The Golden Globes, Ho Hum
December 13, 2007 08:57 AM
The Golden Globes Awards are among the least credible of all the many awards that Hollywood bestows on itself. And these awards are Hollywood's awards to itself, not independent awards granted by an independent press corps. The crop of extremely violent movies they have nominated for awards have been shown by my research and the research of others to be the least profitable; they are stochastically dominated by G-, PG- and PG-13-rated movies. And big budget R-rated movies with a star are the worst of the lot when it comes to profitability.
The foreign journalist association who give the awards are a strange lot. Many have no press credentials or experience in their own countries, as I am told by an experienced journalist who can't get in. So, they are completely dependent on the studios for access to press releases and interviews or any sort of face time with anybody in the business who matters. A "journalist" from a foreign country can't do stories if the studio doesn't like what they write. The reporter might then have to go back to meat cutting or truck driving in an obscure country. So, they tend to be non-critical and almost fawning in their praise of all things Hollywood. They even like George Clooney movies.
LINK · The Movie Business · Comments (0)
Cascades of Ideas and Beliefs
December 5, 2007 10:07 AM
Having worked out some of the dynamic belief and information processes in the movies, I have seen how powerful learning from others can be. I can't resist: 1. trying to feed ideas and opinions into the cascade on global warming to try to move it onto a better dynamics and to reduce the relative frequency of bad ideas to good, 2. pointing out this insightful analysis of the global warming-we have to do something belief dynamic from The OpinionJournal of the WSJ.
The role of journalism, as I have seen it from my perspective as someone who is now and then asked to comment, seems to be promoting consensus through Kahneman's availability heuristic. The more an idea is "out there" and available, the more plausible it becomes. Not true, but that is how our minds work.
LINK · Everything ~ · The Movie Business ~ · Uncertainty · Comments (3)
Contingency
November 18, 2007 09:12 AM
As I think I show in my book, Hollywood Economics, Hollywood has long been a leader in developing payment schemes that are contingent on outcomes. The "nobody knows" principle implies the contingency principle: if you don't know what a movie will earn, then you pay when you do know. That is to say, you base the pay of participants on the outcome through a contingent compensation contract.
The present Screen Writers strike is all about the inability of a platform-based form of compensation to reflect the value of the outcomes from movies and TV shows. As technology introduces new platforms or media through which to consume movies, the old platform based formulas cannot be stretched to reflect the value created in the new medium. So, there is the uncertainty of the outcome of a movie and the uncertainty of new platforms coming along to generate more revenue streams and higher-valued outcomes.
Daniel Mitchell, a professor of business at UCLA, has a brief piece about these sources of uncertainty and the writer's strike, from which I quote an excerpt below. There is a broader point in Dan's article however and that is to take the contingency principle over to government budgeting. Instead of funding programs at fixed levels over future years when the budget is not known, it is superior to fund them on a contingent basis. Bonds are bad because they lock the state into fixed payments which may become difficult to meet when revenues fall. State government revenues do fluctuate, enormously so in fact, so "nobody knows" what they will be in the future. The contingency principle says that you pay when you do know. Therefore, state programs should be funded on a contingent basis as a share of revenue rather than at a fixed level. This applies at the federal level too.
What is good for Hollywood is good for government since they both live in an uncertain world. The excerpt from Dan's article follows.
Read More »LINK · The Movie Business ~ · Uncertainty · Comments (3)
Climate Forecasting, no better than economic forecasting
July 13, 2007 09:44 AM
I have seen a lot of forecasting models over my career; every government agency seems to have one and there once were a host of economic forecasting models that were sold to corporations and government agencies. I began to become a skeptic when I found that one of the most famous regional forecasting models, which had been developed for the state of Hawaii, was a stack of cards in the basement of the government agency it was built for. No one knew how to run it and the only forecasts ever done with it were done by the author for his publications. Graduate students in regional economics all over the world were required to read these publications, but the model was never used by the agency it was built for. From that point foreward, every new project for the state began with a new forecasting model so there was never any validation of the prior model and there was no incremental learning or skill development in the staff. Consultants came and went and the models just piled up, never tested or used beyond the study they were built for (a good thing probably, because they were just trend equations).
Then I reviewed a host of forecasting models for air traffic. It was a nightmare. The FAA's model had failed to anticipate the explosive growth of air travel and Congress got angry at the agency. From that point on, every FAA forecast became optimistic and, still, equally inaccurate. Regional forecasting was even worse. In reviewing airport by airport forecasts, usually done by a consulting firm for the local airport authority, I found that every one of them forecast that the local airport would experience a rising share of the national market. This is an impossibility. So, there was no consistency at all. It was all local boosterism. I solved this problem by modeling the air travel system as a network. So, I was the only economist who had seen that the hub and spoke system of air travel would emerge. But that is another story.
Then I had a look at motion picture forecasting and things really got ugly. It essentially took opinion and mathematized it. It was never "wrong" there was always something wrong with the trailer, or the release pattern, or something else. So, no one ever learned anything. But, still they forecast, institutionalizing group think in the form of Excell spreadsheets.
With this experience, I was not suprised to read this fine review of climate forecasting. They have no clue. Nobody knows anything. Most "expert opinion" is just opinion. The experts are usually worse than Joe the Bartender or you or me in forecasting because they are wedded to a certain view. The dominant forecast for just about any complex process is that tomorrow will be like today. The best forecast of a stock price tomorrow is its value today.
LINK · Complex Systems ~ · The Movie Business ~ · Uncertainty · Comments (1)
Hollywood and Communism
May 3, 2007 10:42 AM
A superb review of Hollywood's strangely ambiguous atttitude towards Communism in Reason Magazine.
LINK · The Movie Business · Comments (2)
Four Fat Millionaires
May 1, 2007 09:24 AM
I saw Wild Hogs the other day. What a waste. I give it four bombs. I have learned now that what Wonder Woman hears from co-workers about a movie is not useful. Here is a review I agree with by Buttersworth.
I actually find few movies I enjoy these days, though I keep trying. My question is why spend the money to put Tim Allen, John Travolta, Martin Lawrence (the only one in the group who can act and is naturally comedic) and the over-rated William Macy in a movie with such a bad script? Ray Liota makes a bad boy appearance too and fattened up for the part. Does he look bad. Which reminds me. Have you seen any of the current generation of actors with their shirts off? Every one of them is larded over and completely off-putting in their incipient obesity. When was the last time you saw a lean actor look like Paul Newman in "Cool Hand Luke"?
The movie has grossed a bit over $135 million in its seven weeks. It can't make money with the cost of that cast.
What was the producer thinking when he/she cast these fat old-looking millionaires in this bad movie? It is the usual inside thinker blindness: they see these four guys as a casting coup that will create industry attention. Maybe, but that doesn't matter when the film hits the sprockets.
Travolta, as it happens, is one of the least profitable actors to appear in movies. Among his more unreasonable contract demands is to receive 8 Armani T shirts ($200 to $300 each) every day of filming. The reason: he is religiously opposed to wearing washed clothing. Even though he owns 5 airplanes he is an environmentally deep thinker, having been quoted somewhere by the dumb media on our need to explore other planets when our world gets too hot to live in. Never going to happen.
Here is an excerpt from an article I reviewed entitled "A Different Tale: Hollywood's unresolved business model" that calculates Travolta's share of revenues of his movies. His pay, relative to the grosses of his movies, is beyond belief. I don't know the author's name as the review process is anonymous (I recommended that it be published with some fixes).
Read More »LINK · The Movie Business · Comments (2)
Seeing things at Disney before they got rid of Eisner
February 28, 2007 08:02 AM
I wrote this as an Op Ed to the LA Times back in 2002. They declined to publish it and I can see why. It was not so long after this that Eisner was removed as Chairman. But, the malaise still lingers and, more troubling, financial markets still over react to forecasts of earnings and "earnings disappointments." They ought to fire the forecasters.
