.2% annualized).
- 10Y Assets: These exhibited lower median returns (7.3% annualized) due to significant capital depreciation as contracts approached expiration, despite higher dividend yields (13–17%).
- Volatility and Correlation: While the dispersion between the 10th and 90th percentiles was substantial for music assets (reflecting unpredictability), the median returns of LOR assets were more stable year-over-year compared to the S&P 500. The authors note that music asset performance appears largely uncorrelated with traditional equity markets.

Significance and Claims
The paper claims that Life of Rights music assets, when held over a five-year horizon, offer risk-return characteristics comparable to S&P 500 equities. Crucially, the authors posit that because music asset performance is likely uncorrelated with stocks and bonds, these assets could serve as a valuable diversification tool for traditional portfolios.

The authors maintain a modest tone regarding their findings, acknowledging several limitations:

The models do not explicitly incorporate qualitative factors such as genre, specific popularity metrics, or industry growth rates.
The backtesting is "in-sample," as model parameters were estimated using data available through 2025, which would not have been accessible to an investor in earlier years (e.g., 2018).
The assumption of constant interest rates is a simplification.

The paper concludes that for music to mature as a true asset class, pricing models must evolve beyond financial history to incorporate measures of musical quality and cultural relevance, as the drivers of performance are ultimately artistic rather than purely financial.