December 05, 2013 - 12:54am EST by
2013 2014
Price: 47.39 EPS -$0.54 -$0.41
Shares Out. (in M): 40 P/E N/A N/A
Market Cap (in $M): 1,875 P/FCF N/A N/A
Net Debt (in $M): -98 EBIT -20 1
TEV ($): 1,777 TEV/EBIT -88.0x 1,269.1x
Borrow Cost: NA

Sign up for free guest access to view investment idea with a 45 days delay.

  • Recent IPO
  • Online Advertising
  • Margin compression


Your application was:
I think Rocket Fuel (“FUEL”) a recently IPO’d AdTech company is currently a really compelling short.  

FUEL completed its IPO on September 16th of this year at $29 per share (the initial range was $24 - $27) on the first day of trading they shot up to $56 per share.  Post IPO they have been pretty volatile shooting down 20% on certain days or rallying 50% in certain weeks.  They closed recently at $47.39.  I think they are worth around $30 per share – for a potential return of approx. 37%.  FUEL has a market cap of about $1.9B and an enterprise value of $1.8B.  They have around 40m shares outstanding on a fully diluted basis.  The quick summary: they operate in a really crowded market that is becoming increasingly commoditized and eventually will be completely disintermediated, there is no customer switching costs, they are benefiting from the rising tide of the real-time-buying environment,  they operate in a grey area and engage in practices boarding on fraud, they trade at a wild premium to their peers based on unrealistic growth assumptions, and finally there will be huge selling pressure from insiders as the lock-up approaches (lockup expires on March 15th 2014).

With the executive summary out of the way let’s now learn how the sausage is actually made. 

The company operates as a demand side platform (“DSP”) what this means in practice is that they are a trade execution platform that sits in between the buyside (marketers and agencies) and the exchanges.  On the opposite end of them is the sell side (where there are sell side platforms that help consumers/publishers sell ads on exchanges).  DSPs help buyers purchase ads and do so using software (“algorithims”) to produce a portfolio of ads geared towards individual customers advertising goals (i.e. so many views or a specific ROI based on brand lift or retargeting). They do this with three primary products: brand booster (helps initiate contact and foster brand engagement), direct response booster (helps with prospecting and retargeting), and other offerings such as analytics and brand assurance which ensures that they are not buying up ad inventory on inappropriate sites (porn, etc.).  Right now about 2/3 of their revenue comes from direct response with the remaining 1/3 coming from brand campaigns (brand campaigns are harder to do since the ROI isn’t as clear as with direct response).  The big growth promise for the company is mobile and video which are currently 26% of revenue.  The economics on these are vastly different from display which the market hasn’t yet fully grasped (mobile is far less profitable and video has a smaller spread due to the higher ASP for high quality videos – see Tremor media for the recent challenges in the video space).  Overall, the space is really crowded – to get an idea of how crowded it is I encourage you to look at the AdTech lumascape from Luma Partners (a tech focused investment bank):  As you can see the space is really crowded.  There are 90+ DSPs out there.  FUEL is neither the largest nor the fastest growing.  There are only two “pure-play” publicly traded DSPs – FUEL and Criteo (Criteo is European focused and about 3.3x the size of FUEL but growing at lower rates).  The public investors don’t really understand the company or space.  They are just blinded by growth rates, which to the company’s credit have been awesome (2012 revenue of $106.6m vs. 2011 revenue of $44.7m).  We will talk more about the financials later on.  

The market doesn’t realize how crowded and undifferentiated the space is.  There are probably four or five private DSPs that have in excess of $100m and they are all growing at similar rates.  Additionally, Google, Yahoo!, and AOL all have in house solutions.  Large customers like P&G and Kellogg even use their own in-house solution since they found the DSP platform so easy to replicate.  The big advertising agencies used to have in-house DSP solutions but have all exited the space as they realized how crappy the economics are and how likely they are to be put out of business.  The ad agencies all use internal trading desks to help purchase and sell inventory for their clients.  Xaxis (WPP’s trading desk) recently (summer of 2013) shut down their DSP solution (WPP has historically been a big proponent of in-house proprietary technology) calling the space “commoditized”.  Xaxis Global COO Mark Grether said, “As technology evolved, we came to the conclusion that the DSP is becoming a commodity, but there are other technologies that make a difference. We've shifted our focus into other areas.”  WPP (along with the other agency trade desks) realized that they need to pivot away from technology enabled point-solutions into data management and technology integration (basically they want to make the same pivot IBM did in the 90s towards business services away from point solution).  A result of this move is that everyone uses multiple DSPs (upwards of 10).  There is in fact zero customer loyalty and no switching costs.  The average contract length is not measured in months or days but literally trades.  People will engage a DSP to acquire a set of ads and when that is done they walk away.     

