Are Your Investments Hits or Misses?

On July 2, 2007, Blockbuster replaced its CEO—John Antioco—with James Keyes. Investor Carl Icahn pressed for Antioco's firing in response to the company's escalating debt and to his focus on developing a service with potential but uncertain returns: streaming. This service—like Netflix, up to that point—was never profitable, and Keyes closed it down shortly after arriving. This decision played a part in Blockbuster's eventual closure, which occurred about seven years later.

Since we know Netflix’s undeniable success, it is easy to criticize Icahn’s lack of vision. But it is also easy to appreciate Icahn's sensible logic: in times of crisis, companies must abandon the businesses that produce their greatest losses. Perhaps the exception is when these businesses have the potential for future returns.

The real unknown Icahn faced was the profitability of streaming. At that time, it was already clear that companies like Netflix and Blockbuster could attract millions of users to their platforms, but no profits had been reported for that service. This means that Icahn needed to estimate the profitability of a newly created business model, without having any data to guide him.

In a recent study, I examined the real mistake made by Icahn. Blockbuster's business model was well known: Families bought soft drinks and snacks while renting movies they would return late and then pay the corresponding fee. Therefore, Icahn could estimate the profitability of Blockbuster's basic model by reviewing the returns reported by multiple companies in the sector. Furthermore, Icahn could determine what the best corporate practices were by detecting small variations between different business models. Finally, the type of information to be considered before making a decision like this could be acquired in the market given its public availability or by purchasing it.

However, streaming is an evolution of the Internet-based business models developed in the late 1990s, represented by a large number of companies belonging to the dotcom bubble that burst in 2001, e.g., Google and Amazon, as well as some that went bankrupt shortly after, such as Pet, eToys or Webvan.com. In addition to depending on the Internet, these models are similar to streaming in that there were no (or very few) benchmarks for annual profits. Therefore, it was difficult to infer, from earnings and cash flows, which models could be profitable in the long term.

My result suggests that, faced with a new business model, investors should implement surveys, focus groups or other quick data collection methods that make it possible to estimate the demand for the goods and services produced. 

For new business models, waiting for accurate information to be produced –such as that offered by consulting firms– has little added value, since the benefit of its accuracy is eliminated by the cost of not making a quick decision. For example, waiting for a company like Google, Amazon or Netflix (before 2010) to report its first earnings could discourage investment in a business model for which profitability has not yet been proven—at least in the short term. However, the cost of not making a quick decision means that, by the time you look at the earnings, the stock price already reflects the value of the company and the possibility of generating alpha is lost.

The academic paper’s results are much more general the Blockbuster example or any given business application. My main result is that the type of information that a decision-maker needs prior to making a well-informed choice crucially depends on whether pertinent information can be acquired from an external source. 

Any external data that ought to be acquired needs to be highly accurate, even if it arrives infrequently (Zhong 2022)[1]. An example is when a regulating body requires a new drug to pass external safety and effectiveness tests. This principle also applies to investments in an established company: the investor should invest when the company’s earnings, with respect to its market value, remain at a relatively low level for a significant length of time.

When such data cannot be obtained from a public source, the company or individual must produce it themselves. This process is known as experimentation. Having to internalize the costs associated with experimentation qualitatively changes the incentives to acquire information. In this case, the decision-maker benefits from acquiring data that can be obtained relatively quickly, even if it has low precision. For example, a company that is searching for the active ingredient for a medication benefits from running several rapid, inexpensive experiments. Speed ​​makes it easier to abandon compounds with little revenue potential and look elsewhere.

However, the difference between the entrepreneur and the regulator is that the latter does not internalize the cost and time required to experiment. Venture capitalists investing in startups face the same incentives. This observation implies that they must quickly search on their own for a great deal of information and evaluate the profitability of numerous companies, since most of them will be unprofitable.

On July 2, 2007, Blockbuster’s board made the final and erroneous decision to elect a new CEO, James Keyes, who quickly abandoned streaming. This date will forever be infamous in the business world, not because of Icahn’s lack of vision, but rather for his inability to discern the right type of information to compare new versus established business models.  

[1] Zhong, Weijie. 2022. Optimal Dynamic Information Acquisition. Econometrica 90(4): 1537-1582.
 

The author is assistant professor at the Department of Finance and Business Economics, EGADE Business School.