Researchers have linked watching Netflix and stock declines. What questions do you have for the study?

The release of new TV series on Netflix and other streaming platforms provokes a decline in stocks the next day, according to a study by a group of scientists from the University of Hong Kong Business School. They concluded that traders who are exhausted after a night of "inebriated" watching fresh episodes are more inclined to close positions than to open them. Psychologist Mikhail Tegin breaks down the methodology of the study for Oninvest and explains what questions it raises.

Read more about the study: Netflix series premieres lead to stock declines, researchers said. How is this related?

***

This study uses events from popular culture to model a behavioral shock that is controlled for as much as possible in finance. The authors test how this shock affects decision making in the market.

The study uses an indirect indicator of the peculiarity of market behavior after a shock, namely sleep deprivation. The authors attribute this sleep deprivation to night premieres on streaming platforms. The authors manually collected the release dates and times of new seasons and episodes of popular TV shows over a 10-year period, which is quite a long period. They also defined what constitutes a late-night show - something that airs between midnight and 5 am.

Initially, more than 6,000 facts of releases of different shows were collected for the study. But since this is a very large volume and not all premieres force the audience to sacrifice sleep, the authors decided to limit themselves to the 108 most popular shows from Netflix, Hulu and Amazon. This limitation is probably necessary to ensure that the premiere of a show is in fact a strong exogenous shock capable of affecting a significant cohort of investors, and is not just a random event.

As for monitoring the market situation, the authors used an extensive set of data: daily returns of various portfolios on S&P 500 and so on. There are practically no questions to the research methodology here.

For the model itself, the authors use a regression analysis approach with a so-called dummy variable. They assume some event has already happened, rather than looking retrospectively as in classical event study. Due to regression analysis at the portfolio level as well as panel regressions with fixed effects at the level of individual stocks, this method seems to be very reasonable.

However, the study has a couple of limitations. First, and this was mentioned above, is the number of potentially relevant shows - 108 out of more than 2,000 shows released on platforms over 10 years were selected. Because of this, the influence of investors' specific preferences and tastes in shows and movies may be excluded. At the same time, this is an understandable and somewhat necessary compromise to ensure the strength of the effect, but not to limit the external validity - the validity and reliability of this indicator. That is, the effect is only studied from very high-profile events.

Another limitation that I would say is key: it's voluntary viewing. This is not directly stated in the study, but it can be seen. The authors cannot measure exactly how many investors watched the show and whether they (these investors) are part of the market being analyzed. The researchers only have two facts that happened one after the other - the premiere of the show and the stock market decline. Strong evidence is shown to suggest that these factors have a causal relationship. But we can't clearly state here how significant a cohort is affected by the new release on streaming - is it 51% of investors watching it, 60% of investors? Is it just institutional investors or is it all investors? In other words, the massiveness of the phenomenon can be questioned in this part.

On the other hand, researchers justifiably use the behavioral approach and, among others, research in cognitive science, which indeed shows and proves that diminishing returns occur through the behavioral channel, namely: lack of sleep in the case of investors may make them less inclined to make a decision to buy securities, since buying requires more cognitive effort than selling. This is because buying requires more sophisticated analytical thinking: you have to analyze the macroeconomy, the sector, fundamentals, etc. And selling is a faster, more intuitive and heuristic decision based on simplified rules of thumb. Roughly speaking, selling is easier than buying. So, if we assume we have a causal relationship between the two events, and investors really don't get enough sleep watching shows at night, they do have limited cognitive resources the next day and are more likely to make heuristic sales than "more labor-intensive" purchases. This can actually cause the price of securities to fall.

But we stipulate that it's unknown how many investors a) watch the show and b) don't get enough sleep because of it.

Moreover, the figures suggest that it is more about price drops that are insignificant for the market volume, although the phenomenon itself is obviously taking place.

This article was AI-translated and verified by a human editor