As I write this, Disney is downgrading its quarterly earnings forecasts because its animated motion picture Treasure Planet opened to lower than expected revenues. This event reveals the depth of the troubles at Disney and more---revised earnings estimates and tumbling stock prices have become a common occurrence in financial markets. In light of recent revelations of the errors and misdeeds of financial analysts, I have to ask: when did a company like Disney and when did financial markets generally adopt the belief that financial forecasts are infallible?
In science when a model predicts incorrectly you get rid of the model and look for a better one. And, you don't allow yourself to be fooled by randomness. What Disney did was shoot their movie rather than their financial analysts when they got the forecast wrong. It isn't the movie's fault the forecasters got it wrong and every forecast has an error band around it that should be acknowledged. But Disney chose to announce their forecasting error as a ``disappointment'' in Treasure Planet's opening gross, elevating the model to infallibility and ignoring randomness. The irony is that what began as a forecast error may become a self-fulfilling prophesy if Disney's actions make people believe Disney doesn't think the movie is any good.
The syllogism is strange. Let me see if I can get it down right. First, the accountants or forecasters at Disney predict opening revenues and get it wrong. Then, rather than acknowledging that the model was wrong, they say they are ``dissappointed'' in Treasure Planet's opening. This shifts the blame from the model, where it belongs, to the movie, where it doesn't. Then they use this model (that got it wrong at the opening) to forecast revenue far into the future, predicting all of Treasure Planet's theater revenues through its run, compounding the initial error exponentially. This series of errors then becomes the basis for revising Disney's expected earnings downward.
This is troubling because these kinds of mistakes are made at other studios too and in financial markets in general. First, financial models are not very good at forecasting earnings and they become worse the farther into the future they project. Second, financial variables are volatile. This means that the models will often be wrong because of randomness rather than changing fundamentals. Third, financial models fail miserably when it comes to forecasting motion picture revenues because even God Almighty doesn't know what a motion picture will gross.
On these grounds, there is little basis for Disney's announcement that they were revising their earnings estimates downward and less reason for the price of its stock to move. Since the stock price is the discounted present value of a company's future earnings, a random variation or forecasting error for one quarter's earnings should have little effect on the share price. This could only happen revised earnings estimates for one quarter significantly altered expectations over a long future. But, Bayes theorem tells us that a random variation in one quarter's earnings shouldn't carry much weight. This fundamental analysis suggests there is altogether too much reaction in financial markets to ``disappointing news.''
Where Disney went wrong is in having too much faith in their financial analysts. No one knows what a movie will gross. My research (Hollywood Economics: Chaos in the Film Industry) shows (and every movie fan knows) that motion picture revenues are not forecastable; the forecast error is infinite. There is no correlation between opening revenues and total revenues for any movies that are successful. Opening revenue is only a good predictor of total revenue for movies that die in their second or third week. All successful Disney movies (most of the animated ones are) enjoy long runs so that opening revenues become a small portion of their total revenue. And, all successful movies reach a point (about four or five weeks into the run) where they separate rapidly from the pack. This bifurcation point in the mapping is a signature of chaos in the non-linear dynamics of motion picture revenues.
Given how little information is contained in a movie's opening, the weight studios place on it is troubling, particularly at Disney which is a studio where film revenues depend on long runs. Disney movies endure and do not have to open big. The difference between opening big and enduring is like the difference between a one-night stand and a long relationship.
Joseph E. Levine was the master of the one-night stand. Levine promoted his movies heavily so they opened big (pumping up the yokels it was called). By the time the audience found out it was a stinker, the movie was out of town. The movie ``road-showed'' in a series of one-night stands with the usual results: big expectations, short runs, and broken hearts. Disney movies have never been like one-night stands; they bond with the audience in a long-term relationship and do a lot of repeat business.
There is a difference between a movie that opens big and a movie that is good and enduring. If Disney is trying to make movies that open big, then its heritage is at risk. So it is troubling and, perhaps, symbolic of Disney's malaise that some bean counter there is writing off a \$140 million movie in its first week and management is falling for it. When the message is wrong, shoot the messenger not the movie. And, don't be fooled by randomness.
LINK · The Movie Business · Comments (1)
FLIFF
November 30, 2006 08:43 PM
The Fort Lauderdale International Film Festival was a nice event. I got to award the Mallen Prize to my former graduate student and colleague, David Walls, and give a little speech about him and the implications of his (and my) research. Few heard it as the audience was involved in their own conversations in the large room. All the speakers had this problem. No matter, David and a few guests near the front who mattered heard it.
I also met Jason Squire, the editor of The Movie Business Book, which has gone through several editions. I sold many issues in my Movie class at UCI. It was a pleasure to meet him and all the other new attendees of the Workshop at the Festival. You know which one of the two above is me.
Me giving my talk. A lot of hand waving. Where is the information on that slide?
LINK · The Movie Business · Comments (1)
Movie investors
November 16, 2006 08:05 AM
A paradox in the movies is that their rate of return is around 4 percent and the risk is higher than most industries. Why are investors drawn to a business with such a poor return when the risk is so large? In his LA Times article, Chew, Spit, Repeat, Patrick Kiger draws on my research to talk about how Hollywood chews up new entrants and investors.
The high risk and low return go together quite naturally when you realize, as I show in my book Hollywood Economics, that the return distribution is dominated by a few huge movies. The returns probability distribution is a Pareto distribution. It has an infinite variance and the mean is up around the 75th percentile. This means most movies earn negative or small returns and a few high grossing movies earn the bulk of the positive returns. Thus, the expected value is far above the most likely outcome (the mode of the probability distribution) and the expected value has no predictive value (because the variance about the expectation is infinite).
The expected value is a large sample property that cannot be attained with the small number of movies an investor can produce, or that even a studio can produce in a single year. In these small portfolios, the most likely outcome dominates and it is located near zero.
Why then do investors still flock into the business? They may suffer what I called the "illusion of expectation" in my Journal of Business article: the expected return is dominated by about 6 percent of the movies and this gives a high expectation relative to the most probable outcome, which is the one most movies will earn. It is an error to base investment decisions in this business on the expectation. And, the infinite variance should be a warning to all that any sample statistics, such as the mean from past movies, is volatile and almost 'wild' in behavior. This means you can see anything you want to by selecting the time frame or movies in the sample.
LINK · The Movie Business · Comments (0)
My Harvard talk on extreme statistics in the movies and pharmaceuticals
The figure is a log log plot of the probability density of motion picture revenues; the log of probability density is shown on the Y axis and the log of motion picture revenue is shown on the X axis. This is a stable distribution with a tail weight of 1.33 and an infinite variance. Note the long tail to the right, where the blockbusters are located. These are low probability events of large magnitude that are so far out on the tail they could never occur in a Gaussian world (Titanic was about 20 standard deviations above the mean). They can and do occur in the Levy Stable world of the movies.
Pharmaceutical firms look for blockbusters, just as movie studios do. The Technology Management group at Harvard Business School noted this similarity and invited me to speak on modeling the wild statistics of the movies and what that might mean for the behavior and organization of the pharmaceutical industry. I gave a similar, but less lengthly and detailed talk at UCLA on the same topic. The slides for the talk are available as a PDF file. Unfortunately, it lacks many of the tables and graphics on the movies. But, it does have some fascinating statistical properties of pharmaceuticals that show dramatic similarity to the movies.
Looking through the slides you will find one that shows the growth (and decay) rate of pharma companies is Levy Stable distributed; the distribution is leptokurtotic and has infinite variance. It is known that the growth rate of pharma companies depends primarily on a blockbuster in their portfolio. This is true of the movies as well; a single movie can change the market share of a distributor dramatically. The downside of this blockbuster effect is that a firm's sales can plummet with the withdrawal of a single drug such as we have seen with Vioxx and Bextra.
Click here for the talk Hollywood Economics: Dealing with `Wild' Uncertainty in the Movies and Pharmaceuticals.
LINK · The Movie Business ~ · Uncertainty · Comments (0)
Review of Hollywood Economics
Orley Ashenfelter, Professor of Economics at Princeton and the former editor of the American Economic Review wrote a brief review of my book in Barron's (December 4, 2004). Here is the gist of what he had to say.
