Description

Short The Trade Desk (TTD).  TTD is an expensive provider of ad tech that has benefitted from addressing certain trends in digital advertising well the last few years, but is inherently a commodity that will see its market position erode.

TTD provides a self-serve technology platform for ad buyers, known as a demand side platform (DSP), focused on ad agencies.  The bull case on TTD is that its focus on the agencies aligns their interests with the agencies and that TTD is more transparent than other DSPs.  DSPs in general are benefitting from the overall growth of digital and an ongoing transition to programmatic advertising.  TTD’s data and tools are a competitive advantage.   TTD is taking an “omni-channel” approach to digital advertising, which will open up other growth avenues beyond display.   The competitive advantages highlighted in this thesis are drastically overstated at best.

TTD operates in an extremely competitive environment.  TTD is focused on an agency self serve model that is supposedly more transparent than peers.  This approach has worked well for TTD thus far and agencies have appreciated the degree of service and support that TTD provides, but focus and service do not translate into a competitive advantage.   Generally, the consensus of my discussions with industry participants has been: we like the service, but the DSP itself is undifferentiated and/or a commodity, as in basically every DSP has access to every exchange and all of the pertinent information that comes with them.  Many of TTD’s competitors also take a multi-channel approach, as well, and there is nothing proprietary about it.  Likewise, most large agencies use many platforms’ APIs, so again this does not impart any special advantage.

TTD has gained market share, particularly with larger agencies, but there is reason to believe that that will no longer be a tailwind.  In 2015, TTD grew its same client spend by 135%.  This coincided with some larger agencies consolidating providers, with some reducing the number from a couple dozen down to several of the DSPs with the most scale and deepest resources.  So this does show that clients do like the product and service, but also shows they are pretty much interchangeable as decisions were made based mainly on price and secondarily on level of service.  This transition seems to have nearly run its course, so this consolidation will no longer contribute to growing same client spend going forward.  Scale helps, but there are no real barriers to entry or intellectual property, so there will be a constant threat from existing competitors and new entrants.

Buyers have considerable power over TTD.  OMC, WPP, and IPG are approximately 1/3 of gross billings and are critical clients for any company in ad tech.  As a result of this competition and lack of differentiation, most industry participants expect and will demand take rate compression.   TTD currently enjoys a take rate of 19.4% as of 2Q17, which has declined from 19.7% in 2016 and 20.6% in 2015.  From my discussions with industry participants the trend is for continued pressure on take rates.  One of the large agencies is currently paying a take rate of well under 10%, which they describe as only a technology fee to use the platform, although they still enjoy a significant amount of hand holding from TTD in terms of service. 

Customers also are not quite convinced on the alignment of interests and transparency of TTD.  Second price bidding is employed by most DSPs where the winning bid pays the second highest bid as the clearing price.  The winner of an auction still does not have complete transparency into what all of the bids were and whether the price it paid was fair.  Advertising buyers are concerned about price manipulation and currently the DSPs are basically a black box and the malfeasance can be a significant contributor to margins.  There are currently efforts underway to try to eradicate fraud and increase transparency.  This is an industry issue and general and nothing specific to TTD, but my sources indicate that most agencies are skeptical of the process across the board and looking forward to the coming changes.

Suppliers have considerable power of TTD and many of its largest suppliers are also competitors.   Much of the best inventory comes from “walled gardens,” which the Association of National Advertisers defines as platforms where the carrier or service provider has control over applications, content, and media and which restricts access to its data; think GOOG, FB, TWTR, Linkedin, Pinterest, basically anyone well positioned on the internet.   Most of these operate their own DSPs and make advertising inventory available to third parties only as it best suits them.  The GOOG/FB duopoly is extremely powerful, is growing market share, controls what inventory they release to third parties, and can squeeze the margins of the DSPs.  If TTD’s access to this inventory was limited or the terms of the relationship changed, TTD could be adversely impacted.  

Working capital is a potential serious problem.   TTD is in the disadvantageous position of having to outlay the entire gross spend.  Its clients generally pay slower than it has to pay its suppliers, so accounts receivable are enormous compared to net revenues, i.e. for every $1 that an advertiser spends with TTD, TTD recognizes $0.20 because its take rate is ~20%, but the entire amount is a receivable.

