|Shares Out. (in M):||42||P/E||0.0x||0.0x|
|Market Cap (in $M):||782||P/FCF||8.2x||7.7x|
|Net Debt (in $M):||-31||EBIT||0||0|
Blucora is well known to the VIC community. The stock is currently trading at around $19, down from $29+ for much of November and December 2013 and we thought it warranted a fresh look. This write-up focuses mainly on recent events; we direct readers to Madler934’s excellent post in late 2012 as well as the lengthy associated message board for further review. We believe Blucora is a terrific buy at these prices; we view it as a collection of five high free cash flow generating businesses with varying risk profiles, very low annual capex (<5% of EBITDA), a $643M NOL balance, and a fair value of $27 representing over 40% upside.
The Company’s current TTM EBITDA multiple is 5.6x and the FCF yield is 13.2%. Using a sum-of-the-parts analysis, we get a price target of $27 for year-end 2014.
Blucora is a collection of five different businesses and the table below breaks out actual and estimated revenue and EBITDA for each business, from largest to smallest by EBITDA. Note that we choose to divide the reported Search business into three sub segments: Search Distribution, Webcrawler, and Dogpile.
Why It’s Cheap
Blucora’s stock price has dropped 25% in the last two months. Two events have taken place to create this downward pressure:
- Feb 18: a short report came out on the stock, written by Gotham City Research LLC. The full report is available on Gotham City’s website.
- Feb 20: the Company re-signed its Google agreement with a change to contract terms, resulting in lowered annual guidance. Blucora also released 1Q14 estimates with a midpoint that was a little lower than consensus.
The Short Report
There were four allegations in the Gotham City report that we investigated further. These were the allegations that we could not immediately explain, despite our familiarity with the business. After conversations with management and various other contacts, it is our belief that each of these allegations is invalid.
1) Blucora buys keywords associated with illicit activities, which drive a substantial amount of traffic and revenue
The report’s case for illicit keyword purchasing is based on the finding that Webcrawler purchased questionable keywords as part of its participation in Google’s AdWord service. These keywords are used to drive traffic to Webcrawler’s landing page where it displays ads that generate revenue for Blucora. A full four pages of the report (pages 18-22) are dedicated to this issue although we note that one table seems to confuse Webcrawler with Dogpile, which does not purchase keywords at all (page 22).
Keyword purchasing, which is part of the search arbitrage process, is exclusively part of Webcrawler’s business. We estimate that Webcrawler represents a total of ~$56M out of $650M of total revenue for Blucora. Webcrawler purchased nearly 1M Google keywords algorithmically in 2013, according to Spyfu. This short report highlights 2-3 terms out of ~1M purchased terms. Management indicated that these keywords escaped the automated filters that were designed to prevent this type of thing from happening. Frankly, it is our opinion that these three terms are innocuous enough that it is easy to see how.
Furthermore, these keywords also slipped through Google’s own keyword filter; it was their AdWords service that held these terms in inventory to sell. The same goes for Blucora’s largest competitor, Ask.com, which also purchased these same keywords from AdWords as the Gotham City report clearly illustrates. A leading Internet fraud expert we spoke with confirmed that both Google and Blucora use automated keyword systems (“robots and software”), and that it’s likely that neither “knew what the heck was going on”.
In conversations with management it was made clear to us that Blucora has no intent to buy illicit keywords and that three terms out of 1M terms are not driving the Webcrawler business, let alone Blucora’s business. As management pointed out on the 4Q conference call, “Totally removing illicit keywords from a search network is an ongoing battle for all industry participants. This is all the more challenging in the case of code words and euphemisms that are ever-changing. While our systems and filters are robust, they require constant evolution.” Googling these same terms now invokes no advertisements from any providers, indicating closure of whatever filter loophole may have existed.
2) Blucora has set up click farms to generate revenue
The Gotham City report case for click farming seems to be based on three things: 1) international traffic for Webcrawler, Metacrawler and Dogpile is significant, 2) international revenue for Blucora is minimal and 3) top search terms are generic and originate from a single IP/block. These three facts are somehow pieced together to suggest that Blucora is engaged in click farming/fraud to generate revenue.
