Making My Predictions

Share on facebook
Share on twitter
Share on linkedin
Share on whatsapp
Share on email

One of the challenges of “big data” is that it is so…big. For any given subject, we have so many ways to measure things. I can pull one set of data to prove my point; you can take the same set of data and pull a different metric to prove your point.

Take gun violence: gun control advocates have their set of data and analysis showing how guns increase homicides, suicides and violence in general. Pro-second amendment folks have their own data proving their own points.

This applies even for something as innocuous as picking TV shows for a streaming platform. In one recommendation I authored, I counted over 1,400 numbers in one powerpoint presentation. How do we figure these issues out with so much data to choose from?

Well, I have a way. It’s unscientific, as far as I know, but it works for me.

Anytime we come across a significant issue with tons of metrics and variables and data, we can employ this method. I call it the “as many measurements as possible” approach, and I’d love to find out there are other more scientific ways to do this. Here’s how it works:

Take as many measurements as possible and determine if they support or nullify the issue under question. If the majority, super-majority or vast majority support the case, then the phenomena is probably real.

The point isn’t to take just one measurement as our gospel but all the measurements we can. If 9 of our 10 metrics indicate that a phenomena is real, then it probably is.

Take global warming/climate change. If you measure temperature, in 95 percent of measurements or ways to measure it, the climate is heating up. Sure a handful of scientists can find one or two ways to show the world isn’t heating up. Meanwhile, 99% of the rest of scientists measure the data in hundreds of different ways from daily highs increasing to the average temperature averaged over the year from city temperature to countryside to oceans and say, “Man, no matter how you measure this, this impact is real”.

Same with gun violence. Guns lead to increased gun homicides and suicides.

Same with stream video: a show that does well in total viewers probably has the most hours viewed and attracted the most new customers and gets the best customer reviews and so on.

I bring this up to put us in the right mind to do our last dive through the data on mergers & acquisitions. I clearly have a hypothesis that a “tidal wave” of M&A isn’t coming because the tide has been coming in for a while now. Like most of the last decade. But now we need the data to really show us what has been going on. As I can see it, we’ve set the terms, reviewed the narratives, gathered the data and now we need to ask the data what it sets. Since we have so much data on M&A and so many different ways to measure it, we could easily pick one or two metrics and have them change our minds. I’d rather apply the “as many measurements as possible” approach: interrogate the data in as many ways as possible and let the overwhelming conclusions, if they exist, be our guide.

So here are our two final questions to answer:

– What is the historic rate of M&A? (Partly answered in Part I.)
– Is that rate increasing, decreasing or can we tell?

The latter question in particular gets to basically the question at the heart of making this prediction. If M&A activity has been steady for most of the last decade, or if it has been increasing, then we should use that knowledge to make our predictions of future M&A activity.

Fortunately, M&A lends itself well to “as many measurements as possible” analysis. M&A can be measured by total number of deals by the size of deals by the types of industry or by percentage of concentration. So let’s look at as many metrics as we can.

Metrics in Opposition

We should start with evidence that M&A has either been low or decreasing over time. And there is one data point that makes this case:Slide11

I’ve put this table in each article because I’m that stoked by it. You’ll notice it shows three key ways to measure this data: by total number of all deals, by deal value and by number of mega-deals. I focused on the latter two metrics in Part I to set the terms of debate because I felt they captured the feeling of a “tsunami” best.

But what if we focus on the total number of deals? Two articles on M&A activity from 2007 and 2008, also using PwC reports, indicated that, well have a look yourself.

Slide12

Yikes. The articles said that in 2007, M&A activity exceeded 1,200 deals in 2007 and 1,000 deals in 2008. (I couldn’t find reliable numbers for total value of deals or mega-deals in those articles so didn’t have the table go back that far.)

Far from showing M&A increasing, this shows the total number of deals decreasing over time. Though, if we exclude the 2007 and 2008 years, the number of deals looks flat, hovering for the most part in the mid-800 deals per year. That said, even 800 deals per year seems like a pretty frothy M&A environment, so even if M&A activity is flat, it isn’t decreasing. This is the main metric I could find that showed that M&A was potentially decreasing, and again it depended on how you measured it. Which is why I love this approach.