Management calculates their own DSO based on gross billings, which is fair, but it is also worth being aware that TTD is on the hook for collecting the entire amount, so credit risk can become a real issue.   As of June 30, 2017, by management’s metric, DSO increased from 88 to 95 yoy.  TTD will always have elevated DSOs, but this increase is still a red flag.  Agencies typically wait until they get paid after ~60 days and then wait a month to pay TTD.  The DSO on net revenues is a little more alarming at 479 days in the most recent quarter.  It isn’t often that you hear someone say DSOs improved to 479, but that’s what happened here, as DSOs were well over 600 by this metric in 2015 and 2016.  Clearly, this was a very high growth period for TTD, so it is understandable that receivables grew, but at this very high level there is the possibility that TTD eased payment terms to customers in order to drive growth to dress itself up to go public.  We will probably know one way or another in the coming quarters.  Regardless, the timing of payments does create some risks, given that FCF was ($20.5) million in 1H17, although cash conversion usually improves in the second half.  Due to the imbalance of DSO and DPO, TTD relies on a credit facility to fund working capital and grow the business.  Also, allowance for doubtful accounts was $5.8 million in 2Q17 or 1.4% of revenues, up from $2.6 million or 0.7%, and it would not surprise me if this went materially higher.  TTD went public with material weaknesses, which given the level of DSOs would be very concerning to me, if I were long.  Growing companies need to consume working capital, but these collection and payment terms leave TTD in a precarious position.  I attached some of the risk disclosures on this topic in an addendum, which are worth reading in their entirety.

Bulls overestimate the TAM and underestimate current penetration.   In 2016, eMarketer estimated that Total Media Ad spending would grow to $551 million globally with $195 million coming from digital, or about 35%, and digital is forecasted to grow at a CAGR of 15%.   However, in the US eMarketer estimates that in 2017 80% of digital display dollars will transact programmatically at just under $33 million, so programmatic penetration is already high.  Beyond that, GOOG and FB are getting virtually all of the growth, so while digital’s share gains will continue its inexorable rise, it will not necessarily be a tide that lifts all boats.  Major agencies corroborate this, saying that penetration levels for programmatic display is already high and that their spend is slowing.  Display advertising has been declining as a percentage of TTD’s overall income, from 100% in 2011, to 48% in 2016, but is still a material part of the business.   Display advertising spend is growing mid-single digits and TTD earns higher than corporate average margins on this segment.  With take rate compression, which seems inevitable, there is a chance that display revenue growth could actually turn negative.

As display slows, management hopes that other verticals will continue to provide growth avenues.  However, other channels could present a number of challenges and end up being less attractive businesses for TTD.  

Mobile will continue to be a higher growth channel.  However, management is disingenuous about the threat of Apple’s Safari Intelligent Tracking Prevention (ITP), which will make it more difficult for third parties to track users.  This exchange is from the 2Q17 earnings conference call and management’s commentary seems misleading:

Safari is estimated to have about 5% market share of desktops, so the changes from ITP might not have much of an impact on TTD’s display business, but management has touted mobile as one of their growth areas.  Safari has about 50% market share in mobile browsers, so it does seem like ITP will have a negative impact on this crucial growth avenue.   AAPL was motivated to do this out of concern for user experience; it does not seem too concerned about publishers of advertising content.

Social and video are two of the highest growth areas, but for the most part are dominated by “walled gardens.”  Those outside the walled gardens are clearly not happy about it, but there is not much they can do about it, so growth rates here may end up disappointing.  GOOG and FB account for almost 2/3 of total digital spend and virtually all of its growth, effectively crowding out the DSPs, so as the industry evolves away from desktop display, TTD will find it harder and harder to compete.

Despite a major PR push from TTD, the feedback that I received from agencies also indicates that TTD is uncompetitive in Programmatic TV (PTV).   In fact, one ad executive at a major agency I spoke with says they do zero activations with TTD on PTV and that nobody thinks of TTD for PTV.  TubeMogul and a number of private players apparently have an edge in this vertical.  The key difference between digital advertising for display, which is a commodity, and television is that it is quite easy to target a desktop with a cookie and have the personal identifiable information (PII) for that desktop and more difficult to do with television.  Connected televisions are still in their infancy and longer term there will be a transition to targeted advertising via television, but people tend to watch the same types of things repeatedly (whereas on a desktop people often browse for a wide variety of things and with an intent to purchase, which is crucial), so there might be less need to buy advertising programmatically, since the PII stays relatively constant it does not have to be done dynamically or in real time.  PTV also has much lower take rates, so any success that TTD has here will be at much lower margins.  It is possible that TTD will make an acquisition in PTV to address its shortcomings.

International expansion is an area that should continue to provide TTD with solid growth.  However, these market are smaller and more fragmented.  They also have different laws and regulations also, which are often more stringent, so these conditions generally mean that more resources have to be dedicated sales and administration on a percentage of revenue, so they will not scale as well as in North America. International generally has lower penetration in programmatic advertising, so there is a longer runway, but they are mostly likely to remain well behind North America adoption. 

