• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Dynamic analysis of box office revenue

Student: Klimov Kirill

Supervisor: Andrei Sirchenko , Dean Fantazzini , Dean Fantazzini

Faculty: International College of Economics and Finance

Educational Programme: Master

Final Grade: 9

Year of Graduation: 2014

<p>Almost everyone on the planet has been exposed to the movie industry in one way or another at least once in his or her life. Ever since the creation of the first film in 1906, and the first studio in 1911, the industry saw tremendous growth, quickly turning into one of the most prominent areas of business.</p><p>Creating a new film often requires a huge amount of investment, which is quite risky since it is&nbsp;difficult to predict the box office revenue of the movie before its production begins. Some factors, such as celebrities&rsquo; participation, a well-known director or a scriptwriter, big budget or even favourable market conditions may raise potential revenues. Nevertheless, they are far from being either necessary or sufficient conditions for a movie to be a success. For example, &ldquo;Paranormal activity&rdquo; (2007), with the budget of 15000$ and no celebrities in the cast, earned more than $193mln, which implies the profitability ratio of over 12 000! However, there are also a lot of examples when the movie investment was not so profitable. For instance, &ldquo;Sahara&rdquo; (2005), despite substantial production costs of $241mln and Penelope Cruz participation, earned just a little more than $110mln. &ldquo;The Terminal&rdquo; (2004), ranking 147 in the list of 250 best movies at kinopoisk<a href="#_ftn1" name="_ftnref1" title="">[1]</a>, would be another example. Although it was directed by Steven Spielberg (3 Oscars and 12 Nominations) and included such celebrities as Tom Hanks (2 Oscars and 3 Nominations) and Catherine Zeta-Jones (1 Oscar) among the cast, the film earned a relatively modest amount of $77mln in domestic total gross revenue, barely exceeding its production costs of $60mln. It is interesting to note that, while the movie earned about&nbsp;$13.74mln during the opening weekend, which was quite similar to some other Tom Hanks projects, such as &ldquo;The Green Mile&rdquo;&nbsp;($13.37mln) and &ldquo;Road to Perdition&rdquo;&nbsp;($15.47mln), the total gross of &ldquo;The Terminal&rdquo; turned out to be about twice as low as the average total domestic gross of all the other movies with the same actor.</p><p>These examples, in particular the latter, show that a simple consideration of the standard exogenous static parameters might be not enough for accurate and robust prediction of the box office revenue. Instead, we should consider changes in dynamic parameters throughout the whole screening period. Potentially, this could allow us to promptly identify the reason and the moment when something starts going wrong with the movie produced, and aptly respond to stabilize or even increase its box office revenue.</p><p>Arguably, one of the most important dynamic factors for a movie&rsquo;s financial success is the marketing policy and, respectively, public attention. When discussing related literature, we demonstrate that Google search volume indices (SVI), provided by Google Trends service, could be successfully used to measure the degree of public interest. There is a set of papers which show that performance of the models with Google SVI is significantly better than the performance of the models without this indicator (e.g. Varian and Choi (2009), Kholodin et al (2010) etc.). In a different setting, it has been demonstrated that Google SVI has a substantial predictive power in the influenza epidemic spread forecasts (Ginsberg et al. (2009)), which led to the subsequent creation of a new service, Google Flu Trends, with the aim to notify people and prevent a further spread of the disease.</p><p>With widespread technological growth, however, there has emerged another important factor that could substantially affect the box office revenue - piracy. Illegal downloads have been the driving force behind many changes in the media industry during the last decade or two. Before the piracy became really widespread, musicians, for instance, could earn enough money by issuing new albums and making small concert tours in their support. Today, when every person with internet access can easily listen to any song on the web (sometimes even before the official release), the situation has changed dramatically: now musicians are forced to make big tours all over the world after the album issue to support the revenues.</p><p>The problem of piracy is especially acute in Russia due to historically relatively weak enforcement of the intellectual property laws, with only&nbsp;recent antipiracy law enactment. However, with the widespread internet penetration and ever increasing data exchange speeds, piracy is no longer limited to a particular country, but is a global phenomenon.&nbsp;</p><p>Substantial resources are spent on fighting the piracy.&nbsp;There are special persons whose work involves protecting a movie from being stolen on the pre-release step or during the box office period in the form of a sample copy by the cinema employees. They also monitor social networks and torrent trackers in order to send requests to block the illegal content should it become available after the movie theatre release. Such proactive monitoring may prove it more difficult for the users to find an illegal copy of a movie, essentially increasing the corresponding searching costs and thereby reducing consumption of the illegal content and mitigating the negative effect of piracy on the revenue.</p><p>Going back to the example of &ldquo;The Terminal&rdquo;, piracy might have indeed played a significant role in its financial failure: after all, this was the year when The Pirate Bay, probably the largest torrent tracker, was founded and instantly gained a huge popularity. In fact, the first illegal copy of &ldquo;The Terminal&rdquo; in good quality appeared on the site only 8 days after the actual movie release. This could explain, for instance, why&nbsp;the initial movie revenue was comparable to the similar Tom Hanks movies, but subsequent returns turned out to be substantially lower. Therefore, we consider piracy as one of the important factors influencing financial success of the projects in the movie industry, and incorporate it into our model.