"This is a remarkable assembly of two decades of Arthur De Vany's efforts to study the movie industry using the tools of modern economics. Okay, movie lover, if that sounds dry, what about a hard-headed, dollars-and-cents answer to film critic Michael Medved's question: "Does Hollywood make too many R-rated movies?" De Vany's answer: a resounding "yes."
His answer isn't based on a red state versus blue state discussion, but on careful analysis showing that, at the production rates in the period he examined, G-rated movies had lower risk at each rate of return than did R-rated movies. From 1985 to 1996, De Vany found, Hollywood churned out more than 1,000 R-rated movies. If it had made more than a mere 60 (you read that number right) G-rated movies in that stretch, De Vany asserts, the industry would have been far better off economically.
Hollywood Economics brings to the movies what some call the New Economics of Art and Culture. A key ingredient in his approach is what he considers the industry's key characteristic, what screenwriter William Goldman (Butch Cassidy and the Sundance Kid, A Bridge Too Far) calls the "Nobody Knows Anything" principle. The basic idea: In Hollywood, a movie's revenues, costs-and thus returns-are extremely uncertain.
It appears that De Vany came to appreciate both the evidence supporting this proposition and the power of its implications only late in his academic career. (Earlier this year, I met him over dinner to discuss his work. A retired economics professor from the University of California at Irvine, De Vany, who's in the same age bracket as Clint Eastwood, looks a bit like that actor-and every bit as fit!)
The kind of risk De Vany describes is associated with the Stable Paretian hypothesis, in that once you know that revenue, costs or returns have reached level X, the expected value of future levels increases in level X.
Basically, it's about exponentials: The downside is called the "angel's nightmare." De Vany estimates, for example, that if you have spent $20 million on a film and are already over budget, you shouldn't expect to spend less than $32 million before you finish! The upside: Once you reach a certain level of revenue, you can expect even more, which leads to the virtually unimaginable financial success of movies like Titanic.
One of the most fascinating parts of Hollywood Economics is De Vany's analysis of superstar actors and directors. Due to the nature of the blockbuster phenomenon, a director like James Cameron (Titanic was his biggest hit) has made films with total gross revenues that equal those of movies by Ron Howard, who's made twice as many films.
On the other hand, De Vany shows that it's far more likely for a blockbuster movie to produce a superstar than for the presence of a superstar to produce a blockbuster movie.
He also argues that paying superstars a sum equal to the increase in revenues they're expected to generate usually means a movie won't be profitable. Indeed, in one chapter, he shows that not all superstars are worth such payments. Though Tom Hanks, Tom Cruise and Clint Eastwood may be; Jack Nicholson, Robert DeNiro or Bruce Willis may not be.
As venture capitalist Kip Hagopian (a/k/a B. Kipling Hagopian, producer of the blockbuster Ron Howard-Mel Gibson hit Ransom) told me last summer, the most difficult part of the Hollywood process is "to get the movie made."
My advice: Read this book before you try."
ORLEY ASHENFELTER is Joseph Douglas Green 1895 Professor of Economics at Princeton University.
LINK · The Movie Business · Comments (0)
New Hollywood Math?
October 31, 2006 10:10 AM
Peter Bart, the editor of Variety, has a slightly biased but nonetheless realistic comment on investors in the movie business in this Suits vs. superstars.
It is not the indiosyncratic behavior of Hollywood superstars that is the issue. It is that they are paid so much and bring too little to the production value and content of a movie to justify their pay. And, they bear too little risk. Investors bear too much risk and receive too little for it.
Now the firings are beginning, but they do not touch the superstars. Only Tom Cruise has lost his deal. ostensibly over his off-screen behavior, but equally over his pay and the return this leaves for the studio. Profit margins in the movies are low; they are in the range of 4 to 10 percent in a business that is so risky the returns ought to be far higher.
Hollywood has been used to keeping money among members of the club and also to accepting money from outsiders and keeping that too. It does seem to be changing, but see my other posts on hedges and Hollywood.
Bart is right that executives are over-paid too. Someday, someone will write the paper that shows how hard it is to distinguish between luck and talent among corporate executives as I showed in my book for actors and directors.
LINK · The Movie Business · Comments (0)
Hollywood and Wall Street
October 14, 2006 12:25 PM
Hollywood and Wall Street are getting together again NYT. The lure of Hollywood as an investment is the dazzlilng variance of returns. That is also the source of much pain. Most movies will do poorly, a few will earn most of the returns. If you could only predict the outcome, you could earn absolutely stunning returns.
But nobody can do the predictions. There is no formula. Returns in Hollywood are harder to predict than earthquakes or stock market returns. How can a Wall Street analyst do what the science says you cannot do? They can't even do the simpler task of predicting the stock market.
Analysis works for Wall Street only when they are doing arbitrage. Most hedge funds don't do arbitrage any more, they act more like leveraged speculators. A portfolio is no protection because the variance rises with the number of films in the portfolio.
You won't see Warren Buffett doing these deals.
The key line in the article may be this one: "But the new investors are hoping that with enough analysis, they can avoid the fate of some of their predecessors."
That is a fool's goal.
LINK · The Movie Business · Comments (1)
Handbook of the Economics of Art and Culture
September 19, 2006 09:51 AM
Elsevier has just published the Handbook of the Economics of Art and Culture. It is wonderfully edited by Victor Ginsburgh and David Throsby. My chapter The movies is in it and it contains a new synthesis of research that goes beyond my book a bit, though the depth is not quite the same.
LINK · The Movie Business · Comments (0)
Moving the Stakes: Changing Risk Bearing and Reward in Film Finance
September 12, 2006 08:26 AM
This is a summary of my upcoming talk at the Strategic Research Institute organized Film Finance Conference in Los Angeles on November 27th.
Dr. De Vany's talk will begin with an exposition of risk and return in the movie business, demonstrating that it lies on the boundary between quantifiable risk and uncertainty (unquantifiable risk). This is a durable feature of the business over time periods, cultures, genre, budget and virtually every aspect of the movie business. He will show how the stable Paretian model, a non-normal, high-kurtosis probability distribution, captures these features and is the key to understanding the structure, uncertainty, and returns in the movie business. He will also show that many other innovation industries, such as patents, pharmaceuticals, and real estate share these properties.
The second part of his talk will focus more closely on the distribution of risk and reward in the movies to demonstrate that stars are not "bankable" and that they bear too little risk relative to the returns they bring and are paid too much for their contribution to box office revenue. Risk takers receive too little for the large risks they bear. As more sophisticated outside investors enter the business they will move the stakes and reallocate risk and returns. The compensation of stars will change and there may be more Tom Cruise-like firings.
The third part of his talk will show the pitfalls for hedge funds and investors and what they must do to guard against failures to correctly understand the implications of the stable Paretian probabilities they are up against. Common errors are 1. a belief that there is a model that can predict box office or other sources of revenue, 2. a belief that a large portfolio offers protection (they do not because the law of large numbers does not hold in the stable Paretian model), 3. believing that the average return is a good predictor of the mean return, 4. a false belief in the precision of forecasts.
In the fourth part of his talk, Dr. De Vany will show how Extremal Securities are designed for motion pictures and other innovation industries where returns have a non-normal stable Paretian probability distribution.
Dr. De Vany will describe this new class of assets and show their three broad areas of application:
Principal funding --- film finance, oil field exploration, pharmaceuticals
Securitization --- real estate appreciation, artist contingent contracts, stock options
Forward selling --- professionals selling a conditional interest in future earnings
He will discuss how banks, insurance companies, hedge funds, investors and principals will package ES and ES portfolios in each of these market models. The differences between these models will also be described.
Extremal Securities are an instrument for repackaging risk and reward in the film and other innovation industries so that the market can bear the risk.