The connection between the aforementioned facts and Gotham City’s click farming allegation isn’t completely clear to us or to any of the experts that we spoke with. Our fraud detection expert’s strong view is that it is nearly impossible to identify click-farming activities externally, likening it to putting a camera in every bank in the world and waiting around for a suspect to commit a robbery, which may or may not happen when and where the camera is focused.
In addition there are several alternative, plausible explanations for the facts above that are not suspicious. For example, international traffic is around 75-85% of most search engine traffic. The US has around 250M Internet users and the world has around 2.5B Internet users so this makes sense. India traffic, which was called out as particularly suspicious, is the second largest search traffic source after the US for every English-language search engine that we checked on Alexa. Blucora’s low percentage of international revenue reflects that 1) TaxAct (and now Monoprice) are US-based and nearly 40% of Blucora’s overall revenue, and 2) search revenue is almost entirely comprised of payments made by Google and Yahoo (US-based companies). In other words, Blucora’s reported international revenue does not tie to its international traffic at all. Lastly, the Webcrawler and Metacrawler generic search terms that the author claims are suspicious, appear to be misunderstood (or mislabeled). These terms are not “top search terms” for Webcrawler or Metacrawler. Instead, they represent the most popular search keywords that send traffic to a Webcrawler or Metacrawler landing page (from an upstream search engine). These are simply terms that are very frequently searched in any search engine. These search terms are unrelated to IP addresses.
It would be a major issue for Google if Blucora had click fraud issues. According to management, Blucora’s compliance team (and Google’s click quality team) actively monitor for serious indications of click fraud. We take management’s comments at face value after speaking with them at length – there is no click fraud taking place and the risk of alienating Google is just not something they would ever consider doing.
3) Blucora’s ad displays violate its Google agreement
The Gotham City case for ad display violations is based on a comparison between the current number/placement of ads that are displayed upon a webcrawler.com search query versus the allowable number/placement of ads that are displayed upon a Google search query per the (now outdated) Google contract filed in the 1Q11 10Q. Four full pages of the Gotham City report are devoted to this issue (pages 27-31).
The simple explanation for this discrepancy, per our discussion with management, is that there are different advertisement real estate restrictions for Webcrawler versus Blucora’s search distribution landing pages. The Google agreement filed within the 10Q refers specifically to BCOR search distribution landing page restrictions while the Gotham City report shows exclusively Webcrawler landing pages. Management also indicated that Google can (and does) change the rules of their contract, periodically, and these changes do not require contract updates. Blucora speaks with Google on a daily basis and sometimes changes and adaptations are made informally to things like allowable number and placement of ads.
As with click fraud, it would be a major issue for Google if Blucora was blatantly ignoring rules in its Google agreement. Blucora’s management explicitly indicated to us that they could not violate Google’s policies and still renew their Google contract. We know that the Google contract was renewed as announced on the last quarterly call and that Google has periodically elected not to renew contracts with partners that violate their agreements (i.e. Babylon). It seems highly unlikely that Blucora has been blatantly violating Google’s contract terms.
4) Blucora engages in “cloaking”
The Gotham City case for cloaking, on page 25 of the report, shows a toolbar search query for the term “cars” that returns an Amazon display ad but that has an Infospace web address (shown in the bar at the bottom of the page). The accusation is that Blucora is tricking the end user into believing they are clicking an Amazon link, when in fact they are being directed to an Infospace landing page.
The example cited above is not an example of cloaking and instead is just a “standard redirect” that all search distribution partners like Blucora use to count click-throughs to advertisements, as part of the normal billing process. Thus users that click on the Amazon link are, for a split second, redirected to the Infospace link, and are then sent along to the Amazon page. Our fraud detection expert substantiated this error, and that indeed this is just a simple standard redirect. In his eyes, confusing cloaking and standard redirects suggested a significant misunderstanding of common practices in Internet search, and called into question the entire credibility of the author.
5) Final Thoughts on Gotham City’s Report
The discussion above highlights a few of the highly questionable accusations made in the Gotham City report. There are many more. The entire report seems to consist of a set of loosely connected, circumstantial, and in some cases completely erroneous accusations and misinterpretations. Our fraud detection expert said it best; that while the Gotham City report was indeed “a lot of pages”, the many flaws in the report call into question whether the author is really qualified to do this type of analysis. We also reached out to the author of the report. Our conversation was short and the author was uninterested in having a discussion with us.