DSPs make their money solely by buying up a portfolio of ads on exchanges as cheaply as possible and then reselling them to their customers at a significant markup.  The model is as simple as that and with the advent of real-time-bidding (“RTB”) which is also known as programmatic buying is simply the digitization of buying and selling ads.  The best analogy to think of is the transition away from floor brokers at the NYSE.  So it’s obvious why there has been fast growth in the industry as of late (the growth will rapidly dissipate as this conversion runs its course).  The industry is full of opacity and “buzz words” that DSPs use to make their products sound more complex than they really are.  My favorite example of this can be seen in scanning websites of the top players in the space (all the private DSPs below are $100m in revenue or above):

•       Rocket Fuel describes their business as using artificial intelligence to do big data driven real-time programmatic buying (Yikes what a mouthful).  
•       Criteo (a public DSP) describes their business as predictive software algorithms coupled with big data to deliver advertising results in real-time.
•       Collective (a private DSP) describes their business as employing data driven programmatic advertising through data science.  
•       Data Xu (a private DSP) describes their business as programmatic optimized real-time buying leveraging big data.     
•       Media Math (a private DSP) describes their business as using algorithmic optimization in a real-time buying environment.
•       Turn (a private DSP) describes their business as data driven advertising using programmatic solutions with big data infrastructure.

As you can see from above there is no differentiation across the industry as everyone claims the exact same advantages.  I called a bunch of industry folks during the diligence process and the clear theme that came across was that it was universally recognized that DSPs are playing an arbitrage game and that the arb is quickly eroding (due to automated buying and wider-spread use of purchasing algorithms throughout the industry).  The technology is being commoditized and pure-play DSPs are not long-term viable business models.  The trend in the industry is towards platform solutions (i.e. not working in the vertical silo of a DSP but offering a horizontal solution that plays with both the buy and sell side).  Rocket Fuel has made zero effort to diversify their business and has made it clear time and time again that they are a DSP through and through.  When asked about it during their roadshow the response was we are diversifying by expanding internationally.  It seems every other player in the industry gets that this is a quickly ending game and the time to diversify is now (if not yesterday).  To make matters even worse the industry is rife with lots of shady tactics that are borderline fraud.  Google believes that 30% and 40% of all ads are never seen which is a function of fraud, others have estimates that are even higher than this.  There was a good WSJ article describing the fraud practices in the industry and what is being done to counteract the historical bad practices:

To make matters even worse there were a couple reports that came out this summer from (an industry researcher) claiming that DSPs are benefiting from botnets serving fraudulent views in that they can charge higher prices for much lower quality inventory.  This was actually corroborated on some calls with customers of FUEL where they basically alleged that FUEL was overcharging them for inventory.  While fraud is not a huge key to this story what is important to recognize is that participants are tired of the self-imposed opacity in the DSP space and they are pushing for clarity which will come at the DSPs expense – especially as the arbitrage game plays itself out.  

There is also no risk of FUEL being taken out by someone.  Google has its own DSP through Invite Media.  Yahoo has its own through Right Media, AOL actually just shut theirs down, large agencies are outsourcing this since they view it as a bad business to be in.  To make matters worse DSPs are being disintermediated by other solutions pivoting to become platform solutions (SSPs, ad networks, publisher tools, etc.)   So there is really no one left who would want to buy in.

FUEL’s key backer Mohr Davidow (the main VC investor) owns 33.4% of the outstanding shares and they are sitting on nearly a 50x so they will sell at almost any price.  This will be the first win for that vintage fund which is important since it was a 2007 vintage so it should be in harvest mode now, plus Mohr Davidow is in the market to raise a new fund so any win they can show will help with the age-old LP reup.  To further investigate this phenomenon look at Mohr Davidow’s recent other IPO wins (Rally Software, Shutterfly, and ProofPoint) they were pretty big sellers within a year or less of the IPO.  Given their position here I would expect a big distribution to LPs, probably not the whole position but a good 10% or so of the outstanding shares which will cause real selling pressure.  Mohr Davidow’s stake today is valued at nearly $500m which is in excess of 50 days of trading volume.  So their distribution (let alone the other VCs which own around 18% of the company) will cause real downside pressure.  If you look at VC backed IPOs over the last two years they do tend to drop by about 15% from the IPO price (using longer dated data from 2000-present the drop is around 10%) – so this effect is real, coupled with large insider selling it can be severe.  