</p><p>This paper suggests a reliable dynamic reduced-form model for the movie revenue as a function of various filming and screening characteristics, including the degree of public attention, the distribution chain features, star participation, runtime, various calendar effects, etc. Most importantly, we pay particular attention to the impact of electronic piracy on the financial success of the movie. We discuss how the effect will depend on the various characteristics of the latter, i.e. film grade, the quality of the illegal copy and its availability for an unsophisticated internet user.</p><p>In order to model the effect of various film characteristics on the box office revenue, we have created a new database of the movies released in the USA during the last decade. This involved writing a special script that would extract the relevant information according to particular algorithms from the various publicly available sources and combine it in a single dataset. The database contains a lot of information regarding particular films, including (but not limited to) the presence of stars in the cast, daily revenues dynamics, distribution details, calendar effects for the state holidays/weekends, movie genre, runtime, budget, and other characteristics. Piracy presence is measured by screening the contents of The Pirate Bay tracker and VCDQ illegal content list, paying particular attention to the pirated content release date and its quality. Finally, we also trace the Google Search Volume Index that relates to a particular movie from a month prior to the film release up to the end of the screening period. This allowed us to create a unique dataset with the information covering over 1200 movies produced during the last decade&nbsp; (to be more precise, in average each movie had approximately 70 time points which resulted in more than 80,000 records total, used for the dynamic analysis). The full description of the dataset and its composition/algorithms could be found in the Data section.</p><p>We build a dynamic reduced-form model for the daily stream of box office revenues and make a number of interesting empirical findings.</p><ul><li>The effect of an additional cinema theatre on screening revenues is highly nonlinear, and there appear to be decreasing marginal returns.</li><li>As expected, calendar effects have a substantial impact: during the public holidays the revenue on average increases by 57% and on Friday/Saturday/Sunday by 86/125/79% respectively.</li><li>Public attention, as measured by prior values of the Google Search Volume Index, seems to have the highest effect on the future movie attendance within the span of one to 2 weeks.</li><li>Piracy impact on the film revenue depends on the quality of the movie (as measured by the IMDB rating with a threshold of 6.0) and its electronic copy. In particular, other things being equal, the availability of a low quality illegal copy of the movie decreases daily revenues by 33% for &ldquo;bad&rdquo; movies compared with only 26% for &ldquo;good&rdquo; ones. High quality copy leads to revenue changes of 52% and 33% accordingly. We also prove that the effects of &ldquo;bad&rdquo; and &ldquo;good&rdquo; quality illegal content vary significantly inside and between these groups.</li><li>Searching costs have a substantial impact on the illegal content consumption and its effect on box office revenue. We find that the availability of a movie on a renowned tracker (such as The Pirate Bay) has an additional negative effect on the daily revenues of 12-20% (depending on the quality of the film and its electronic copy) and this effect is statistically different between the groups.</li><li>Overall the results suggest that the piracy is related to a substantial negative impact on the box office revenue, and that the worse is the movie, the stronger it seems to be affected by it.</li></ul><p>The model developed in this paper has impressive explanatory power (as measured by the in-sample R-squared), and could be empirically tested to develop the recommended distribution scheme depending on the movie features. The economic impact of the piracy is assessed and shown to be decreasing with the quality of the movie and associated searching costs.</p><p>The paper is organized as follows:</p><p style="margin-left:57.3pt;">&middot;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Section 1 provides a literature review</p><p style="margin-left:57.3pt;">&middot;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Section 2 discusses the database collected for the research</p><p style="margin-left:57.3pt;">&middot;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Section 3 introduces the main reduced-form model for the numerical dynamic analysis</p><p style="margin-left:57.3pt;">&middot;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Section 4 presents the results of the dynamic analysis</p><p style="margin-left:57.3pt;">&middot;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; In section 5 we discuss the static modeling results and the effects of the static movie characteristics on the box office revenue path-defining individual variables</p><p style="margin-left:57.3pt;">&middot;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Section 6 provides main results and conclusions</p><p style="margin-left:57.3pt;">&middot;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Section 7 contains a small survey of possible ideas for the future analysis and unresolved questions</p><div><br clear="all" /><hr align="left" size="1" width="33%" /><div id="ftn1"><p><a href="#_ftnref1" name="_ftn1" title="">[1]</a> http://kinopoisk.ru</p></div></div><p>&nbsp;</p>

Full text (added June 26, 2014) (1.01 Kb)

Student Theses at HSE must be completed in accordance with the University Rules and regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses