The purpose of this article is to evaluate the amortization of streaming content assets for Netflix (NASDAQ:NFLX) and its impact on earnings. I have explained in a prior article the accounting process involved in streaming content assets - please review that document for the background data. This article evaluates a mundane aspect of accounting, and it proposes that management could provide substantially better information on lifecycles of their digital assets as well as a more modern approach for amortizing the assets.
Streaming content assets and related obligations continue to rise for Netflix, approaching $10.9 billion at the end of 2015 (including the obligations which are off-balance sheet). In addition, the amortization of content is increasing (28%) faster than revenues are growing (23%). Given that the major component of CGS for Netflix is the choice for amortizing content, any implicit error in the approach is material to the evaluation of profits/loss.
The current approach of using accelerated depreciation is a relic of a prior time when computation engines and big data were not available to be used in evaluation and modeling of potential useful lives. Historical methods of depreciation are based on wear and tear on physical assets - things like bearings wear out, engines get dirty, tires wear thin, etc. Digital entertainment assets are vastly different and do not "wear out" over time - rather, a digital asset becomes passé as and when the potential universe of consumers have viewed it and it ceases to be viewed in a meaningful way.
But first a note on the actual process of accounting for streaming content. In the 10-K, the following is noted:
"The Company's estimates related to these factors require considerable management judgment. Changes in estimates could have a significant impact on the Company's future results of operations. In the third quarter of 2015, the Company changed the amortization method of certain content given changes in estimated viewing patterns of this content. The effect of this change in estimate was a $25.5 million decrease in operating income and a $15.8 million decrease in net income for the year ended December 31, 2015. The effect on both basic earnings per share and diluted earnings per share was a decrease of $0.04 for the year ended December 31, 2015."
Source 2015: 10-K www.sec.gov
The key takeaway from the above note is that Netflix maintains a fairly robust record of estimated viewing patterns - as they should. And yet, these details are not consistently applied to the amortization of streaming content assets.
Management typically chooses between straight line and accelerated depreciation for their content. And the above note is most likely due to some shows being placed on accelerated from straight line.
The current accounting choices for annual depreciation are summarized below:Amortization Approaches
The percent is applied to the asset each year and the amount flows into profit and loss via amortization. So, for example, a $100 million series of House of Cards would be amortized at $25 million per year under straight line, or $40, $30, $20 and $10 million per year under the accelerated method.
From the above, it appears that management has chosen a conservative path by depreciating more of the content in year one than would be followed in a straight line approach. Yet, given the current mode of binge viewing, the approach of only amortizing 40% of the costs in the current year for a newly released series is debatable.
Let's assume that the useful life of four years is valid, but who really knows? The accountants in Netflix have a large database of viewing patterns which can be used to identify the lifecycles of their digital assets. And, what if an evaluation of viewing patterns suggested a much different deceleration?
Amortization - modeled on total estimated views
Let's create a model of potential views for a show taking into account two factors:
- A diffusion model as an estimator of viewing habits
- The fact that new customers are being added
What could the adoption curves look like?
Or perhaps a similar:
By Krishnavedala - Own work, CC0, here
For additional information, please consult:
Although the above models are not related to consumption of digital content, the hypothesis is that they might prove to be useful to consider as a surrogate for modeling of viewing of content at Netflix.
The key takeaway is that the first 80% of adoption or consumption happens quite quickly, and that the remaining is fairly static.
With that in mind, let's imagine a scenario of a hit show for Netflix which gets viewed by 80% of their customers in year 1, then has a steady stream of views from the original customers over the next three years. Let's further assume that 80% of the new customers over the next year view the content, and that there are similar small view patterns over the subsequent years. Finally, let's apply some latency to the "fad" and assume that in years 3 and 4, only 50% of the new customers view the content, and that there are similarly small constant views occur thereafter.
The forecast would look something like:
Source: Author's estimate of a diffusion-based model with new customers being added each year
Yes, there are many assumptions embedded in the above forecast, and rest assured that management at Netflix has a much more accurate view of these factors.
One also can rest assured that Netflix's management have chosen to not report or utilize such metrics for their amortization. It is strange that the "new paradigm" company is basking in "old paradigm" models of amortization, presumably because it serves their purpose and because the accountants have signed off on the approach. In the above example, where the content was evaluated based on potential future views, the amortization pattern would have been:Diffusion Amortization
Comparing the three methods, the annual rates of depreciation would differ as follows:
In terms of P&L, this would imply that NFLX lost money in 2015. The impact on reported earnings of a change in amortization expenses from 40% in year 1 to 53% could be as high as $1 billion, although it would be offset by a smaller year 2 depreciation as compared to the current approach. Without additional disclosures, it is not possible to have a definitive answer, but the size is such that it could easily swing the company into an operating loss.
Disclosure: I am/we are short NFLX.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.