Travel Series: AIrBnB – “Baller IPO” or Desperate Hail Mary (including a 3.5 bn USD accounting time bomb) ?

Intro

Long term readers know that I have covered the (online) travel industry intensively and that I actually have build up a “post pandemic travel basket” recently. Therefore, I was really excited to look at AirBnB’s S-1 going public filing.

Airbnb is one of the most prominent Unicorns of the last decade. The company was founded in 2007 and has since then become one of the really big names in online travel. It describes itself as having established a new category of travel called “home sharing” and that all the hosts on the platform as well as the clients are a big “community” that make travel “Human”.

However the big “elephant in the room” is the question: Why do they go public now after 13 years ? Why didn’t they go public earlier or wait a few more months once the travel recovery really kicks in ?

There was already a lot of press coverage already for Airbnb in the past weeks. I think in general one could distinguish between the Bull Case and the Bear Case:

The Bull case :

  • It’s a “positive” global brand with strong growth potential and a huge TAM (all travel lodging globally ) 
  • People will rent apartments first if travel rebounds 
  • Restrictions maybe less a problem in cities after Covid-19

One of the biggest cheerleaders of the Bull case is clearly Prof. Scott Galloway who wrote a big post some days ago, putting the value of AirBnB at 120 bn USD with the following statement:

The only firms I can think of that have global demand/supply and brand equity, and an asset-light high-margin business are the credit card companies, which trade at 20+ multiples of revenue. Airbnb projects 2021 revenues of $5-6 billion, yielding a potential $100-120 billion valuation. 

In 2020, in the private markets, Airbnb shares have traded at a valuation ranging between $15-30 billion. The media pegs the IPO valuation at $30 billion. Put on your seatbelts, as the bankers will have no excuse to not price the shares at the high end of a recast (higher) range, and then reward their institutional clients with a Snowflake-like flurry when it begins trading. There’s something about weather and cruise analogies that becomes more appealing as you get older.

Some facts from the S-1 Prospectus and the Bear Case 

Scott Galloway is funny and successful, however some times it also makes sense to actually read a prospectus. Here are some of the observations I made and I will explain them in more detail: 

  • the business doesn’t really seem to scale 
  • compared to a similar company (at the same stage), the company is quite unprofitable
  • Their platform seems to be not so dominating and expensive to maintain
  • strange employee stock option program and a 3 bn “accounting bomb” is lurking
  • They have a Google problem, too

Numbers first

As I am a primarily numbers driven investor, let’s look at 2018 & 2019 pre covid numbers of AIrBnB

Airbnb 2019 2018
Sales 4805 3651
Cost of rev 1196 864
Operations and support 815 609
Product development 977 579
Sales and marketing 1622 1101
G&A 697 479
Total Cost 5307 3632
Operating profit -502 19
Interest income 85.9 67
Interest expense -10 -26
other 13.9 -12
EBT -412.2 48
TAX -263 -63.9
Loss -675.2 -15.9
     
Cost of Revenue in % 24.89% 23.66%
Op & supp. In % 16.96% 16.68%
Prod dev. in % 20.33% 15.86%
Sales & Marketing in % 33.76% 30.16%
G&A in % 14.51% 13.12%
Nights booked 327 250
GMV 37962 29440
Take rate 12.66% 12.40%
Avg. amount 116.09 117.76

What we can see from the numbers, that in 2019, despite a nice increase in sales, all cost factors went into the wrong directions, both from an absolute and relative point of view.

The only positive was a very small increase in the take rate, all other cost items increased quicker than sales which is really bad and shouldn’t happen with a company that uses any kind  of technology. In the prospectus, there is very little explanation. Only that they had to “upgrade their IT Platform” which for a “technology company” is odd, as I would assume that a technology company has always top notch technology and doesn’t need specific upgrades. 

