Translate in another language

Long Tail from music catalogues to hedge funds going bust: Behavioural implications of small-world networks

Long Tail from music catalogues to hedge funds going bust: Behavioural implications of small-world networks

Of Music Charts, and Assets Class

Hi There,

Justin Timberlake is dancing his way to the bank as this week he sold his music catalogue to a fund backed by the private equity behemoth Blackstone reportedly for a US$ 100 Million. He joins music stars such as Bruce Springsteen, Madonna, The Chainsmokers, Bob Dylan, amongst others in this select club. Artists who have sold their music catalogue to if you permit me a bit of economics speak to “de-risk” their futures.

At the heart of this is a very interesting conundrum, how do we measure popularity? How does the popular music lead to a choice-overload? Ever since there has been a popular music scene there has been a conventional way of measuring popularity by looking at the charts in some form or other. Some bands seem to live on the charts, some have a niche following. The Rolling Stones, and Pink Floyd seem to be made for the charts. Some like the “Sugarman” Sixto Rodriguez need to be rediscovered. Some make it to the charts by a more circuitous routes.

For example the band Dire Straits released a song in 1978 called the Sultans of Swing, it failed to chart. The song was re-issued in 1979 and it reached the fourth spot in Billboard 100, and 8th on the British charts. The song also helped push the sale of the album, and even today remains not just the staple of classic rock, but is usually identified as the band's signature song. The unusual bit was that the song required six months to make it to a chart. Even in the much slower pre-internet world of music it was a very lucky break for the band, and the listeners!

Contrast it with today where if a song or album does not reach the top immediately post its release chances are it will never be able to. The earlier music albums made it to charts in a span of a month, where they usually climbed from lower positions to reach their peaks. This does not happen very much. This is the argument made in a fascinating paper by Lukas Schneider, and Claudius Gros, where they analyse five decades worth of data from music charts to map out the acceleration in cultural processes. The paper can be found here.

The authors of the paper look at the charts in different countries spanning the United States, and Europe. The figure below is the chart data before the 2014-15 split to include a multi-metric consumption including streaming. With streaming added to mix the very definition of popularity became very fragmented. This is something that the authors posit is due to the acceleration of time horizon and one time effects are more of a behavioural reason for making streaming decision.

An album the authors argue spends almost half the time less on the charts than it used to. By breaking down the data they illustrate the acceleration in number one songs, three decades ago there might have been a dozen top of the pop songs, now the number could be as high as 40, or more. There is therefore more competition in the music industry, and the tastes are more varied. There are so many charts that there are popular sites keeping a score of different charts with different metrics. Music has become very personalised so it is impossible to track the rise and fall of popularity, an artist could be on different positions on multiple charts, is it better or worse than being high on just one chart? With too much choice, comes too much confusion. Making the job of valuation just that much harder.

However what does this have to do with selling back catalogue? Well, two things artists are now more than ever dependent on utilising the heyday as best as they can, as the time they have on the popularity rung has come down by half as per the study. The other is that with music going digital it makes more sense to capitalise on the nostalgia factor. The entire catalogue can be bought and used for different kind of royalties from advertisements to cross-licensing. It also ensures that they can remain active while someone else can look at the business of managing their creative asset portfolio and the problems of dealing with it such as copyright infringement, licensing etc.

There is another factor attached to it. A cautionary tale of a theory failing to properly replicate. In 2004 the editor of The Wired magazine proposed the idea of The long tail, in the context of music it meant that 20% of the transaction would come from the current popular hits, and the rest from the catalogue. This was supposed to mean that niche bands, niche music, and one hit wonders on the archive would be available on demand and since the opportunity cost of digital versus physical storage is minimal it would be a game changer.

The theory will be two decade old in just two years, and it has not been able to replicate well at-least in the music scene. Three economists led by Will Page formerly the Chief Economist of Spotify, looked at transaction data of music records and found that just 3% of the songs available accounted for 80% of the streams/sells. Almost 80% of the songs in the catalogue had no transaction data at all. So no one had bought them, streamed them, or licensed them. Sometime in 2019 the massive cost of keeping a huge catalogue started pinching the digital streaming platforms, and they have been downsizing the amount of repository they hold ever since. Thereby making the catalogue of popular artists a valuable investment. Add to that the presence of very low interest rates and you have a booming market in a new asset class.

So for any company making a decision based on popular artists it makes sense to bet on those that have a wide appeal, and those that have massive fan followings. Mr. Timberlake has sold around 150 million records and has had a storied career. The back catalogue value of Mr. Timberlake is lower when compared with Bob Dylan's whose career has now spanned six decades, a few brick-bats but many accolades including a Nobel Prize.

