Streetwise Professor

The most recent derivatives/speculation kerfuffle involves something novel—futures contracts on movie box office receipts.  Two entities, Cantor Fitzgerald and Movie Derivatives, Inc.  have announced plans to introduce such contracts.  The film industry is in a tizzy at the prospect, and has enlisted the help of the usual anti-speculation suspects on Capitol Hill.

The virulence of the reaction is interesting, and deserves explanation.  Here’s my initial stab at the problem.

The most likely effect of the introduction of this contract is that it will facilitate “box office discovery.”  That is, information about the likely success (and quality) of a movie prior to its release will improve.  This will occur because those with private information about films, and those who can better utilize public information to predict a film’s performance, will be able to trade on this information.  The resulting price will aggregate this information, leading to more precise predictions about box office receipts.

Moreover, and crucially, potential filmgoers can utilize the information in the prediction to choose which movies to attend.  If the predictions are good signals of quality, they will have a certain self-fulfilling feature: bad (good) movies will have low (high) contract prices, which will result in low (high) attendance.*

At first blush, this sounds great, but it may not be good news for film producers, distributors, or theaters.  To fix ideas, consider this simple model.  Assume that when movies are released, there is no information that allows potential patrons to determine which movies are good, and which are bad.  Making their decision about which films to view from behind this veil of ignorance, consumers will randomize, and all movies will earn (roughly) the same revenue when first released.  Subsequently, word of mouth and other information will lead to revelation of the true quality of movies, and good (bad) movies will earn high (low) revenues in the weeks after release.

Thus, ignorance effectively reduces the variability of film revenues in the first days after release.

With better information, available immediately upon release due to the existence of a market, however, good (bad) movies will earn more (less) right off the bat.  Thus, one would expect that the variability of receipts will increase due to the new derivatives.  This will make risk averse firms and individuals who are long a piece of the revenues worse off.  (Diversification across films is one way of addressing this risk, but due to the nature of the business, diversification possibilities may be limited.  Nonetheless, it would be reasonable to expect an increase in diversification, perhaps through consolidation of producers, distributors, etc., and perhaps through expansion of equity-like sharing relationships.)

Of course, there is already some information about film quality that is available to consumers before release, so the above characterization is extreme.  Nonetheless, to the extent that the new derivatives improve information, the variability of receipts soon after release should increase.

This revenue risk increasing effect could explain the hostility of the industry to the innovation.  Against this, it could be argued that the contracts would permit hedging.  I am skeptical, however, that these contracts will become sufficiently liquid to make hedging practical.  Virtually the entire industry is long; there is nobody who is short receipts.   This natural hedging imbalance would require substantial speculative participation to offset.  Moreover, given that movies are heterogeneous and relatively numerous, no single benchmark can emerge that can serve as a liquid hedging and speculative market the way that CBT corn or NYMEX WTI can.  Put differently, basis risk is huge here, almost by definition.  The fragmentation of the market is antithetical to liquidity, and without liquidity hedging won’t thrive.  That is, the market is likely to remain a fringe, boutique market that generates valuable information that increases the riskiness of producing and distributing films, but does not work as a hedging mechanism.

There are other potential sources of opposition.  In particular, incumbent producers of information do not like competition from a new source.   The incumbent producers of information include studio marketing departments, advertising firms, and the trade media.

In brief, the angry response is quite understandable.  Film derivatives will likely gore many quite wealthy and well-connected oxen who are quite adept at self-promotion.  Indeed, it’s hard to see a natural constituency that would favor them.

It is also consider the effects of better information resulting from a market on the allocation of resources.  One effect will be to lead a better match between moviegoers and movies (i.e., fewer disappointments).  This is beneficial (and would tend to increase the demand to see movies, because there will be lower risk when choosing which movie to see, which would be something of a benefit to the industry.)  Moreover, it could influence the incentive to increase quality, through, for instance, greater selectiveness in choosing projects to produce, and greater effort to ensure quality during casting, filming, etc.   In my veil of ignorance story, movie quality is essentially in the public domain at release, reducing the value of quality internalized by producers (opening revenue is based on expected quality, not the quality of any particular film).  Finally, there is always the possibility that the ability to trade on information leads to rent seeking, as is true of any market; the social value of private information does not necessarily match well at all its private value.   (I won’t go into manipulation possibilities here.  Most of the horror stories told about this seem overblown.   Very Hollywood, in fact.)

My story has some empirical implications.  It implies that the variability of movie receipts (in the cross section) in the days after release should be higher with derivatives than without them.  It also implies that the variability in receipts in the time series should be lower, because the disparities in information about film quality will be smaller over time with derivatives.  These effects should be more pronounced, the more informative the derivatives prices (net of any effect that their introduction has on the amount of information produced from traditional sources, which should decline due to the new source of information.)

All in all, this will be a fascinating experiment on the informational effect of the introduction of a new market.  If, that is, the powers that be don’t strangle it in the cradle.

* One could argue that this self-fulfilling effect could lead to a perverse outcome.  Let’s say that all trading in the new derivatives is noise trading, and the prices are uninformative, but moviegoers think they are informative.  Then they will go to movies with high derivatives prices, and avoid ones with low prices.  Voila—the uninformative prices look informative ex post.  I doubt that this will in fact occur, however.  Post release, word of mouth and other sources of information will exist, which will permit an ex post evaluation of the accuracy of the derivative price predictions of movie quality.  If consumers learn that the prices are in fact poor measures of quality, they will not rely on them to choose what new releases to attend.