LINK · The Movie Business ~ · Uncertainty · Comments (1)
Predicting Box Office Success: The Failure of the Neural Network Model
September 7, 2006 04:27 PM
People keep sending me a neural network model published by Sharma and Delen in Expert Systems that claims to predict box office success. I think they are just looking at the title, not the model. The title far overstates what the model accomplishes. The modelis not helpful to a studio or an investor trying to choose movies to make. My take on it? Choose a good story and tell it well. The rest is close to nonsense. A good movie can make someone a star. But, no star can make a turkey into a success. Just finance it in in the right way, not by paying huge costs up front before you know how the movie will do. Pay when you do know, after box office and other revenues come in. That is to say, pay stars on contingent contracts. And bring the movie in on budget. Money is not creativity.
Neural networks are useful classifiers (things that say "this thing is black or white") but they cannot even begin to approach humans as classifiers. A neural network is like a committee of idiots that you give a signal to. The answer can only be given in a discrete category. A simple message going in, really simple decision makers inside the black box, and a really simple answer coming out. After letting the committee of idiots give responses to the signals, you make some of them more influential than the others. You do this by giving their opinion more weight in the neural network. Initially, this will be the loudest idiot, but eventually if you keep changing the influence each has on the others, the choices get better. But, remember, it is a committee of idiots, no one in the black box has seen any of the movies that are classified or any movie at all, ever. They are just supposed to classify the movies based on some information about them. The information is given in discrete chunks too; like, this movie has a really big star (based on past movie grosses). So it could be Sandra Bullock playing Terminator and the neural network would not know the difference between her and Arnold in the lead role. [Come to think of it, it might be a good movie. Not difficult since there has been no good Terminator movie.]
The authors "train" their network on 9 out of 10 sets of movies and then test what they know against the remaining set. The movies cover 5 years. Now get ready for a shock: they discover that the predictors of box office success are high star value, high technical effects, and number of opening screens. The latter variable is the only one treated as a continuous variable and was not discretized as the others were. Anyone could have told you that.
If you were going to use this information (by the way they do not show you the magnitude of the effects of increasing screens or adding effects), would you take an arbitrary movie and say all I have to do is add an actor, technical effects, and book more screens and I will have a hit? Or, would you say the neural network has simply recognized how Hollywood positions movies in the market place? Which means that Hollywood already knows and the neural network (it is a committee of idiots after all) has discovered it.
What their "high technical effects" variable measures is animated and science fiction movies. There are three categories, high, medium and low. So, this is picking up animated movies first of all and then Star Wars and the like. Hardly news, either as a prediction or as an acknowledgement of how these movies are marketed.
But, they mislead in claiming higher accuracy...
LINK · The Movie Business · Comments (0)
Movie Stars and Profits
August 28, 2006 09:04 AM
After quite a few emails and phone calls with the NYT, this story about stars and movie profits is out. I see a lot of my friends and colleagues in the story; I know and have talked extensively with every one mentioned but for Alfred Marshall (I may have spoken with him in my sleep I spent so much time reading his work in grad school).
There is nothing new in the story, but basically it is right. What troubles me is how little of the depth comes through. And the amount of time I, and doubtless the others, spent relative to the amount of information revealed in the story. I am cutting back on time with reporters; it has been a large drain with little to show for it.
Anyway, here is the story.
LINK · The Movie Business · Comments (2)
Nobody is a star forever
August 24, 2006 05:29 PM
If you have read my Trimming Hedges in Hollywood post, you will find this to be a familiar story Where does Cruise work now that Paramount fired him, or I told you nobody is a star forever.
Even Cruise can't be a star forever. I expect this story to eventually appear about Cruise because I did the research. And it shows that nobody is a star or a genius forever. If your movies lose money you are gone. If you structure your deals so that you make money when the studio, who risks the money, loses, then you are gone. It should not be otherwise, even in the unreality of Tinsel Town.
I estimated the half life of a star. It is not long, but it is an odd statistic, rather like Mandelbrot's Cemetary of Poets metaphor. A poet who dies leaves about half of her work unfinished (the conditional expectation is proportional to what she has done and the coefficient is about 2). In the movies a dying actor (and they die in the business before they die) leaves just less than half of her work unfinished. A dying director (and they die in the business before they die too) leaves about two thirds of her life's work unfulfilled. Don't we all?
Harold Vogel, who complimented my book by saying that I had explained the industry in a way that had never been done before and who features my research in his Entertainment Industry Economics, makes the right, and most important, point in this story. The first dollar ought to go to the one who puts up the money. This had been true for some time since the studio got its distribution fee, first dollar, while they were the one who put up the money.
But, when an outside investor comes in, he/she will insist on first dollar and who could argue? They take the risk. The star takes no risk, unless he/she is on a contingent contract in which pay depends on outcome. I think many stars realize that is risky too. They would prefer a fixed payment, but they make their most extravagent paydays when they take a contingent pay off. They are few in number, but huge in total pay. They ought to have a way to lay off some of this risk and retain part of the high returns too, when such events happen.
Extremal Securities® (patent pending) and an exchange for them will solve these problems. But, that is another story I will develop when the exchange is up and running. I hope to announce the opening of the Extremal Securities Exchange® at the Fort Lauderdale Film Festival when I speak there in November. My student, David Walls, is the winner of the Mallen Prize that I won some years ago.
The NYT will have a story Monday on this funny business and my thoughts about it (with others too I would guess). I spent quite a bit of time today with the reporter on the phone. Smart guy. It will be worth reading I think. But, you have seen it here and in my book, Hollywood Economics. You do have a copy, don't you? It is about wild statistics, the improbable, and how to cope with them. My themes about chaos are there and it isn't the usual pop science stuff. This is real and responsibily tested and developed.
So, is there anything this says about fitness? Yes. You will see how complex dynamics are the stuff of life, at least my life and the way I live it.
LINK · The Movie Business · Comments (0)
Spinosaurus
Some body is bigger and badder than T. Rex.
When will Spinosaurus have his own movie? He makes even that other giant lizard, Godzilla, look silly.
LINK · Evolutionary Fitness ~ · The Movie Business · Comments (1)
Trimming Hedges
July 16, 2006 10:23 AM
Hedge funds are supplying substantial funds to Hollywood now. They employ different approaches, but in the end they are bringing outside funds to the movie business. The movie business has what I described as a "never ending need for money" in Patrick Kiger's article linked below.
You will see various estimates of profitability and of the internal rate of return. Eisner once put it at 4%. Others, quoted in the articles below, put it at 14%. On a single film, the internal rate of return can reach dazzling levels, but that has little meaning because the studio has to pay the bills for many productions that have dazzling negative returns.
So, what is the real story? Nobody knows. Vogel, Entertainment Industry Economics, Sixth Edition says that "a few big winners pay for many losers." I, Hollywood Economics, estimated that 5% of films earn about 80 to 90% of profits in the industry. Vogel's calculations of operating margins for the major studios fluctuate between 4 and 18%. And he points to my research to show that there is no "average" film nor average rate of return. Everything is volatile. He quotes from my book, "There is no typical movie and averages signify nothing..."
Few people understand three points about the stable-Paretian distribution that characterizes the movies (and a lot of other industries):
First, the average is dominated by large scale events and moves all over the place. In fact, it is also a stable-distributed random variable with its own expectation and infinite variance.
Second, a stable-distributed random variable is stable only in the self-similar form of the distribution, not stable in the sense of the law of large numbers. This means it looks the same from every point of view. Hence, there is no escaping the wild variance even if you alter the budget, the advertising, the cast or anything. You always have wildly divergent outcomes.
Third, self-similarity implies that small pieces of the time series, say, of revenues or rate of return, have the same distribution as large pieces and they all look the same. This self-similarity is rather like you see in the stock market where the ror series on the 10-minute window look like the 24-hour window, which look like the 30-day window, which look like the year to year window over 2, 5, 15 and 20-year windows. They all show the same volatility, never settling down as they would if they were normally-distributed random variables.
Self-similarity means you never escape the underlying wildness (fluctuating average and infinite variance) no matter what you do (see my "The Movies" in Ginzburg and Throsby eds., The Handbook of the Economics of Art and Culture (2006)).