Metrics in Favor

The number of metrics showing M&A increasing are much more numerous than that one above. (Another great way to test your data/correlation is to move the dates by a year or two in either direction. If suddenly your correlation disappears, then you were probably cherry picking the data as opposed to finding a true relationship.) Here are the measurements showing M&A increasing:

– Total value of deals has increased not accounting for inflation, across a variety of years.
– Total value of deals has increased accounting for inflation, across a variety of years.
– Total number of mega-deals has increased. (Or at worst stayed flat.)
– The size of the biggest mega-deals has increased, well surpassing the rate of inflation or the entertainment cost of capital.

Another way to look at it would be the “peak year” in any given metric. You should account for inflation here, when dollars are used, but I could point out that the highest year in total deal value was 2016 for announced deals and 2018 will likely be the peak year in closed deals. Here’s my take on when a year set a new “record”, with new peak years in green. (Yes, I can’t go back to 2000 based on my data set and potentially the year 2000 would break all records in M&A in entertainment due to the Time-Warner-AOL merger.)

The final way to manipulate the data is to smooth everything out to five year averages. When I set the terms of the debate, I wanted to use actual years to see the data for myself. But as I learned yesterday, this is a noisy data set and one giant deal can swing any given years data considerably. To smooth it out, I made a table with the five year rolling averages.

Slide13

Since I only have data going back to 2009, the rolling averages got shorter than five years in 2011-2013, and I didn’t bother just putting the year for 2010. Overall, I like how the five year averages show the steady, consistent growth of M&A in media and entertainment.

But there are other cool ways to look at the data that also support that M&A is increasing.

The unscientific way: “Look at these big names”

This is the least scientific way to look at this, but may be the most impactful. It was for me. The table below is just a reminder of all the huge deals of the last six years. Even President Obama’s Department of Justice couldn’t stop these deals. Here’s the table:

 

Slide14

For the data heads, this is the top 2-4 deals in a given year above $1 billion dollars. The main takeaway I have from this table is that big conglomerates in entertainment or communications didn’t need clearance by the Justice Department to make deals. If you insist on only specific quantified data, just learn from this chart that, it will take a lot of really big deals to exceed some of these huge deals. (Also, the 2015 data is light because I just couldn’t find a lot of good summaries from that year. That said the Charter moves were huge.)

The metrics way: Changing the definitions

One fun trend I noticed in reading all the PwC articles was that it seems like PwC changed it’s metrics in the last year for how they define “mega-deal”. In 2018, it seems, they moved the definition from deals valued over $1 billion to deals valued over $5 billion. I see the logic given that you have deals like AT&T-Time Warner merger that are 85 times the value of the Disney-BAMTech deal.

That’s still moving the goal posts and a sign of how much industry consolidation is already effecting entertainment. Having to change a measurement in favor of consolidation is additional evidence that indeed M&A has been steady for years.

The slicing way: Look at sub-industries

One explanation that could show why it looks like M&A is increasing is that perhaps a few sub-industries are consolidating but the rest of the larger industry is not. This isn’t the case.

No matter how you slice the data you see consolidation. Distribution? Yes, in movie theaters, cable provides, and wireless providers. Aggregators? Yes, in TV broadcast channels, in cable networks and in radio channels. Producers? Yes, in TV production, film production and publishing. Conglomerates? Yes, new ones are forming and the big ones are getting bigger.

Here’s two tables to provide just two examples of the above, one from Variety on TV production and one from the Hollywood Reporter on film and tv content producers:

Slide15

Source: Variety/IHS Markit

Slide16

Source: The Hollywood Reporter/Thomson Reuter

Finally, making our prediction

So M&A activity has been frothy for the last few years. Now it’s time to turn that assessment into concrete predictions. In the first post in this series, I laid out two shells of predictions:

The rate of mergers and acquisitions in media, entertainment and communications valued above $1 billion will increase by X over the next Y years.

The dollar value of mergers and acquisitions in media, entertainment, and communications will increase by X over the next Y years.