Even more interesting is a comparison with a relatively similar business:

Airbnb 2019 2018 Company XYZ 20xx 20xx-1
Sales 4805 3651   5260 4355
Cost of rev 1196 864   1177 1275
Operations and support 815 609   0 0
Product development 977 579   43.7 33.8
Sales and marketing 1622 1101   1569 1171
G&A 697 479   639 475
Total Cost 5307 3632   3428.7 2954.8
Operating profit -502 19   1831.3 1400.2
Interest income 85.9 67   4 8
Interest expense -10 -26   -62 -32
other 13.9 -12   0 0
EBT -412.2 48   1773.3 1376.2
TAX -263 -63.9   -338 -309
Net Income -675.2 -15.9   1435.3 1067.2
           
Cost of Revenue in % 24.89% 23.66%   22.38% 29.28%
Op In % 16.96% 16.68%   0.00% 0.00%
Prod dev. in %+ 20.33% 15.86%   0.83% 0.78%
Sales & Marketing in % 33.76% 30.16%   29.83% 26.89%
G&A in % 14.51% 13.12%   12.15% 10.91%
Operating margin -10.45% 0.52%   34.82% 32.15%
Net margin -14.05% -0.44%   27.29% 24.51%
Nights booked 327 250   350 235
GMV 37962 29440   39100 28400
Take rate 12.66% 12.40%   13.45% 15.33%
Avg. amount 116.0917431 117.76   111.7 120.9

So despite being somehow similar businesses, the economics are totally different. The only parallels are gross margins and Sales and marketing cost, but anything else looks much much much better at company XYZ. All in all, company xyz seems to be way ahead in terms of profitability and scalability than Airbnb.

I don’t want to create too much tension, but company XYZ is actually Booking.com in the years 2012 and 2011 where they had around the same size as Airbnb has now (and were called Priceline). 

The comparison is not perfect, as the cost categories of the costs don’t fit. For G&A at Booking, I summed up all personell costs plus G&A plus depreciation. For product development, I took all IT costs for Booking.com at that time.

The almost shocking thing is that AirBnB has two significant cost lines, Product development and Operations and support, that do not have any counterpart at Booking.com at that time. This is what Airbnb writes about these cost categories:

Operations and Support

Operations and support expense primarily consists of personnel-related expenses and third-party service provider fees associated with community support provided via phone, email, and chat to hosts and guests; customer relations costs, which include refunds and credits related to customer satisfaction and expenses associated with our host protection programs; and allocated costs for facilities and information technology. We expect that operations and support expense will continue to increase on an absolute dollar basis for the foreseeable future to the extent that we continue to see growth on our platform. We also expect operations and support to increase in the near-term as a percentage of revenue, as we continue to invest in trust and safety programs. We are also investing in the near-term in initiatives to reduce customer contact rates and improve the operational efficiency of our operations and support organization, which we expect will decrease operations and support expense as a percentage of revenue over the longer term. We anticipate additional operations and support expense during the year in which we complete our initial public offering as a result of the stock-based compensation expense associated with our RSUs as described in the subsection titled “— Critical Accounting Policies and Estimates — Stock-Based Compensation — Restricted Stock Units,” as well as additional stock-based compensation expense going forward.

Product Development

Product development expense primarily consists of personnel-related expenses and third-party service provider fees incurred in connection with the development of our platform, and allocated costs for facilities and information technology. We expect that our product development expense will increase on an absolute dollar basis and will vary from period to period as a percentage of revenue for the foreseeable future as we continue to invest in product development activities relating to ongoing improvements to and maintenance of our technology platform and other programs, including the hiring of personnel to support these efforts. In addition, we anticipate additional product development expense during the year in which we complete our initial public offering as a result of the stock-based compensation expense associated with our RSUs as described in the subsection titled “— Critical Accounting Policies and Estimates — Stock-Based Compensation — Restricted Stock Units,” as well as additional stock-based compensation expense going forward.

Some business model considerations

Both, Airbnb and Booking are two sided market places which can exhibit good economies of scale and network effects.