To make matters even worse the AdTech space is mired with high-fliers out of the gates who then immediately crash and burn.  Millennial Media came out at nearly 5x EV/R with lots of pomp and circumstance and they now trade at 1.4x forward EV/R.  It should be noted this is not because their growth has slowed (they are projected to grow 51% in 2013, 53% in 2014, 40% and 2015, etc.) nor is it because they grew into the multiple since IPO their 2015 consensus revenue estimate has dropped by about 15%.  If you are wondering if this is unique for the AdTech space it’s not.  This can be seen with Marin Software (IPO’d at 6x EV/R and now trades at just south of 3x), Tremor Media (IPO’d at 3x and now is at 0.5x) and most recently YuMe (2x EV/R and now 1.5x EV/R).  The industry as a whole trades at 2.0x forward EV/R vs. nearly 4.6x for FUEL – this is why the industry is growing by about 30% per year for the foreseeable future.  So the market pretty quickly reverts to the mean once it realizes that their lofty growth assumptions were too high and the businesses are far worse than they imagined.  During FUEL’s IPO roadshow they guided to 20% long-term EBITDA margins whereas Criteo their nearest and dearest competitor guided to 15% long-term EBITDA.  When speaking with industry contracts the consensus is that FUEL has unsustainably high gross margins due to the arb game and this will likely feel significant compression going forward making it nearly impossible to hit 15% let alone 20%.  

Finally if you look at valuation the assumptions required to validate FUEL’s current price are just breathtaking. FUEL will probably do $232m in revenue this year with an EBITDA loss of $5m or so (they are not nor have they ever been profitable).  This is nice 100%+ growth (mainly due to RTB transition which is temporary).  Additionally, the company has debt of about $30m with another $30m or so in operating leases.  The company burned about $8m in the most recent quarter in operating cash flow (they burnt $20m in 2012) and they aren’t projected to be cash flow positive until 2016.  While they do have $125m in cash if you include debt and operational cash burn there is a chance they will need to raise more either via equity or debt.  In my model I assume a 35% revenue CAGR from 2012-2022.  There aren’t too many businesses that can compound at 35% over 10 years when they are already starting from a $100m+ base.  To be perfectly clear in the AdTech space there is none (Google has done it in search but let’s face it that is not comparable at all here).  Beyond that you need to assume that they have constant ARPU increases which is quite a steep order considering that ARPU over the past few quarters have gone from 9% to 3% to 0.5% - this is due to customers spreading spend out across multiple providers.  You also need to believe that the gross margin is able to stabilize at 45% which is in excess of the industry average (Criteo for frame of reference is at 37%).  Then I add on 8x terminal multiple against 2022 FCF, discounting back the cash flows at an 11% discount rate results in a price target of around $30 per share.  It’s really entertaining to go through the sale-side models and look at what they do to justify their valuation.  After FUEL came out with decent numbers in their most recent quarter (they dropped due to negative industry commentary from Tremor Media on the video space) analysts actually upgraded on the selloff and lowered price targets.  This just shows how at the IPO they couldn’t come out with sell ratings but couldn’t really justify the current price so they resorted to financial witchcraft.  My personal favorite is BMO Capital which uses a 30x terminal multiple to justify the current price.  So as best as I can understand it if you believe the company can compound revenue at 30%+ over the next 10 years, maintain gross margins despite industry pressures, and expand EBITDA margins to 21% which is beyond FUEL’s target and what industry insiders and competitors say is feasible, and slap on an excessively high terminal multiple then you can justify the current price.  But, if not all those works the stock will sell off sharply.

One item worth noting - the borrow has tightened considerably since I first submitted this to VIC for review.  What started out at 3%-7% borrow cost can now vary greatly based on your broker and the availability of borrow.  I also already pinged the CBOE about listing options (as I think they represent far more attractive risk/return dynamics) and they said they were not listing them at present due to volume issues (I suspect this is a float comment more than a trading volume comment).  
I do not hold a position of employment, directorship, or consultancy with the issuer.
Neither I nor others I advise hold a material investment in the issuer's securities.


So to summarize.  This is an unprofitable, over-hyped business, in a challenged market, that is becoming increasingly commoditized overtime.  To make matters worse they are trading at an unreasonable valuation due to Wall Street’s unachievable expectations.  They will revert to their peer group and trade down substantially due to one of the following catalysts:
1.      Missed or reduced growth estimates (due to customers switching away or reducing allocation as they utilize multiple platforms)
2.      Poor industry commentary from a competitor (sold off 20% on video comments from Tremor even though video only represents 1% of current revenue)
3.      Reduced margins (the margins will compress as the ad buying arbitrage collapses
4.      Realization from the momentum investing community of how crowded the space is and a re-pricing of the multiple (i.e. reversion to the mean similar to other AdTech IPOs)
5.      Selling pressure from major shareholders specifically around the lock-up expiry

I don’t think there are a lot of risks to the counter at the moment.  The only real negative is borrow is tight but depending on when you call it’s not that bad.  The ranges I have been quoted are 3% -5% when borrow is available.  

Seems to me the deck is stacked in your favor at the moment.  The market is forgetting when playing with Rocket Fuel you can get burned.
    show   sort by    
      Back to top