As much share as digital is taking from traditional channels, there are growing concerns among advertisers about the effectiveness of certain spending.  Most notably perhaps was PG cutting its digital spend by over $100 million.  Even Gary Freidman of RH fame is sounding uncharacteristically rational and analytical when it comes to digital ad spend.  Digital growth will continue, but the industry is imperfect and in a constant state of flux.  What has worked the past few years is not necessarily will work in the next few.

AMZN is another “walled garden” that is becoming increasingly powerful in advertising.  I have been told that AMZN is in the process of hiring 3,000 ad tech professionals.  I think it is more likely than not that AMZN is going to be taking a lot of share from everyone.   One agency told me that AMZN is now getting 10% of their total digital spend, up from practically nothing a few years ago.  Some agencies had been loath to work with AMZN because AMZN did not provide the advertisers with data, but AMZN is now giving data back to advertising clients, which they are evidently very pleased with.

There are regulatory concerns regarding privacy.  In the EU in May, 2018, the General Data Protection Regulation is scheduled to go into effect, which restricts the usage of personal data and could force TTD to change its operations.  There are numerous other proposals around the world that could become potential headwinds to tracking internet users’ data.  For instance, in the US some regulators have suggested creating “Do Not Track” standards and the FTC has criticized the advertising industry’s self regulatory efforts to protect consumers and address these standards.  It is difficult to say what the long-term impact of any new legislation or regulation might be, but the current trends seem to be towards more consumer protection, which is not good for digital advertising and something worth watching.

All of which is to say that this is a nascent industry, which is changing very rapidly where the independent DSPs are not well positioned for the long haul, which brings us to valuation.  Could DSPs in their current form just be a stopgap measure for a few years as the digital advertising space evolves and matures?  Maybe, maybe not, but would you underwrite it at 40-50 years worth of current earnings? 

TTD will grow its top line 50% this year, but given the challenges I have cited I expect that it could begin to disappoint in the near future.  TTD is trading at 9x sales, 30x EBITDA, and 46x earnings for 2017, so any shortfalls could be impactful.  I expect as take rates compress and some other the other issues create headwinds that will cause revenue growth to come in below estimates in the out years and margins to fall, but at this point it is pretty difficult to predict much beyond the direction. 

Valuing TTD based on comps is challenging.  On the one hand, the sell-side seems to agree that to justify their price targets (more or less in line with where the stock currently is) it should be compared to SaaS companies, despite not having recurring revenues, or comparing it to ADBE, FB, GOOG, etc, which are very different business models and much higher quality companies.   One analyst in his bear case has revenue growing 15% through 2023 with adjusted EBITDA margins increasing from 30% to 47%!  Let me reiterate that is the bear case and he is the one guy with a hold rating on the stock so he is not that bullish by comparison (there are eight buy ratings by the way).  Commodity businesses generally do not earn >30% EBITDA margins for very long and I expect once margins begin to decline the stock will rerate very quickly.

On the other hand, most of the other ad tech companies have never been profitable and trade for less than 1x revenues.  The general lack of profitability of the industry might foreshadow the direction of TTD’s margins, but it would not be appropriate to compare it to these companies.  CRTO is the one profitable stand out, which has recently had some unflattering issues exposed, but the stock has more or less bounced back from and it is trading at 2.75x EV/Rev and 18x this year’s earnings.  If TTD manages to increase its revenues by ~65% in a few years to $500 million while somehow maintaining its margins, and it traded for the same multiple as CRTO, then the stock would be cut in half.  Also, TUBE was acquired by ADBE in 4Q16 for $540 million or about 2x revenues.  TUBE was unprofitable, so one could argue that TTD’s private market value is higher, but 10x rev might be a stretch.

Corporate governance issues: Insiders continue to be large sellers of stock.  There were secondaries in 3/17 and 6/17 for 7.3 million shares at $35.50 and 4.8 million shares at $52, respectively.  The company did not receive any proceeds.  There is also a dual share structure.  Insiders still hold about 10 million B shares with 81% of voting power, but at the rate they are selling it might not be an issue much longer as the B shares convert to A shares when sold.

Addendum: risk factors with regard to payment terms:

We are subject to payment-related risks and, if our clients do not pay or dispute their invoices, our business, financial condition and operating results may be adversely affected.