With the interest rates in the market not remaining so low it will be interesting to see the asset class hold up, it could be a great deal for Mr. Timberlake. For economists and historians however it will be an interesting experiment to see the creation of a new asset class, and re-defining popularity for the post-internet age.

As the problem of figuring out popularity will continue to plague the internet streaming services for quite some time to come as there are two network effects at play. One of them called the small world network effect we will take a look at in the next section.


The Weird World of Small Networks

In September 1998 Russia had just gone bankrupt. A few days after that a massive bailout scheme was announced not to save Russia, but an obscure hedge fund called Long Term Capital Management (LTCM).

Had LTCM not been rescued the financial markets were at the brink of collapse! One hedge fund with a total borrowing of US$ 4 Billion would have lead to a massive correction (read terribly large losses) in the financial markets amounting to almost a Trillion US$!

LTCM was started by Nobel prize winning economists who used their insights on derivatives trading to start a hedge fund. Derivatives are an interesting asset class, and very simple to explain but require a mathematical firepower to compute. Derivatives are based on an underlying asset class, say Apple stock, an investor might think her stock might fall, another investor might bet it would rise. A contract might be stuck between them to trade at a future date, at a specific price point. The work of Black, Merton, & Scholes found a way to apply the principles of statistical physics to financial markets and thereby solving the great problem of incorporating future in the pricing mechanism. It helped that the model This was the insight which won them the highest accolade in economics.

So a hedge fund with these people as founders should not have failed. Fail it did. The fund was invested heavily in the Russian government bonds known as the GKO. When Russia defaulted the fund continued to maintain the position even after hundreds of millions of dollar losses everyday. Waiting for a turn around. Talk about sunk-cost fallacy.

The problem was an over reliance on computer models based on their work, which failed to take into account the weird world of power-law distribution. The long tail argument used in the music piece above is an illustration of this, though in streaming it is becoming more skewed in favour of a few rather than the old maxim of 80:20 pareto distribution.

This is where the weird and wonderful world of small world networks comes into effect. In 1967 Stanley Milgram started an experiment to map out the average path length for people in the United States by sending a letter from a distant location both geographically, and socially to another. The conclusion of the experiment suggested that human society is a small-world network, it also popularised the six-degrees of separation phrase. See the image below which is an approximation of the experiment.

The experiments asked strangers to send letters to each other in order to map out the shortest path possible. The findings of the experiment have been very influential in social networks, and popular culture (Source: Wikipedia)

The small world nature of some network means that the basic principle of inter-connected nodes does not exist. Think of vaccinating against measles if you vaccinate enough number of people, the chances of an outbreak are minimised. This is a bedrock of vaccination policies against just about anything from measles to Covid-19.

However, what if the connected networks have shortcuts and follow a power-law distribution? What that means is there is no critical number to infect in order for it to explode. Think about the early days of the Covid outbreak, where one single carrier could lead to extreme percolation. This is the insight that Marvin Du spoke about in a very important paper here. These are a small probability event, but the likelihood is very, very, very small but not zero. This is what makes them very powerful, and extremely paradoxical to understand.

The small world network in more ways than one defines our life today than we care to know. Think of any fad you know and chances are you have been introduced to a strong evidence for a small-world network. Let me help you, Clubhouse, and the myriad audio based networks came, conquered, and are now in the process of disappearing. Small world effect. People making a particular kind of coffee during the lockdown, and some people dancing to a certain song by jumping from cars which is very unsafe! All highly clustered with low degrees of separation. Where things either explode, or fizzle out.

LTCM hedge fund believed in what is best described as the outdated theory of perfect equilibrium which does not exist. The pricing models of the fund failed to take into account the relatively small change somewhere in Russia could lead to a default and a widespread loss on their books. Since the contagion was fairly localised but LTCM was inter-connected to most of the large financial institution a small probability event shook the world. A decade later it came back for an encore, known as the 2007 Great Recession.

The ubiquity of network effects is still not fully understood. Even now after a few economic recessions the evaluation of risk is extremely atomistic, and examined node by node, which leads to a certain confidence that these networks are robust, and not fragile. Yet they are as some investment bankers call them robust-yet fragile. Go figure!

Behaviour Science can help break down the issue of scale, and help with communicating the paradox of inter-connectedness with more real world evidence. Till then look out for the next big thing.


Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to IP Wave.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.