Airbnb prides itself that they have 4 mn hosts who offer 5.7 mn properties, but economically that might be part of the problem: Maintaining a good relationship with this 4 mn hosts seems to be a lot more work intensive and expensive compared to the maybe 20-30 k hotels that Booking.com back then had on the offer side of its market place.

One big problem is of course also that similar to Uber drivers, the supply side can always list on different platforms. That is what we see right now with big pushes from Expedia and Booking into the “Vacation rental” market which is an important sub-sector of Airbnb’s offering and they seem to have no problems in adding supply. 

In Airbnb’s case I would argue that both cost categories mentioned above are rather “cost of revenue” and that AirBnB in my opinion significantly inflates economic gross profit. My assumption is that running a platform for millions of hosts with only one or two properties to share is not a very scaleable business.

The Google Problem

As Booking, Expedia and all the other OTAs, Airbnb has a Google problem too. Yes, some people might go directly to AIrbnb’s website, but many people use Google as a starting point for research on future trips.

This is what they write in their prospectus:

We also face increasing competition from search engines including Google. How Google presents travel search results, and its promotion of its own travel meta-search services, such as Google Travel and Google Vacation Rental Ads, or similar actions from other search engines, and their practices concerning search rankings, could decrease our search traffic, increase traffic acquisition costs, and/or disintermediate our platform. These parties can also offer their own comprehensive travel planning and booking tools, or refer leads directly to suppliers, other favored partners, or themselves, which could also disintermediate our platform. In addition, if Google or Apple use their own mobile operating systems or app distribution channels to favor their own or other preferred travel service offerings, or impose policies that effectively disallow us to continue our full product offerings in those channels, it could materially adversely affect our ability to engage with hosts and guests who access our platform via mobile apps or search. 

What we have seen in the past is that especially Google is going deeper in the the search funnel and pushing back organic search results. Airbnb seems to have stopped all paid Google advertising for the time being, which gave its Q3 2020 results a nice uplift bu this is not a valid long term strategy. In general, Online Travel has the big problem that many people start with Google and this really limits the attractivenss of the whole sector, as Google has become better and better extracting as much value as possible from these kind of businesses.

Other products

Back in 2016, Airbnb announced that they want to become a “full service” travel company. Looking at the S-1, I would say: Nice try but there seems to be nothing left of these ambitions. Despite their huge spend on product development, they still only offer one main product. Even the much hyped “attractions” segment is nowhere to be found. Again an indication that a lot of these “product” costs are maybe cost of revenue.

The brand

I am not a brand expert, but I mostly associate a brand with a certain guarantee how the end product will be, i.e. quality, design etc. This is not the case with Airbnb. Yes, Airbnb is great to use, but the ned product can be very good or very bad as Airbnb only has limited control over the end product and scams and frauds are clearly a problem. What does it help me if I have a great experience booking an appartment only to find out that it is a sxxxhole with faked ratings ?Therefore I would disagree totally with Galloway’s assumption that Airbnb is a great brand. 

Employee Stock option/RSUs and the 3.5 bn USD accounting “time bomb”

Now comes the really weird part: Airbnb has a very complicated capital structure with a lot of stock options and Restricted Stock Units (RSU). The RSUs are the biggest bulk of the employee incentive programs and carry an interesting vesting condition. This is what Airbnb writes in their S-1:

We have granted RSUs to our employees and certain non-employees, with substantially all of such RSUs vesting upon the satisfaction of both a service-based vesting condition and a liquidity-based vesting condition. During the quarter in which this offering is completed, we will begin recording stock-based compensation expense for these RSUs with a liquidity-based vesting condition. If this offering had occurred on September 30, 2020, we would have recognized $2.7 billion of cumulative stock-based compensation expense related to RSUs for which the service-based vesting condition was satisfied or partially satisfied, and the remaining unrecognized stock-based compensation expense relating to these RSUs would have been $0.8 billion as of September 30, 2020.