Many of our contracts with advertising agencies provide that if the advertiser does not pay the agency, the agency is not liable to us, and we must seek payment solely from the advertiser, a type of arrangement called sequential liability. Contracting with these agencies, which in some cases have or may develop higher-risk credit profiles, may subject us to greater credit risk than if we were to contract directly with advertisers. This credit risk may vary depending on the nature of an advertising agency’s aggregated advertiser base. We may also be involved in disputes with agencies and their advertisers over the operation of our platform, the terms of our agreements or our billings for purchases made by them through our platform. If we are unable to collect or make adjustments to bills to clients, we could incur write-offs for bad debt, which could have a material adverse effect on our results of operations. In the future, bad debt may exceed reserves for such contingencies, and our bad debt exposure may increase over time. Any increase in write-offs for bad debt could have a materially negative effect on our business, financial condition and operating results. Even if we are not paid by our clients on time or at all, we are still obligated to pay for the advertising we have purchased for the advertising campaign, and as a consequence, our results of operations and financial condition would be adversely impacted.

A substantial portion of our business is from advertising agencies that do not pay us until they receive payment from the advertiser, resulting in an increased length of time between our payment for media inventory and our receipt of payment for use of our platform, and our ability to collect for non-payment may be limited to the advertiser, thereby increasing our risk of non-payment.

Substantially all of the spend on our platform is from advertising agencies. Generally, we are contractually required to pay advertising inventory and data suppliers within a negotiated period of time, regardless of whether our clients pay us on time, or at all. Additionally, while we attempt to negotiate long payment periods with our suppliers and shorter periods from our clients, we are not always successful. As a result, we often face a timing issue with our accounts payable on shorter cycles than our accounts receivables, requiring us to remit payments from our own funds, and accept the risk of bad debt.

This payment process will increasingly consume working capital if we continue to be successful in growing our business. In addition, we typically experience slow payment by advertising agencies as is common in our industry. In this regard, we had average days sales outstanding, or DSO, of 95 days, and average days payable outstanding, or DPO, of 73 days at June 30, 2017. We compute our DSO and DPO as of a given date based on our average trade receivables or trade payables, respectively, for the trailing 12-month period divided by, for DSO, average daily gross billings for the period, and for DPO, by the average daily cost of media, data and operating expenses over such period. The average trade receivables or trade payables are the average of the trade receivables or trade payables balances at the beginning and end of the 12-month period. Historically, our DSOs have fluctuated over time. If our DSOs increase significantly, and we are unable to borrow against these receivables on commercially acceptable terms, our working capital availability could be reduced, and as a consequence our results of operations and financial condition would be adversely impacted.

Due to this imbalance in our DSOs and DPOs, we rely on our credit facility to partially or completely fund our working capital requirements. We cannot assure you that as we continue to grow, our business will generate sufficient cash flow from operations or that future borrowings will be available to us under the credit facility in an amount sufficient to fund our working capital needs. If our cash flows and credit facility borrowings are insufficient to fund our working capital requirements, we may not be able to grow at the rate we currently expect or at all. In addition, in the absence of sufficient cash flows from operations, we might be unable to meet our obligations under our credit facility, and we may therefore be at risk of default thereunder. We cannot assure you that we would be able to locate additional financing or increase amounts borrowed under our existing credit facility on commercially reasonable terms or at all.

Addendum:

Company Overview – SunTrust Robinson Humphrey

The Trade Desk is an advertising technology company founded in 2009 that provides ad-buyers, especially agencies, enhanced tools to improve the buying process using a self-serve platform. Ad buyers are able to create, manage, and optimize ad campaigns across a variety of ad format like video, display, audio, native, and social across devices and geographies. The company works with an assortment of partners to integrate with ad inventories, data providers, and publishers to give its ad buyers reach and decision making capabilities to help them effectively leverage the platform. The company is headquartered in Ventura, California and has offices across the North America, Europe, and Asia. The company went public in September 2016 and had a secondary offering for existing shareholders.

Ad buyers are able to leverage The Trade Desk’s tools and third-party data to show the product effectiveness and its delivery of higher returns on ad spend. Typically, the technology is deployed directly with ad buyers, but management believes the great advertising ecosystem will benefit as publishers are able to generate higher yields on available inventory while consumers receive more relevant advertising.

FIGURE 13: AD ECOSYSTEM SHOWS THE TRADE DESK’S FOCUS ON THE BUY SIDE

Source: SunTrust Robinson Humphrey

The Trade Desk offers a host of services such as: 1) allowing for the building of a proprietary technology on top of the Trade Desk platform, 2) utilizing third-party data to effectively measure and optimize digital ad campaigns, 3) using first-party data from the ad buyer or their client to improve campaign performance, 4) observing and/or managing digital ad campaigns on a realtime basis, 5) ability to purchase digital media programmatically across media exchanges and sell-side platforms (SSPs), 6) access services such as The Trade Desk’s data managemen platform (DMP) and publisher management platform marketplace, and 7) provide attribution of offline sales as a result of digital campaigns.

Catalyst

take rate compression, slower growth