Then there is the other point about the internal rate of return. The equation often has multiple roots or solutions. The internal rate of return is the discount rate that sets the present value of the revenue stream equal to the present value of the cost stream. According to Descartes' rule of signs, there will be as many roots (internal rate of return solutions) as there are points where the series crosses the origin. One has to be very careful in checking for multiple solutions in the long and complex series that constitutes projected revenue and cost flows for a movie or a production slate. And, of course, there is much room to affect the internal rate of return by placing revenues or costs at different points in the streams. Finally, all this is projection and guess anyway. So, you can get just about any rate of return you want if you put in the right numbers. Or, ignore the multiple roots of the equation and just pick the biggest, positive one.
Finally, an apparent self-contradiction exists here. If the studios earn an internal rate of return of 14% to 28%, why must they depend on outside sources to finance their productions? Maybe they don't require outside funding, though Epstein below argues they do. But, they are happy to take the funds either way. Aleck tells me he could stand at Rodeo and Wilshire at 3 am saying "I have money to finance movies." and a crowd would form.
Both Vogel and Epstein (The Big Picture) conclude that the heart of the movie business is finance and distribution, the "clearing-house" idea. Both suggest that production and marketing cost risk is transferred to outside investors.
I am not so sure. Production risk is more easily controlled than revenue risk. They are transferring some of the revenue risk, the wildest and least controllable risk a film faces. Reallocating risk is one of the primary purposes of financial markets. Nothing wrong with that. But, why the allure of investing in the movies? I suppose some players want to come in and mingle with stars and call some of the shots. They either buy baseball teams or invest in movies.
But, there is a more subtle issue. It is the illusion of the expectation (see my book). The average of revenue, rate of return, profit and other variables is far from the typical outcome. The most probable outcome, the mode, is a negligible value. The average is up around the 75th percentile. So, there is a large difference between the most likely outcome, a low number, and the mean, a number completed dominated by a few huge events far out in the tail of the distribution. Many businesses with this feature draw more investment than is justified by the returns. Investing in a portolio is of some help. But, it doesn't reduce the variance. In fact, the variance increases with the size of the portfolio. This is where Extremal Securities, Extremal Security Exchange, and Extremal Security Portfolios (trade marks of Arthur De Vany, patent pending) come into their own. You have to get it right though.
Some articles about Hollywood hedge funds in the media that are worth a glance.
Read More »LINK · The Movie Business · Comments (2)
Extremal Securities ®
July 15, 2006 08:30 AM
Now that I have trade marked Extremal Security ®, Extremal Security Exchange ®, and Extremal Security Portfolio ® I will do a series of posts explaining their advantages. The actual design work in creating an extremal security involves a variety of tests and mathematical/statistical analyses to characterize the structure of the outcome space and the required attributes of the extremal security and extremal portfolio. Optimization of the attributes of the extremal security and extremal portfolio then are done to achieve the desired gain/reward exposure from the frontier of efficient outcomes. These methods are part of the new financial technology embodied in the concept of extremal security (patent pending).
It turns out that a number of funds are working in the movie business creating extremal security portfolios without acknowledging it as my invention and even claiming the extremal security portfolio as their own idea. Now that I have learned this, I'll have to look into it a bit. Sounds like intellectual theft to me.
Of course, they are playing with small potatoes. Extremal securities could transform finance and financial markets.
LINK · The Movie Business ~ · Uncertainty · Comments (1)
Financing Movies, Part 3
July 13, 2006 10:59 PM
This post is about Extremal Securities ®. I designed these securities in my book, Hollywood Economics (2004). I am seeking a patent on the design because, though it is grounded on contingent claim theory, it is novel and of great use in innovation industries, of which motion pictures are only one. Extremal Security ® and Extremal Security Exchange ® are my trade marks. The extremal security and a portfolio of extremal securities is my invention. Both inventions are available for licensing.
I built these extremal securities from the Arrow-Debreu idea of state-contingent claims, that is a distribution of claims on revenues over probabilistic states of the world. An old idea I put together with the stable Paretian distribution to create an instrument Hollywood and other innovation dependent industries need. I have lectured about this idea for several years, most recently at Harvard and UCLA and to the Orange County Intellectual Property Bar.
I showed that other industries with stable-distributed outcomes, such as pharmaceuticals, patents, and books could be financed and even reorganized through the use of extremal securities. Aleck Grishkevich is licensing the invention and service mark from me to set up an extremal security exchange where contingent revenue claims may be listed by anyone who has a verifiable contract, patent, innovation, or new venture to finance.
So, here is the text from the Epilogue of my Hollywood Economics, pages 271 and 272, where I laid out the general design. The details of the design are subtle if the security is to offer some degree of central area pooling and capture the massive, lower-probability outcomes in the Paretian tail of the stable distribution. Other features of the design provide access to extreme tail outcomes while decreasing the variance. The design works for stable distributions, Paretian tail of stable distributions, extreme value statistics. The log-normal distribution can also be carved into tail events readily because, 1. it is in the stable class, and 2. for this distribution one can use conventional options pricing models.
"Studios are really banks that have a distribution arm for leasing movies to theaters. They finance movies based on what they think they will cost and how much they expect they will make. All the risk is borne by the studio unless they grant points to artists in exchange for a lower fee. To shed some risk they might pre-sell one or another revenue stream. This turns risk into cash. A film has many such streams and a studio can choose which ones to sell for cash and which ones to keep. Turning these unknown streams into cash may seem comforting, but the studio may regret taking cash upfront instead of retaining some part of the uncertain gain. If the movie is wildly successful, they will have sold the stream for a pittance, transferring to the buyer all the uncertain, but potentially very large gains. Remember, the rewards lie in the far upper tail of the probability distribution. The heavy tails tell us that a studio should always retain some part of the upper tail of any revenue stream they pre-sell because that is where the really big gains are.
There are better ways to finance movies. We know that we should take portfolios of projects and that there are extreme statistics to deal with. So, investors in motion pictures must manage their risk through the use of portfolios; these can be created by packaging projects together into one investment package. Or, the projects can be securitized in some manner so that an investor can manage risk by diversiying securities.
Except for the odd portfolio offering, such as the Silver Screen limited partnerships (where the junior partner typically does not fare well), few movies are financed by packaging them as part of a portfolio. Studios implicitly build portfolios of from four to fifteen films in their annual productions, but they do not use portfolio analysis in picking movies because they greenlight them individually.
So, is there a better way? We can design securities to solve many of the problems in film finance. How should such a security be designed? Design must be grounded in the specifics of motion pictures. What is to be securitized is the revenue or profit of the revenue streams that flow from a movie for the whole of its economic life. It is desirable to securitize a stream that can be monitored by the investor and verified by a court. Revenues, as reported by any of the independent sources of box office revenues, is an ideal variable on which to condition payoffs. Profits are more controversial and are difficult to allocate over a portofolio (this leads to what has been called the studio accounting problem). Conditioning security payouts on revenue is more likely to draw investor support because studios or producers have less room to influence revenue than profit and independent sources publish information on revenue. By reducing moral hazard, revenue-based securities are more likely to support a market. The importance of a market is that it would bring greater liquidity to motion picture securities and, thereby, increase their value.
Read More »LINK · The Movie Business ~ · Uncertainty · Comments (1)
Motion Picture Finance, Part 2
July 11, 2006 08:31 AM
This is a picture of the distribution of estimated motion picture profit. The Stable distribution is compared with the data and with the Normal distribution. The Stable distribution is a superior fit over the visible range of the variables. What you cannot see is the vastly superior fit of the Stable distribution in the far left and far right hand tails where the Normal distribution has essentially zero probabilty for movie outcomes that actually occured. Note that the distribution is not symmetrical either; it has a heavier tail to the right (positive profits) than to the left (losses). This is because losses are limited by what you spend, but profits have no upper limit.
