It’s always tempting to invent an elaborate story for why a particular investment is brilliant, why a trend is about to start (or reverse!), or why some incredible, sweeping change will alter the world and make savvy investors a profit.
But there’s an easier place to start, which is to assume that trends are going to be pretty much the same as they always were.
Investors diving into the music royalties business for instance may wonder: how can you predict which catalogs will perform? It might be tempting to “buy what you know,” but tastes vary — and there’s an easier way.
As it turns out, royalty investors can take advantage of a phenomenon known as the “Lindy Effect.”
The Lindy Effect is the observation that the best predictor of how long something will last is how long it’s already been around. If a book is a bestseller today, it might be populating used bookstores and stoops in six months. But if it’s been in print for a year, it has staying power. If it’s been in print for a decade, it’s going to be around longer. (The bestselling book published in 2019, Where the Crawdads Sing, sold 4.5 million copies. But that’s about 1/20th of the number of copies of the Bible sells in a given year. And it’s been in print since the invention of the printing press!)
The Lindy Effect works for music, too. If you want one analytical tool to estimate how well a given catalog will perform, look at the “dollar age” of royalties.
Computing the “dollar age” of a royalty stream is simple: take each royalty payment, multiply it by its age, and compute the average.
Here are two examples:
In this case, Catalog A has produced lower overall results, but in a smaller range; Catalog B has a lower dollar age — it’s getting more popular.
We can see this in live data as well. Here’s an example of a longstanding catalog with a dollar age of 4.6 years:
As this chart shows, the catalog had declining revenue for a long time — but continued to produce a stream of royalties. Then, in late 2018, royalties picked up again.
The beauty of this calculation is that it does the research for you: you don’t have to understand why a given catalog keeps on producing royalties to know that it does. Is it nostalgia? Is it a small base of intense fans? A band that keeps cranking out new music so their catalog keeps getting rediscovered?
It doesn’t matter. What matters is that the royalty income stream keeps on coming, year after year.
This is not just a tool for analyzing musical assets. There’s a long precedent for using historical stability as an indicator of future returns. Sophisticated quantitative investors use the “quality factor” to predict future performance: they look at things like the stability and durability of a company’s earnings over time. As it turns out, these signals work: stable companies produce healthy excess returns. And even those quants are riffing off of other brilliant investors: the “quality factor” in equities is a way to partially automate the investment process of Warren Buffett.
Buffett has famously talked up the value of long-term brands. On Gillette, he said, “It’s pleasant to go to bed every night knowing there are 2.5 billion males in the world who will have to shave in the morning.” On Coca-Cola, Buffett cited a Fortune magazine quote: “It would be hard to name any company comparable in size to Coca-Cola and selling, as Coca-Cola does, an unchanged product that can point to a ten-year record anything like Coca-Cola’s.” That quote was from 1938.
Of course, this kind of investing mindset doesn’t directly apply to an asset like royalties: investors using the “quality” signal also look at variables like leverage and profit margins that don’t have a direct analogue in royalties. But the general insight is the same: while the future is uncertain, history is a good guide. And in finance, you can boil history down to quantifiable metrics.
Royalty streams for a given catalog don’t march straight up every single quarter; there are fluctuations as bands experience a burst of popularity or a lull. But, over time and on average, they’re predictable.
But when you factor in dollar age, they’re more predictable still: a royalty prediction model gets 45% more accurate if it incorporates dollar age. And that’s predicting royalty streams on a per-catalog basis — always a tricky business. In the aggregate, looking at a wide set of music assets, a model that predicts royalties next quarter can get halfway there solely by looking at the last quarter’s royalties. But changes in royalties over time are not strongly predictive. However, a model that looks at changes in royalties weighted by dollar age can predict over half of the next quarter’s change in royalties.
In other words, a simple analytical tool looking just at the average time that a catalog has produced royalties will tend to provide a strong investing edge.
Even better, it’s a way to expand the investable universe. Instead of treating music royalties as a specialty asset requiring domain expertise and rigorous qualitative analysis, you can look at it as a simple bet that the best way to predict popularity in the future is to see a lot of it in the past. Analyzing and predicting the popularity of cultural artifacts is hard, but numbers give you a “cheat code” to quickly spot what’s a flash-in-the-pan and what’s a long-term winner.