So these RSUs have not been expensed yet because it seems to be that their vesting requires a liquidity event which is either an IPO or a sale of the company. I have to admit that I haven’t seen this yet. 

For me this implies several interesting aspects:

  1. At first, these people who hold all these RSUs are most likely desperately hoping for an IPO to actually get their hands on the RSUs
  2. the past results of Airbnb should be reconciled to reflect this actual expense (stock based comp). I guess however that they will just record it as “non cash” expense in the fourth quarter. To me, even with its week numbers, Airbnb’s profitability in the past years is overstated
  3. Maybe this is also part of the answer to the question: Why the hell do they want to do the IPO now ?

The RSUs maybe also explain why there is only a very limited lock-up up for employees, who can sell 15% of there shares directly after the IPO which is also relatively unusual

Valuation:

It will be interesting at what valuation Airbnb will be IPOed. Scott Gallaway thinks it will be 30 bn which will then make it a quadruple following his applied valuation metrics (20x revenue).

In the past, Airbnb was one of the first “mega Unicorns” and commanded huge valuations already quite early in its life:

The big question is, what will Airbnb be valued at when the IPO commences? The group was last valued at $31 billion in 2017. But according to the Financial Times, private investors were trading indirect stakes in November 2019, valuing the company closer to $42 billion. However, in April this year, when it raised $2 billion, the equity portion of this was at a pre-money valuation of $17 billion.

In my opinion, the best proxy would be Booking.com. With pre-Covid net revenues of 15 bn USD, Booking trades with its current market cap of 85 bn at ~ 5,7x trailing sales. A similar multiple on AirBnB would justify a market cap of 27 bn, however with little upside in my opinion due to the issues with the business model as discussed above. Maybe they can grow for some years but profitability and scalability is a real issue.

Summary:

After looking at the S-1 prospectus, I am very disappointed. The AirBnB business model doesn’t look very attractive. The company tries to sell itself as a “human travel tech company” but in my opinion is a relatively simple online vacation rental agency with a difficult to scale business model and little success in broadening its offers.

In contrast to Scott Galloway I do think /as AirBnB itself) that its competitors are not Hotels but Booking.com, Expedia and Google who seem to be in general better at this than AirBnB.

However in this market, anything can happen with such a “brand name” IPO and that is maybe the ultimate reason why AirBnB wants to go public now and not wait until tourism recovers. Maybe Galloway gets his 120 bn valuation as currently anything is possible. But long term I don’t think that AIrBnB will be a good investment unless thy can change things fundamentally.

 

25 comments

  • Thanks for that great article! Do you have a view on Tripadvisor actually?

  • Thanks, great analysis! I admit that like you, I was very dissapointed when the numbers came out (thought it will be a profitable company), though from press publications during the past 2 years and silicon valley dismal track record in that regard, it was expected.

    BKNG CEO talked about the issue of scale in “alternative accomodation” almost every quarterly call. He seems to agree with your conclusion, and I think that the issue isn’t the number of properties but the number of rooms. Recall that AIRB has many properties like a single 2 bedroom apartment, owned/managed by one person. If the same person owns/manages 10 similar apartments (20 rooms), it’s easier to scale. See the link below for example.

    https://hotelanalyst.co.uk/2019/03/03/booking-looks-to-alternatives/

  • Last year’s article, but good background info on why management is under pressure to get the IPO done asap.

    https://www.businesstimes.com.sg/garage/news/for-airbnb-employees-dream-turns-into-disillusionment

  • W.r.t. unrecognized stock-based comp: DoorDash has the same. US GAAP is really being ‘abused’ heavily nowadays. Investors should therefore treat the reported numbers of such companies with great care. DoorDash’s numbers look a lot worse than many people currently assume (including the press headlines) if correctly adjusted for the non-cash personnel expenses.