Now have a look at the variance of profit over time. It has the ragged shape of a fractal or mountain range where the variance takes leaps and falls of every size with no typical value. That is the nature of risk on one hand but of return potential on the other hand. For it is in the variance that opportunities for huge profit occur. These low probability, massive events dominate profit in Hollywood. The Stable disribution is the key to this pattern as well.
These pictures give some hint of the wildness of outcomes in the movie business. It is a business where the extraordinary is ordinary and where the small probability, massive events dominate everything. I could show similar pictures of wild statistics on the revenue side as well. The Paretian Stable probability model (and the models of the dynamics that produce this distribution) captures the very essence of the movie business. Among the many properties it implies is Goldman's famous "nobody knows" principle.
Nobody knows what a movie will gross in US theaters, in the world theatrical market, in DVDs, and other revenue windows. If they knew, the business would be easy and few films would ever lose money. But, many or even most films do lose money (it is hard to do the accounting, but many estimates show this).
My research and that of many others now has shown that the nobody knows principle is true. But, there are highly stable patterns as well, if you know where to look for them. I develop them at length in my book (A. De Vany, Hollywood Economics, 2004). Thus, even though motion picture revenue streams are statistically ``wild'', with non-stationary averages and infinite variance, there is a deeper pattern to the underlying probability distributions. With a knowledge of these deeper, universal patterns (I show that they hold for all time periods and countries and for many different variables), it is possible to design effective instruments to gain some pooling among outcomes near the mean while retaining a claim on the huge returns that come from rare events in the upper probability tails when they occur. You can almost "have your cake and eat it too" with the right mixture of contingent claims on parts of the revenue streams a movie throws off.
Revenue streams are sums of random variables, aggregated over theaters, DVDs and other venues and they converge to the stable class of probability distributions. The Normal or Gaussian distribution is a member of this class, but it is the only one with a finite variance. The Normal or log-Normal (as assumed in the options pricing models) do not fit the movies. In fact they are so far off that movies like Titanic or Harry Potter should never occur. Only the stable probability distribution captures the "wild" behavior of the movies. It fits the movies like a glove and it reveals deep patterns that some try to ignore at their peril.
The stable Paretian distribution is the attractor of movie revenue streams and is independent of initial conditions. The dynamics have the stability property and, thus, are also self-similar. The basin of attraction of the process is determined wholly by the tail weights of the distribution. The distribution has ``heavy'' tails and completely non-Gaussian probabilities in the upper tails, where the main revenue generators lie. These remarkable features explain a great deal about the movie business and provide the basis for new ways to finance movies.
It is not hard to see the non-Gaussian nature of the movies: Forrest Gump earned US theatrical revenues 10 standard deviations above the mean; Titanic earned revenues 20 standard deviations above the mean. These events would not occur in billions of reruns of the Earth's history if events were Normally distributed. But, events occur on all scales, with no typical outcome in the movies. The mean is dominated by events of massive scale ("blockbusters") and so is the sample average. The sample variance increases in the sample size and the theoretical variance is infinite. There is also a time-dependency in the statistics, partly because of the influence of extreme events. These and other properties are confirmation of Goldman's proposition. The business is wildly uncertain and, for that reason, it offers unusual investment opportunities.
There are at least two reasons for these opportunities. First, the studios and producers know they face great risk and find many ways to lay it off through their deals. But, they only have so many ways to do this in today's financial markets. Primarily, they adjust their risk exposure by overspending on stars and in the way they release movies on many screens with large advertising expenses. Research shows these are not effective and may lead to a rapid playoff before the film can get its ``legs''.
The second way they adjust their risk exposure is through contingent contracting or what are called participation contracts. If no one knows how much a film will make, then it makes sense to condition pay or pricing on outcomes. This is the option principle of delaying the pay or pricing decision until more information becomes available. Thus, many contracts are written conditional on the revenue streams a film realizes over its various runs through its venues. Director fees, star pay, distribution fees, rental charges, and a variety of other pricing and revenue variables are determined after the film plays in its venues and streams of revenue come in. The long history of the industry in structuring deals this way lays a foundation for the new financial instruments we will bring to the industry.
Studio models focus on forecasting expected values and virtually ignore the variance. And the forecasting is seldom sophisticated and usually wrong (how could it be otherwise given the nature of the statistics?). The sample average is a poor forecaster of the mean in this business. The sample average is highly non-stationary (it even is stable distributed) and a poor estimator of the mean. It is dominated by a few rare events. The mode is well below the mean outcome and the mean is up around the 75th percentile. The distribution is highly kurtotic; excess kurtosis is 15. If these expected value models underlie studio decision making, then they will undervalue the upper tail events which are the heart of the business.
Thus, at one time studios pre-sold the foreign distribution rights to their movies for a flat fee, independent of foreign revenue. The cash advance enabled the studio to finance the movie, but at the cost of giving away the upper tail events that dominated revenues. Eventually, the pre-sold foreign distribution fee was made contingent on US theatrical revenues. But, this brings the money in too late from the point of view of the studio who needs the money well before foreign or domestic distribution in order to finance the production. And it is one factor in the narrower release windows we are beginning to see.
Another factor that distorts studio decisions is that, in this business, the variance is the key to high returns. The Sharpe ratio, variance over mean, that determines a lot of financial decision-making is infinite in the movie business. So, perhaps a lot of hedge funds stay away from the business. But, this is an error. The heavy tails make for the infinite variance and the higher than Gaussian probabilities of large events in these tails drive the mean. Heavier tails, more variance, larger outcomes, higher mean. It is not a Gaussian world and that property can be used to give investors a piece of these hugely profitable upper tail events. In a world where arbitrage has driven prices to low Sharpe ratios, an industry like the movies offers one of the best opportunities to gain higher returns.
The risk associated with these huge potential returns can, in turn, be managed by the way tail events are linked to investor returns. We can structure things so that investors can buy tail events. They never lose more than what they pay for owning a contingent outcome that could pay thousands of times the price of the instrument. I call these instruments extremal securities. But, they are rather like call options on uncertain revenue streams. Each stream in every venue can be colateralized by creating conditional claims on a share of revenue at respective tail events.
Read More »LINK · The Movie Business ~ · Uncertainty · Comments (0)
Motion Picture Finance
July 10, 2006 07:12 AM
I have a series of posts coming this week on financing movies and paying stars. It will develop the issues in a way that lays the foundation for a new market that my partner Aleck Grishkevich and I are setting up.
Our market will let investors share in the high kurtosis of motion picture revenues, but at low risk. It will offer a means for compensating stars that sharply reduces the risk for producers, but lets stars share in the large returns that a few movies realize. It will offer a mechanism for pooling returns among generations of new actors and reduce the uncertainty individual actors face. It can become a primary means for financing training and the early stages of an actor's career.
Banks, funds, and institutional investors will find in this market a tool for capturing the high returns that can be generated in the movie business while taking little risk. Few other industries offer the dazzling variance in returns that has been driven from a picked-over stock market where investors try to minimize variance for expected return. In the movies and other heavy tailed businesses, the expected return is dominated by the high variance, large scale events. Running away from variance sharply reduces expected returns and this is not in the conventional risk avoidance fashion of the CAP model. Expected return rises with variance because it is the primary feature of the distribution of motion picture revenue outcomes in which the expected return is dominated by the large scale, low probability events.
We have solved the puzzle of how to participate in this extreme variance market with little risk. We have designed contingent claims on revenue streams, following the theory and evidence in my book Hollywood Economics, that sharply reduce investor risk while allowing parts of the extreme returns that come to those 5 to 10% of movies that earn 90% of profits in Hollywood.
The market will bring liquidity to what is now a very thin and underdeveloped mechanism for paying actors and financing movies. With liquidity comes higher value for the parties on both sides of the deal and lower pricing risk. A market of the kind we are creating has the potential to revolutionize how movies are financed and open the way to more creative productions. Book publishing, patents, and pharmaceuticals---all heavy tailed, high kurtosis industries---could also benefit from a market of the form that we first will establish for motion pictures and actors. More coming...
