Netflix Change in Amortization Estimate

In its Q3 2015 10-Q, Netflix, Inc. (NASDAQ: NFLX) disclosed a change in accounting estimate made to their amortization of certain intangible assets.

1. Basis of Presentation and Summary of Significant Accounting Policies

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 $12.8 million decrease in operating income and an $8.0 million decrease in net income for the three and nine month periods ended September 30, 2015. The effect on both basic earnings per share and diluted earnings per share was a decrease of $0.02, for the three and nine months ended September 30, 2015.

As noted, the change in estimate accelerated the recognition of amortization expense, resulting in an $8 million decrease to net income for the three and nine month periods ended September 30, 2015. The effect on basic and diluted earnings per share was a decrease of $0.02, which contributed to the quarter’s earnings miss.

This change in estimate came up in the company’s earnings conference call (excerpt courtesy of Seeking Alpha).

Peter Kafka – Re/code

You guys changed your amortization schedule for your license content… what does that tell us about viewing behavior?

David Wells – CFO
Presumably, Peter you’re referring to … the change in accounting estimate. … What we’re saying is there is more viewing in the first month not necessarily less viewing over the lifetime…. And it could be that [with] so much good content over the last few years — you have to make a … choice…. You’ve got 20 good choices, you’re choosing.

So it could be that the merchandising system is concentrating a little bit of that when something comes in.
Regardless, we try to be accurate in our accounting, and we follow the trends regularly on a quarterly basis. And this past quarter, we noted some content that was trending up, so we accelerated the amortization on that.

Reed Hastings – Founder and CEO
Peter, let me give you a little example…. Suppose we pay $1 million dollars for a title for a four year license, and we think it’s going to go 1 million hours of viewing. But the title is just incredibly popular, and instead it does 10 million hours. But many of those hours are in the first month as supposed to spread over the four years. Well, then we would accelerate the amortization of that million dollars, even though the whole thing is doing much better than we ever thought. Its relative proportion overtime is more upfront. …

Peter Kafka – Re/code
And generally what’s happening, are these things over performing or is there a mix where they sometimes are performing less well than you thought?

Reed Hastings – Founder and CEO
It’s not about the total performance that gets us to change the amortization, it’s the relative performance overtime. And so if it performs a lot in the beginning, relative to how it does at the end. That’s what gets us to accelerate. …

Theodore Sarandos – CCO
Well, and a good recent real world example is Longmire. We just released the original fourth season of Longmire. So season one, two, and three, which we’ve licensed for several years on Netflix, they were at the kind of tail end of their life, actually exploded when Longmire season four launched.

David Wells – CFO
And it’s a good lead in to remind everyone that these are dynamic, right. That’s why we look at it every quarter, because things change. …

Reed Hastings – Founder and CEO
And where it’s ambiguous, we would tend to want to amortize it more quickly. That’s advantages to us.

Theodore Sarandos – CCO

Reed Hastings – Founder and CEO
But it has to be justified by the actual viewing patterns.

Here you have the CEO, CFO, and CCO all discussing a particular accounting policy extemporaneously. We don’t see this too often, and it’s an interesting discussion on its own. At first it seems pretty straightforward, if a little vague, but on reading the full excerpt there is an apparent lack of clarity. A quote often attributed to Einstein goes, “you don’t really understand something unless you can explain it to your grandmother.” Peter Kafka from Re/code might not be anyone’s grandma, but I wonder if he was able to deduce what was really going on from the answers given.

Ostensibly, the change is straightforward. Here’s the company’s accounting policy for the amortization of Streaming Content from the same 9/30/15 10-Q.

Critical Accounting Policies and Estimates
Streaming Content
Based on factors including historical and estimated viewing patterns, we amortize the content library (licensed and produced) in “Cost of revenues” on the Consolidated Statements of Operations, on a straight line or on an accelerated basis over the shorter of each title’s contractual window of availability or estimated period of use, beginning with the month of first availability. The amortization period typically ranges from six months to five years. For most of our content, we amortize on a straight-line basis. For certain content where we expect more upfront viewing, due to the additional merchandising and marketing efforts, we amortize on an accelerated basis. We review factors impacting the amortization of the content library on a regular basis, including changes in merchandising and marketing efforts. Our estimates related to these factors require considerable management judgment.  Changes in our estimates could have a significant impact on our future results of operations. In the third quarter of 2015, we changed the amortization method of certain content given changes in estimated viewing patterns of this content. The effect of this change was a $12.8 million increase in cost of revenues in the Domestic streaming segment, for the three and nine months ended September 30, 2015.

Netflix amortizes Streaming Content using two methods: either straightline or accelerated. The straightline method amortizes content costs equally by year over the title’s life. The accelerated method amortizes costs proportionally to the amount of viewership, typically weighted towards the beginning of the title’s life. So presumably, the amortization method for some content was changed to recognize more of the expense in the first year, to reflect that proportionally more views were occurring in the first year.

But, interestingly, the examples given by David Wells, CFO, and Theodore Sarandos, Chief Content Officer, don’t appear to be 100% consistent with that explanation. Wells’s first explanation is that there’s an abundance of good content, which is causing viewership to shift towards the beginning of content’s life. “And it could be that [with] so much good content over the last few years — you have to make a … choice…. You’ve got 20 good choices, you’re choosing.” One would think an abundance of great options would actually have the exact opposite effect. Imagine Netflix only had one great show, say “House of Cards”. If that were the only option, then people who primarily use Netflix for their video entertainment would be more likely to binge watch House of Cards all at once, because there’s nothing else on. On the other hand, if there are a lot of great options, a subscriber might watch “Orange is the New Black” first, and then catch up on House of Cards sometime in the future. This seems to suggest that viewership numbers would actually even out over time as more great content is added.

Further, the example given by Theodore Sarandos is precisely the opposite of a situation where accelerated amortization would be called for. He points out that, with the release of Longmire Season 4, the prior seasons of the show saw a noticeable uptick in viewership. That is, there was a viewership increase in the later years of the amortization schedule. This example argues for decelerating amortization, or at least keeping it straightline.

There was potential here for management to give some real insight into the viewing habits of its subscribers, but unfortunately not much additional information can be gleaned from this conversation. Given Netflix’s general antipathy towards sharing viewing statistics, it’s not surprising that they weren’t more candid here. Or, perhaps, they don’t really understand what’s driving changes in viewership.

The Audit Analytics Changes in Accounting Estimates Database covers all SEC registrants (foreign and domestic) that have disclosed a change in estimate as defined in ASC 250-10-20 and as required by ASC 250-10-50-4 in an annual or quarterly filing. The database currently has over 5,400 records with coverage from 2007 to Present. For additional information or to schedule an online demonstration of Audit Analytics, please e-mail or call 508-476-7007.