Let’s start by answering the first question from my introduction, “What is the historic rate of M&A in media, entertainment and communications?” Well, it hasn’t been slow.

Over the last 6 to 8 years, since the recovery, the number of mergers has either stayed steady at 800 a year. The number of mega-deals increased to roughly 16-18 mega-deals. The value of deals peaked at over $200 billion in 2016, was over $140 billion last year and has a five year average of $143 billion. We could pick any of the numbers in these ranges, but the five year averages will suffice to establish our baselines to show the “increase by X” from our predictions.

Now for the second question from the introduction and the key to making the prediction: What is the rate of growth? Is that rate increasing, decreasing or staying flat?

So let’s try to figure out the growth rates in the first place. The problem is that unlike my approach to measurements, having a ton of different ways to measure the growth rates here will make it tougher to find the “true” growth rate if there is such a thing. Still, we can see if the growth rates cluster around a number and how moving our dates around effects thing. So be ready for a few tables.

First, I calculated both the average growth rates (just the arithmetical means of the average each year) and the CAGR (compound annual growth rate) for various time periods for both the value of deals and the number of mega-deals.

Slide17

I think I see some clustering for the “total value of M&A deals” at around 20%. The highest rate goes all the way up to 33%, if you take the 8 year raw average and down to 8.6% if you take the CAGR for the five most recent years, which again is lower because 2016 was bigger than 2017. The growth rates have been much lower in general for the number of mega-deals, around the low single digits. So the growth in M&A is coming form bigger and bigger deals.

I want to pause for a moment on CAGR. On one hand, it captures the growth rate better because it factors in the growth of interest on interest. On the other hand, it can swing a lot more by which two years you choose to start or end from, as the next table shows:

Slide18

So this shows that if anything the CAGR growth rate is slowing down as we get closer. That said, if you took the NY Times numbers I mentioned in the last post that puts M&A at over $300 billion for 2018, then the numbers would jump back up. This is why I can tend to be skeptical of numbers announcing a CAGR of “150%” in some new industry, like VR. Just move the numbers back two years and it would plummet.

With so many numbers to choose from, well we need to make a judgement call. At this point, this is where I combine the measurements I prefer. I like the rolling five year average because I think it smooths out the highs and lows. That said, I like that CAGR and the average tell similar stories. And I don’t just have to choose one number, I can pick two, what I think makes sense for a high and low. So I’m going to use the averages and CAGR for the last five years on the rolling five year averages. With this, we can use the last year’s average and make projections through 2022:

Slide19

So here’s what I like about this slide so much: I really can’t see us hitting $435 billion in the total value of deals by 2022. It gets to the point where I can’t even imagine what $1.4 trillion in deals over the next five years would look like, unless we only have three entertainment companies remaining in five years…which would be really sad for competition wouldn’t it?

But could we hit over $300 billion in M&A in media, entertainment and communications? I mean, that’s what The New York Times says we’re on track to do this year just through the first half of the year. Given some variance for any given year and the fact that some companies will be so large they can’t acquire each other, but smaller ones will get sucked up, yeah $300 billion seems reasonable.

So my predictions have been revised to:

The rate of mergers and acquisitions in media, entertainment and communications valued above $1 billion will increase by $193 billion (from $143 billion 5 year ave in 2017 to $336 billion) over the next 5 years.

The dollar value of mergers and acquisitions in media, entertainment, and communications will increase by 6 (from 17 in 2017 to 23) over the next Y5years.

Phew. There you have it. My quantified, sure to never be wrong, predictions for how much M&A we can expect for entertainment going forward. Sure it took four posts and thousands of words to get to that humble opinion, but hopefully we all learned something.

And there is nothing more to say. Well, I mean I have one or two more ideas. You know what? I’ll have some final thoughts that didn’t make it in here for Monday.

The Entertainment Strategy Guy

The Entertainment Strategy Guy

Former strategy and business development guy at a major streaming company. But I like writing more than sending email, so I launched this website to share what I know.

Tags

Join the Entertainment Strategy Guy Substack

Weekly insights into the world of streaming entertainment.

Join Substack List
%d bloggers like this: