Q1 Revenue Estimate: $697.24m
Q1 Earnings Estimate: $1.04
I first took on Netflix back in July of 2010. I plugged the company’s quarterly earnings into my models, only to find percentage errors in the teens; generally, I only accept average differences between actual quarterly revenues and model estimates of 0%-2%. I lost that battle.
Recently, I decided to engage Netflix once more. The result was the same, if not worse—my models’ predictions of prior quarters were well below actual results, providing an ominous indication that Monday’s actual Q1 2011 results would be far above my Q1 estimates. Just as I was about to throw in the towel for the last time, an idea was born.
I remembered learning about exponential trend analysis from my old statistics professor while I was enrolled in the MBA program. When I began modeling earnings, exponential trend analysis never worked; the estimates were always far above actual results. Given the rapid growth of Netflix’s revenue over the past four years (YoY increases from 2006-2010 of 9%, 19%, 24%, and 34%, respectively), I felt that this would be the perfect company to use exponential trend analysis on. With this, I was off and running.
To begin, I charted the past five years of quarterly earnings (20 total periods). I then added various trend lines and visually examined which ones best matched the revenue curve. The results of the first three trend lines relative to the actual revenue curve can be seen from Exhibit 1. As evident from the chart, exponential, logarithmic, and linear trend lines were added with respective coefficients of determination (R^2) of 97.25%, 71.72%, and 93.39%.
With this information, it was easy to determine that a logarithmic trend line would not work for Netflix. Generally, logarithmic equations are best utilized when there are sudden increases or decreases in the data before eventually leveling out. The problem in this instance was that revenue had not yet begun leveling out; in fact, revenue was still increasing at an increasing pace.
Next to cut was the linear trend line. Despite an R^2 of 93.39%, actual revenue was already significantly outpacing the trend line. Furthermore, linear trend lines are best utilized when the data is increasing or decreasing at a steady rate. Obviously, Netflix is not a “steady” company. Linear tread equations would best be utilized with a company like Walmart.
And then there was one—the exponential trend line. Yet, despite an even higher R^2 of 97.25%, I was not confident enough that this equation could accurately predict Q1 2011 revenue; therefore, I dismissed it like the others.
Exhibit 1
By now, you are probably wondering if I ever found a trend line that I liked. The answer is “yes,” and it’s name is Polynomial. This type of trend line is best used when the data fluctuates, as in the case of Netflix. Polynomial curves can be manipulated based upon their order; the higher the order, the higher the fluctuations in the dataset. Exhibit 2 shows the results of using two polynomial lines of order 3 and 6.
If you are having a difficult time seeing the different lines, then that means I have found the right equations to predict future revenue, specifically Q1 2011. Also, note that the coefficients of determination for both orders are above 99%. This means that the equations predict more than 99% of the variation in revenue!
For those math geeks out there, the polynomial equations for the third and sixth orders are: y = 73.803x3 - 1554.9x2 + 21239x + 207350 and y = 0.1002x6 - 6.1016x5 + 136.26x4 - 1265.2x3 + 3707.3x2 + 17200x + 201566. Yikes!
Exhibit 2
Now that I have found, for the first time, a way to accurately predict Netflix revenues, I wanted to see just how exact these equations were over the past six quarters. The results are detailed in Exhibit 3. I would like to specifically mention that over the past six quarters leading up to the 21st period (Q1 2011), the “Poly 3” equation had an average percentage error of just 0.03% and the “Poly 6” equation had an average percentage error of 0.29%!
Exhibit 3
|
The above models predicted Q1 2011 revenue to come in between $651.15m and $654.86m. For those who know what current analyst consensus is, you are probably wondering why I am so far off. The answer lies in a November 22, 2010 press release—“Netflix Ups Prices, Makes Bet on Internet.” The difference between average analyst estimates and my above revenue estimate is the increase in Q1 2011 revenue attributable to the increase in rate plan prices that took effect in early January.
With a new streaming-only plan and the increase in existing rate plan prices, I estimated an extra $2 in revenue per average customer for the first quarter. This figure, summed together with the average polynomial estimate, comes to $697.24m.
Once I had my revenue estimate, everything else fell in place. I used a mixture of common-size analysis and good ole’ fashion financial statement reading to determine where the other income statement line items are likely to end up by the end of Q1.
The cumulative results of my analysis are summarized in Exhibit 4. I am anticipating an improved operating margin of 14.42% due to a combination of factors such as: the creation of a streaming-only plan, a lessening rate of by-mail DVD rentals, and other improved efficiencies.
Exhibit 4
Income Statement | |
Three months ended March 31, 2011 | |
Revenue | $697,235 |
Cost of Revenue | 439,258 |
Gross Profit | 257,977 |
Operating Expenses: | |
Tech. & Development | 49,504 |
Marketing | 92,035 |
G&A | 17,431 |
Gain on Disposal of DVDs | -1,523 |
Total Operating Expenses | 157,447 |
Operating Income | 100,530 |
Other Income: | |
Interest Expense | -4,000 |
Interest Income | 850 |
Income Before Income Taxes | 97,380 |
Provision for Income Taxes | 39,926 |
Net Income | $57,454 |
Diluted EPS | $1.04 |
Now it is time to see how my estimates compare to big-time Wall Street analysts. Currently, according to Yahoo! Finance, analysts are anticipating Q1 2011 earnings of $1.08 per share on revenue of $703.60m, with a range of $1.02—$1.17 and $646.70m—$719.48m. My figures are four cents shy on EPS and 0.90% below on revenue.
While I am a long-term long holder of Netflix shares, I do anticipate a drop in the after-hours stock price due to results below analysts’ consensus. However, even if there is a drop in the share price, I believe that Netflix’s run is not yet over.
NFLX Ownership
Author: Yes
Author’s Family: NO
Disclaimer:
· Opinions, estimates and projections contained in this report are of the author as of the date published and are subject to change without notice.
· This report is not, nor should it be construed as, an offer to sell or solicitation of an offer to buy any securities.
· Unless otherwise noted, all research reports provide information of a general nature and do not address the circumstances of any particular investor.
· Opinions, estimates and projections contained in this report are of the author as of the date published and are subject to change without notice.
· This report is not, nor should it be construed as, an offer to sell or solicitation of an offer to buy any securities.
· Unless otherwise noted, all research reports provide information of a general nature and do not address the circumstances of any particular investor.
· It is important to do your own due diligence on any position you enter.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.