Disney-Lucasfilm Deal Part X: You’ve Been Terminated: “Terminal Values Explained” and The Last Piece of the Model
(This is Part X of a multi-part series answering the question: “How Much Money Did Disney Make on the Lucasfilm deal?” Previous sections are here:
Part I: Introduction & “The Time Value of Money Explained”
Appendix: Feature Film Finances Explained!
Part II: Star Wars Movie Revenue So Far
Part III: The Economics of Blockbusters
Part IV: Movie Revenue – Modeling the Scenarios
Part V: The Analysis! Implications, Takeaways and Cautions about Projected Revenue
Part VI: The Television!
Part VII: Licensing (Merchandise, Like Toys, Books, Comics, Video Games and Stuff)
Part VIII: The Theme Parks Make The Rest of the Money
Part IX: Bibbidy-Bobbidy Boo: Put It Together and What Do You Got?
Yesterday’s article was pretty audacious, trying to estimate 6 years of past revenue and 10 years of future revenue. But the eagle-eyed among you may have noticed I left out a crucial detail:
What about the future? 2029 and beyond? Surely Lucasfilm is worth something then too?
Yes, it is. But predicting the far future is the toughest part. Which ties into one of my biggest pet peeves in valuation. I loathe business models that project near term middling performance (or even losses), but a far future of wild success.
Usually, this wild success is summarized in an outsized “terminal value”, one of the most crucial concepts in equity valuation. It can be hyper-dependent on the growth rate. If the growth rate raises by a point, then the model’s value can shoot through the roof. (And yes, many tech valuations follow this model.)
Yet, terminal values are the best tool we have to solve this problem. If we use them properly. Today I’m adding that last piece to the model, but explaining how I got there and what it is.
What is the terminal value? Well, the last number on the spreadsheet that captures all “future” earnings. Look at my model (this is the median scenario), with the new lines added:
In a word, the “terminal value” tries to capture the value of all future earnings after your model stops. Say you feel confident you can predict revenue out five years. Okay good enough. (I mean no one can really predict revenue, costs and hence earnings, though we still try.) But what about 10 years? 15 years? There are too many variables.
You can see the need for this in the Lucasfilm acquisition. Can I really predict what will happen with release dates of films, even two years out? I already had to remove Indiana Jones 5 from my models. Take another line of business, licensing. If you used the toy sales of 2015 to forecast the future, well you’d be much, much too high. (2015 was probably the peak of Star Wars toy sales.) Back when this deal was signed, Disney didn’t know if they were going to launch a streaming service (I assume) but they still could have sold Star Wars TV series. Possibly for even more money. Not selling to others changes the model.
Here is where the science of modeling has come back to the art. (Which isn’t a bad thing, despite current connotation. Good art is really, really hard to make. Great art even harder.) The traditional way to model a terminal value is to use the future cash flows of the last year of the model, and assume those hold steady into the future. In other words, you make a “perpetuity”, a cash flow stream that continues forever. Alternatively, if your company has a large variance in cash flows year to year, you can use a three or five year average to get the base number. To be even more conservative, you can assume instead of a perpetuity, it is an annuity, where the future revenues only last for a given period of time, say 10 or 20 years. (If you need a refresher on “time value of money”, go here.)
The Specific Terminal Value Calculations
How long will Star Wars be valuable? Davy Crockett was the Star Wars of the 1950s, and it isn’t worth $4 billion dollars. Mickey Mouse has been Mickey Mouse since the 1920s, and he’s worth well more than $4 billion dollars. Which way will Star Wars go? I’m going to assume for a long time. Essentially for decades, but with one scenario where it shrinks over time. Which I’ll control for by tweaking the discount rate. (Either having it grow or shrink.)
To calculate the terminal value, I’m going to use averages of the earnings (EBITDA) of the last 5 years of the model. As this model shows, revenues can be very lumpy. See The Force Awakens boost to 2016 revenue, for example. To smooth that out, the five year average makes more sense.
As for the discount rate, for the base model, there’s no reason to change what we’ve been using. That means our 8% entertainment cost of capital. In a perpetuity, you divide the cash flows by the cost of capital, which acts in essence as a “multiplier”. In this case, a 12.5 multiplier, which is about what the “price to earnings ratio” is for many entertainment companies, hence the cost of capital.
For my low case, I’m using a 10x multiplier, meaning instead of growing, Star Wars decays in value at 2% per year. This is the “Davy Crockett” scenario, and part of me wants to make this ratio even more aggressive. As for the high side, I considered a scenario where the brand keeps growing, using a 3% rate. This would impute a 20x multiplier (8%-3%=5%). But in that case, almost all of the value is in the future. And that’s too high. Star Wars isn’t a future tech monopoly. So I used the same 12.5 multiplier as with the base case, but the high case is higher since it’s average for the last five years is naturally higher.
(Also, I used the five year average, and had no growth, while I moved the terminal value forward one year. This is a difference of something like 3% over a 16 year period. In other words, way too small to really effect the final answer in a material fashion.
Are These Terminal Values Reasonable?
I think so. The theme parks are huge capital expenditures ($1 billion per park) that pay off over decades, as I wrote yesterday. The new Star Wars lands inside the parks are really the longest term plays. (And Disney’s best competitive advantage.) And I see that pay off continuing for a long, long period of time. I also do see toy sales as fairly evergreen, though dependent on movies and TV series.
Still, it helps to triangulate values when possible. To understand terminal value better—I don’t think it is intuitive unless you’re reading a financial text book—another way to think of the terminal value is as the price you would pay to own something in the future. In other words, what would Disney, or some other buyer, potentially pay for Lucasfilm in 2028? They would do the calculation for future cash flows, and derive the value from that.
But they’d also look at the price Disney paid for Lucasfilm in the first place. In other words, after 16 years of making movies and toys and theme parks, how much is Lucasfilm worth, compared to what Disney paid for it? The best way to think about that, is to put the terminal values above back into 2012 terms. See here:
That table is a good reminder that the “terminal value” is only important in that it is taken back to present value. In other words, we need to discount any of our terminal values the same way we have discounted all EBITDA profit in this business model by taking it back to the value it was worth in 2012, when this deal was signed.
(What about the current value? As in the 2018 values for everything? I’ll tackle that question tomorrow.)
Either way, that table in 2012 terms makes the problem pretty stark. In 2029, the lowest Disney would pay to acquire Lucasfilm from itself would be $4 billion dollars. That’s the floor, I can’t imagine the value going lower than that. But they wouldn’t pay that in discounted terms, since that’s the equivalent of $1.1 billion dollars in 2012 dollars. But I still like the initial price to consider what that would be in 2028. Here’s that in a table:
The way to think about this, then, is whether after all the exploitation of the franchise—through lots of movies and TV shows and showering store shelves with toys—is if the brand is still as valuable as when Disney bought it. In this case, Disney has a pretty strong track record. Mickey, Winnie, Marvel, Pixar and now Star Wars are all arguably in better places than when Disney bought them. Even if Disney would pay what it paid for the deal in 2016 terms, that’s still a great deal.
With two different ways to triangulate the value—bottoms up from 5 year average and “waht would you pay for it”—I feel good about these terminal values. And they don’t completely overwhelm the model. (Though, they do account for nearly half the value in some cases.)
The Final Models
Ready? Well, here are our final models, including revenue, costs and the terminal values from above.
The Median Case
The High Case: The Biggest Franchise in History
The Low Case: Star Wars Fatigue is Real
You know what we need now? A huge, easily quotable headline. As in, “Analyst Says Disney Will Make XXX% on its Lucasfilm acquisition.” Well, that comes tomorrow.