LINK · The Movie Business ~ · Uncertainty · Comments (4)
Leonard Mlodinow's "Meet Hollywood's Latest Genius"
July 2, 2006 06:28 PM
Then again, in 6 months he could be a loser. Box-office success is more random than you may think.
By Leonard Mlodinow, Special to The Times
July 2, 2006
CHAOTIC: (ka ät ik) adj. 1. in a state of chaos; in a completely confused or disordered condition 2. of or having to do with the theories, dynamics, etc. of mathematical chaos 3. how Hollywood really operates.
You have to love that definition; John Cassady had a similar take on my work in the New Yorker in his "Chaos in Hollywood". Unfortunately, this article is not on their web site.
Leonard spent a lot of time interviewing me and others for this fine article about Hollywood's uncertain grasp on uncertainty. I am quoted often in the article. His exposition of uncertainty or not grasping the role of luck in explaining events is lucid and shows he is a master of the subject, having taught a course on it at CalTech this past year.
Like most of us, Hollywood executives think of successes as due to their skill and failures as bad luck (or a bad trailer, or a dumb director, or whatever). As I say in my book, Arthur De Vany's Hollywood Economics (plug), it is a business where learning is hard because the movies are a complex system where causality has little meaning because so many factors are intertwined in a chaotic dynamics. Nassim Taleb's Fooled by Randomness is a readable and insightful book on the subject.
I may never work in this town again after what I said about the executives in Hollywood. Then again, the end of the story shows there may be hope if the executives quit worrying about their jobs and start trying to make great movies. If only...
Len's article is here Meet Hollywood's Latest Genius (but hurry before he is gone). That's my subtitle, not Leonard's.
I have two of Len's books, having bought them after we had talked for some time via email.
LINK · The Movie Business · Comments (4)
Upcoming Article in LAT Magazine
June 26, 2006 10:00 PM
Leonard Mlodinow is a physicist who taught a course on uncertainty at Cal Tech this past academic year. He is the author of several fascinating popular books on physics, one with Steven Hawking.
He has written an article exploring the themes he taught in his course and ideas expoused by Nassim Taleb and myself on uncertainty and the human failing to be fooled by randomness. His article is primarily about uncertainty in the movies and is due out July 2 as the cover story for that issue of the LA Times Magazine. He tells me my work is a prominent part of the story.
From the interviews we did I gather that he has put some very interesting information together on who was hot in Hollywood and isn't now. Which shows that no one beats the odds for long. In my book, you all know the title by now, I show that there is almost no difference between the number of heads you would get in a row flipping a coin and the number of movies that a director makes. In other words, there is not much left for that thing called talent to explain in the differences between the careers of directors. I am looking forward to the article. I will post the link if the Times puts it on their web site.
LINK · The Movie Business · Comments (0)
Regrets over Brokeback Mountain
April 9, 2006 11:59 AM
How could anyone who was attached to the movie regret how successful it was?
Apparently, Randy Quaid could. He seems to have taken the wrong kind of contract for his participation in the movie. I know none of the details, but Michael Kim has sent me his interesting piece from the film blog Libertas. See Applications of De Vany's Hollywood Economics.
LINK · The Movie Business · Comments (3)
How Hollywood Chews up Naive Investors
March 23, 2006 03:50 PM
Patrick Kiger did a nice job putting his interview with me and others on how Hollywood draws investors and what happens to them if they don't know what the business is like or what the uncertainties really are. The town has taken money from a lot of shrewd business men from Hearst, to Hughes, Joe Kennedy, Bronfman, Morita (head of SONY Corporation), and Mercier when he ran Vivendi (a highly connected French water and sewage company---you would have thought that might have worked, a lot of the sewage could have gone straight to the screen).
I put the Time Magazine article under the Research link at the top of the page. See Kiger's fine article "Chew. Spit. Repeat."
Not only have large corporations and moguls from other industries been chewed up, largely because they do not understand the business or the stable Paretian distribution. Many smaller investors and would-be creative people are chewed up too. Here it may more often be dishonesty rather than the deadly outcomes under the Pareto distribution, though both are at work.
Aleck, who has never sat in a class with me, but must be my best student on Hollywood because he combines experience there with what he has learned from my book [he has lost a lot of weight and inches too from following Evolutionary Fitness], told me a Hollywood joke, a wry comment that a lawyer told a client. "You should take his money before he learns you are going to steal it from him."
LINK · The Movie Business · Comments (4)
Hollywood's Simplistic Plots
December 29, 2005 12:09 PM
Victor Davis Hanson nails it with this comment on the movies Munich and Syriana.
LINK · The Movie Business · Comments (2)
An Interesting Day
November 29, 2005 07:15 PM
I got a call from a Scottish newpaper today, the Sun something, about the latest spate of remakes of old films. King Kong was on the reporter's mind, but another fact was that more than 15 remakes of old movies are slated for release. This is an increase, according to the reporter, over last year's production of about 10 remakes, and slightly less the year before. It seems there is a pattern here. Has Hollywood discovered that remakes are profitable? Maybe, according to some perceptions of studio executives.
If one remake was profitable a few years ago does that mean that 15 this year will be? Or are producers and studios hoping to ride on their perceptions of what King Kong might make? Who knows. Certainly they don't know. Don't expect some kind of model or analysis to indicate that remakes are profitable and that 15 in the market will pan out. It can't be true and there is no model. If one was highly profitable a few years ago, when it was the lone remake, how could it follow that more than one or two of 15 remakes will be?
Most of them will not succeed. What this pattern shows is an old pattern of Hollywood grasping at previous successes with no knowledge of how they will turn out in the future. We saw this recently after the success of Shakespeare in Love which was followed by serious pronouncements by Hollywood and media reporters that the next big thing for the studios is to make character-based movies with slightly flawed characters. Virginia Postrel wrote a great column for the New York Times about the "Next formula" that year using my research to show it was a faint hope for managing the profound uncertainly that the movies must always live with. There is no formula.
It now appears that Hollywood has run out of creative ideas that can be marketed in the blockbuster manner and is reaching for past successes for a repeat. It can't happen. Even a remake is a completely different movie in a different time and place. Nobody knows how it will do.
I also got a letter from Bud Selig, Commissioner of Baseball...
Read More »LINK · Everything ~ · Sports ~ · The Movie Business · Comments (2)
Work
November 25, 2005 07:50 AM
De Solla Price found that half the work done in scientific publishing is done by the square root of the number of scientists. Try that in your own organization doing other kinds of work; according to the law half the work in an organization of 100 people is done by 10 workers.
Pope John Paul had his own version of the law. When asked how many people worked at the Vatican, he replied "About half."
And producer Robert Evans had his own version for the movie business. He said that at any time most of the people in Hollywood are not working.
Now a guy named Jim has come up with a version of the law for the United States. It goes like this;
"The population of this country is 237 million. 104 million are retired. That leaves 133 million to do the work. There are 85 million in school, which leaves 48 million to do the work. Of this there are 29 million employed by the federal government, leaving 19 million to do the work. 2.8 million are in the Armed Forces, which leaves 16.2 million to do the work. Take from the total the 14,800,000 people who work for State and City Governments and that leaves 1.4 million to do the work. At any given time there are 188,000 people in hospitals, leaving 1,212,000 to do the work. Now, there are 1,211,998 people in prisons. That leaves just two people to do the work. You and me.
And you're sitting at your computer reading jokes."
Read More »LINK · Everything ~ · The Movie Business ~ · Uncertainty · Comments (2)
Do Directors Have Tangible Value?
November 10, 2005 09:02 AM
Sometimes my research takes me into strange areas. I was an expert witness in the divorce of John McTiernan (director of the movies Die Hard2, Hunt for Red October, and The Thomas Crown Affair among others) and his wife of 8 years Donna Dubrow, a producer. How did this happen?
Well, the issue was how significant were John McTiernan's reputation and earnings as a director. Dubrow's attorneys argued that he had gained reputation during the years of the marriage and that his wife should get some of the gain in the value of his human capital as part of the settlement.