    DoorDash:
    If this offering had been completed on September 30, 2020, we would have recorded $243 million of cumulative stock-based compensation expense related to the RSUs for which the service-based vesting condition was satisfied on that date, and would have an additional $225 million of unrecognized stock-based compensation expense related to the RSUs that would be recognized over a weighted-average period of 3.14 years. In addition to stock-based compensation expense associated with the RSUs, as of September 30, 2020, we had unrecognized stock-based compensation expense of $24 million related to outstanding stock options which we expect to recognize over a weighted-average period of 1.76 years.

  • Excellent write up. Thank you very much. I guess one question is if their marketing spend ever begins to fall as % of revenue. If it does, then this is a business with a “moat”. If it doesn’t, then the economics are not very attractive.

    • Thx. Yes, marketing costs is one thing, but this will remain very hard because of Google and Facebook sitting on the top of the funnel. But the other costs (product, support) are even more critical in my opinion.

  • “So these RSUs have not been expensed yet because it seems to be that their vesting requires a liquidity event which is either an IPO or a sale of the company. I have to admit that I haven’t seen this yet.”

    Uber did exactly the same thing last year: “Upon the Company’s IPO, the Company recognized $3.6 billion of stock-based compensation expense. Upon the IPO, shares were issued to satisfy the vesting of restricted stock units (“RSUs”) with a performance condition.”

    They obviously presented it as a non-cash, non-recurring expense…

  • Great post! I wasn’t expecting that numbers from Airbnb…

  • GNP-GlobalNosePicking

    I also got to listen Prof G (is it the cousin of Ali G?) on AirBnB. … Actually, I may have introduced Mmi to ProfG in a previous post. x-P !

    Agree on most if Mmi lines. I thing ProfG had a bias towards AirBnB after his stance on WeWork (“WeWTF”).

    For me, AirBnB is a ‘pass’, and for travel&leisure I have my personal bets already. Taking TerrySmith’s approach, better would be Booking.com as it already is a winner, and it has a goid ROCE.

  • Two points:
    1. Why do bulls think that “People will rent apartments first if travel rebounds”? I would expect people to prefer a hotel (cleaner and disinfected) to someone’s apartment even if more expensive
    2. On RSU and liquidity event: it happened exactly the same to Uber last year: “Upon the Company’s IPO, the Company recognized $3.6 billion of stock-based compensation expense. Upon the IPO, shares were issued to satisfy the vesting of restricted stock units (“RSUs”) with a performance condition”. Obviously they recorded it as a non-cash, non-recurring expense….

    There were also rumours in the past that AirBnB was “desperate” for a quick IPO as many RSU were close to expire (this is one of the few articles to mention it: https://www.businesstimes.com.sg/garage/news/for-airbnb-employees-dream-turns-into-disillusionment

  • great write up but I think the AirBnB product is very different than Expedia / Booking product. I usually travel as a group of 6 people and always use AirBnb for accommodations as it is inconvenient to put 3×2 people in 3 different hotel rooms. I know other larger families / groups do the same. Expedia and Booking.com do not come even close to offering the same product as AirBnB. If they get close to it, it will cost them a lot more. Furthermore, I just spent 3 months travelling through Europe and used the AirBnB experience a couple of times a week – it offers again much better experiences than other alternatives. I think the product is great and differentiated but the scalability will be tough.

    • That might be your personal impression, but it is not backed by numbers:

      Booking.com has 6.7m non-hotels on their platform (“homes, apartments, and other unique places to stay”)
      AirBnb has 7m worldwide listings.

      I would call this a draw for practical purposes.

    • I used to travel with the family, and it was rarely a problem to find a family room on booking.com. They also have apartments or even cottages for rent, although not as many as Homeaway. Noticeably, there are a number of people who advertise both on booking.com and on AirBnB.

  • Great summary and really interesting analysis, was really interesting to read. I never thought about it but it makes total sense that it will be very hard to scale such a business.

    I think your last paragraph really hit the nail on the head. Now is the best time to IPO because if cash-burning company like Tesla can be worth $500 billion, AirBnb can sure be worth $120 bn.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.