The Quarterly Earnings Call
Blucora’s Q4 earnings call didn’t go so well. The Company beat estimates for Q4, but its guidance for Q1 was slightly more tepid than expected. Its contract with Google was renewed but with a change that caused the Company to lower search revenue guidance for 2014 from growth in the “low double-digits” to growth in the “low single digits”.
1) Guidance for 1Q and Monoprice
The Company guided for 1Q14 revenue and EBITDA of $213-$222M and $51-$54M versus street expectations of $222M and $55M. One part of the miss was attributable to the Monoprice business. Instead of around $4M of EBITDA in the quarter, the segment is more likely to see $2-3M of EBITDA due to more moderate yoy revenue growth (low-mid single digits versus 21% in 4Q13) and elevated sales/marketing expenses. It is not particularly troubling because of the minimal absolute dollars of profit that are at stake but certainly worth keeping an eye on.
2) The Google change
Blucora extended its Google contract for three years, but beginning April 1st, excluded AdSense network ads from Blucora or Blucora customer mobile pages, which includes all pages on smartphones and tablets. More explicitly, Blucora will no longer fill advertising real estate on any of its or its customers’ mobile pages with Google ads. Instead, this real estate will display ads from Yahoo and other ad networks.
How important is mobile?
Based on management’s comments, we estimate that mobile revenue in 2013 totaled $50-60M, nearly doubling from 2012 levels. Similar dollar growth in 2014 implies mobile revenue of $80-$90M for 2014, or roughly 11% of 2014E revenue, although we believe 2014 mobile revenue will be lower than this figure as discussed below.
Will Blucora lose customers as a result of this change?
Google is sun-setting the AdSense ad network for all of its search distribution partners, for mobile, as their contracts come up. This is not a Blucora-specific change and we have first-hand knowledge of one other company that was similarly notified that mobile Google AdSense access would stop on April 1st. Customers will not be able to switch to alternative search distributors to access Google ads on mobile. We believe IAC, which potentially retains mobile AdSense access through March of 2016 (when their contract expires) is strongly discouraged, by Google, from partnering with Blucora customers. This gives us some comfort that Blucora will not see large customer defections to IAC as a result of the modification to the Google agreement.
Of course, customers can always bypass Blucora and go directly to Yahoo for their ads, but as one of their leading customers said; 1) Blucora commands better rates, and 2) if there’s a “political issue”, it’s better to be part of the larger Blucora network of customers than a tiny customer of Yahoo.
Will mobile real estate monetize at lower yields now that Google ads are no longer available to display?
The answer to this question is an absolute yes, which is consistent with management’s reduced guidance for 2014, but by how much, and to what degree margins are impacted, is not clear. The customer we talked to above confirmed this but also said that over time, it could be a “blessing” because it frees up mobile page real estate for things like display ads, “pay-per-call” ads, and other potentially more lucrative sources of revenue. One customer also suggested that eliminating Google’s restrictive policies on the placement and number of ads could ultimately be a benefit.
How does this change affect our model?
We estimate that Yahoo ads on mobile (and other network ads that Blucora might select) will generate roughly half of the revenue that Google AdSense ads would generate and will also decrease overall search margins by a few percentage points. We spoke with one person who was formerly in the mobile search monetization division at Yahoo, who suggested using a 60% of revenue as the absolute floor instead of 50% of revenue but we choose to be conservative. Both this fellow and Blucora management indicated that they would know the falloff in revenue quickly, once the feed is switched; our mobile search expert suggested within one month, and Blucora management indicated that they have a reasonable idea, based on preliminary tests.
The overall impact to our model is a reduction in revenue of $45M and a reduction in EBITDA of $22M in 2014. Using our 8x multiple implies an overall decrease in valuation of about $4 per share in 2014. Our $27 price target reflects these changes.
The Gotham City short report has numerous flaws. The change in Google’s agreement with Blucora certainly warrants keeping an eye on but our conclusion is that the market has overreacted to both of these issues creating an interesting buying opportunity.
Lastly, we note that the WSJ ran an article on Sunday night regarding a potential acquisition of Brookstone by Blucora. We have spoken with management and our interpretation is that Blucora is not going to be purchasing Brookstone. The Company’s M&A strategy remains focused on internet-based asset-light FCF-generating businesses that will utilize the NOL balance.