Whether she should get part of his gain in human capital was not my position or topic even. I dealt with estimating his earnings and the probabilities of his making future movies, with allowance for the uncertainties of the movie business. Both the Superior Court which first decided the case, calculating that McTiernan's gain in "goodwill" was $1.5 million, and the California Court of Appeals decision, which overturned this part of the decision, cited my testimony. The decision can be obtained through many sources and is part of the public record. It is available Court of Appeals Decision.
The Court of Appeals overturned the "goodwill" part of Judge Montes decision on the grounds...
Read More »LINK · The Movie Business · Comments (1)
Fights in Movies
June 28, 2005 10:06 AM
They are almost always bad.
I read a long article years ago by a doctor who analyzed movie fights and killing. His main point was that humans are extremely difficult to kill and that the movies dramatically cheapen human life by making it appear that humans die easily. They don't and an evolutionary perspective would make you understand how hard it is to kill a human. Those who went easily didn't leave their genes behind.
People who are on PCP are almost super human; they can tolerate great pain and injury and still exert enormous force and power. They are extremely lethal.
The legendary Berserkers, the wild killers of the Viking invaders, were on some form of drug rather like PCP according to scientists who have studied them. They were also a semi-secret, cult-like group who practiced hypnotic rituals before their attacks. They didn't wear armor, just fur covers, unlike the other Vikings.
I suppose they were sold some kind of Valhalla story, rather like the fundamentalists of our day. Another instance of someone forecasting your future for you to their benefit, not yours. Of course, there is no Valhalla on the other side as there is no other side. The mind dies when the brain dies and so do any thoughts or sentience that reside there. The cheapening of the present, the only place where you live, relative to the future is a common trick of those who want to manipulate you. The future is completely unpredictable. And you are only alive now.
Interestingly, body builders are not good fighters, at least if you judge by their movies. Steve Reeves was awful; he lunged and posed more than he fought. Reg Park too. Arnold isn't at all convincing. You action movie fans could name others I am sure. Harrison Ford is a more convincing fighter than any major star, but his victims drop far too easily. If they didn't, the movie would be one long fight. And, no story (come to think of it, there is no story anyway.)
It would actually be an awful movie to see just one victim die in a realistic way for it would take most of the movie.
The fighters in Troy who were said to be bashing each other pretty hard were the Bulgarian weight lifters, probably Olympic lifters rather than body builders, though I don't know and I didn't see the movie.
LINK · Everything ~ · The Movie Business ~ · Uncertainty · Comments (17)
Troy and Me in Time Magazine
June 27, 2005 03:34 PM
Do you remember this movie? It was heavily advertised and featured Brad Pitt and Orlando Bloom.
Just before it came out Time Magazine interviewed me on its prospects. The article by Josh Tirangeal was pretty good.
The link is Troy in Time.
There is a funny bit about the 250 Bulgarian weight lifters they recruited for the movie and the problem they had with the food. Storms, wars, and little tortoises on the beach all add to the mad chaos of making this film.
How will it do they asked me? Nobody knows. But, what about all those screens they asked. That is supply, not demand. The audience has to fill those seats or you don't have anything. I haven't checked, but it didn't do very well as I recall; it was gone pretty quickly.
If you think about it, the ability of movies to temporarily fill the shelves is quite different from selling toasters or bicycles. For a week or two, a movie might fill 20% or more of the available shelf space (screens). What if retailing operated this way? It would be madness.
LINK · The Movie Business · Comments (4)
Who is that actor?
June 23, 2005 05:53 PM
One of our readers guessed that the actor shown on the (discounted at Amazon) cover of my book, Hollywood Economics, is Harold Lloyd.
So here is the little contest. The first one to correctly name the actor on the cover, the movie, and the director gets a (listen to this) free copy of Chapter One of Evolutionary Fitness.
I will use the time date on the email to decide who wins and you must name all three. This is easy. But, the prize is not that great either.
LINK · The Movie Business · Comments (9)
Time Magazine and the Movies
June 22, 2005 09:34 PM
I gave a long telephone interview today to Time Magazine on the pending merger of AMC Entertainment and Loews Entertainment. Both are major theater chains and the combined theaters of the merged entity would be 450 theaters. Not so much compared to the more than 5000 theaters in the US, but one that has a lot of major city coverage. I did the interview to test what effect it would have on the sales of my book, Hollywood Economics. Today its sales rank is 223,380. I'll check back after the interview appears. I plan to turn down future interviews; they usually are not worth my time.
As to the merger. Neither company is publically traded so it is hard to know how the market evaluated the news of the pending merger. In most mergers, one or both companies suffer a stock value decline. This is probably due to the high transactions cost of putting the merger together; fees are incredibly high and this has to take a toll on the value of the merged entity. But, the answer I gave is that nobody knows if or how the merger will pay off. It could go either way, though the evidence in the movie/entertainment business is that mergers and acquisitions do not increase value.
The industry has written off many billions in the past decade or so. SONY wrote off 5 or so billion in its acquisition of Columbia, the ABC-Disney merger saw a stockholder value loss of, I forget, but it was many billions. And then there is the Time/Warner AOL merger that helped to produce a loss of stockholder value in excess of 30 billion. Finally, there was the French sewer company, Vivendi. Its claim to fame was political power within France because of connections to municipal policiticians throughout the country through Vivendi's interests in water systems. Through this influence it gained access to capital and political support in media. Mercier, as a graduate of one of the prestigious Ecole's, leveraged this small base into a world wide media company, in part through the acquisition of Universal here in the US.
Read More »LINK · The Movie Business · Comments (2)
Stars and Bombs
April 22, 2005 10:51 AM
So many people seem to believe that stars can make a movie open big that I used to have to remind my class that they can also make a movie bomb. The bombs go off often, but people somehow cling to their belief in stars. So I often used a little gag to make my point.
I would ask them if they heard about the bomb scare yesterday in Santa Monica. Then I would say don't worry it just turned out to be the opening of Madonna's latest movie.
The formula was general however. You could substitute many names in place of Madonna's and, eventually, nearly all the big stars would fit the formula. Arnold Schwarzenegger, Mariah Carey, Ben Affleck, even Tom Cruise eventually fall victim to the bomb. It even works for lizards like Godzilla and other large creatures such as The Hulk.
The general rule is: A bad movie with a big star bombs faster than anything else. Explanations may be found in my Hollywood Economics, but by now you knew that.
LINK · The Movie Business · Comments (4)
Handbook of the Economics of Art and Culture
April 21, 2005 04:28 PM
This is the introduction of my chapter on the movies for the upcoming Handbook of the Economics of Art and Culture. The Handbook series is edited by Ken Arrow and Mike Intrilligator (one of my old professors) and has developed a good reputation for authorative statements on the state of the art in economic research in various fields.
Arthur De Vany, "The Movies," in Victor Ginsburgh and David Throsby, eds. Handbook of the Economics of Art and Culture, North-Holland (forthcoming).
Abstract
This chapter is an overview of a new kind of economics of the movies; it also is my attempt to lay a new foundation of the economics of art and culture. The essence of cultural goods is that the are creative goods that have no natural limit on their consumption or dissemination; they are information goods. And they are wildly uncertain. I show how this vision may be implemented in a rigorous and insightful way in the study of the movies. A centerpiece of the analysis is the stable Paretian hypothesis and its usefulness as a model of motion picture revenues, costs, and returns. The industry’s organization, contracts, pricing, and compensation deals are also seen as rational adaptations to the uncertainty captured by the stable Paretian probability model.
The essence of the stable Paretian model is that the probabilities of motion picture outcomes are far from Normal. The tails of the stable Paretian distribution are “heavy” and large scale events are far more probable in a Paretian than in a Gaussian world. The large events far out on the probability tails dominate sample statistics. The variance is infinite and, for some variables, even the mean does not exist. Movie box office revenues, therefore, have no natural size or scale and there is no typical or average movie; each is unique unto itself. Revenue and cost dynamics are complex and expectations of cost or revenue at level X are proportional to X.
Read More »