Uber's IPO Looks Attractive

In light of Uber Technologies’ UBER S-1 filings and upcoming IPO, we’ve updated our views on the company’s competitive position and financial prospects, which we initially published in July 2018. After reviewing the additional metrics and disclosures in the S-1, we have even greater confidence in our narrow economic moat rating and $110 billion valuation estimate for Uber.

Despite Uber’s market leadership and high penetration in ride-sharing, it appears that its growth in that space is decelerating faster than we initially anticipated, and the same may be true for its peers, such as Lyft LYFT . On the upside, Uber has grabbed market share in food delivery much more quickly than we expected, and we think strong growth potential for that service remains. We now estimate a $650 billion total addressable market for Uber by 2022, which may grow to over $1 trillion in 10 years, up from our July 2018 estimate of $630 billion. We continue to expect the company to become profitable and generate excess returns on invested capital on the back of moat sources such as network effects and intangible assets associated with user data. While CEO Dara Khosrowshahi has righted the ship after Travis Kalanick’s departure, we maintain our very high fair value uncertainty rating as Uber continues to face various regulatory and legal matters.

Uber Aims to Take You Anywhere and Bring You Anything, Anytime Uber is a technology provider that matches riders with drivers, hungry people with restaurants and food delivery service providers, and shippers with carriers. Its on-demand technology platform could eventually be used for additional products and services, such as autonomous vehicles, delivery via drones, and Uber Elevate, which, as the firm refers to it, provides "aerial ride-sharing."

Founded in 2009, Uber has become the largest on-demand ride-sharing provider outside of China. Since 2016, the company has completed over 10 billion trips in 63 countries. In 2018 alone, trips representing 26 billion miles were initiated on Uber. At the end of 2018, Uber had 91 million users who used the firm’s ride-sharing or food delivery services at least once a month. Uber refers to those users as monthly active platform consumers.

Uber helps people get from point A to point B by taking ride requests and matching them with drivers available in the area. The firm refers to this as personal mobility, which also includes micromobility, or shorter-distance transportation via electronic bikes and scooters, which Uber separately calls new mobility. Uber generates gross booking revenue from this service, which is equivalent to the total amount that riders pay. From that, Uber takes the remaining after the driver takes his or her share. Ride-sharing gross booking grew 32% in 2018, according to Uber’s latest S-1. Revenue increased 33% with the slightly higher average take rate, which we estimate will be declining.

We believe Uber has a 30% share and will be the leader in our estimated $411 billion total addressable ride-sharing market, excluding China, by 2023. It faces stiff competition from players such as Lyft (mainly in the United States) and Didi, in which the firm does have a 20% holding after the sale of its operations in China to Didi in 2016. While Uber no longer operates in China, it does compete with Didi in other regions around the world. The firm also has a 27.5%) stake in Grab, another former competitor in Southeast Asia. Globally, the market remains fragmented, and Uber competes with many local ride-sharing platforms and taxis.

Uber’s personal mobility service is based on two apps, one for riders to submit requests and one for drivers. In addition to vehicles, riders can see other modes of available transportation, like the nearest e-bikes or e-scooters. The types of rides available include UberX, which is a regular ride for one person or one party, and UberPool, in which the driver picks up more than one rider or party to deliver them to different locations in proximity. Uber also started an Express Pool service where different riders are picked up at certain locations, similar to bus stations. One company that focuses on such a service alone is Via. There is also UberBlack, the first service Uber offered, which is basically a luxury ride, priced higher, and provided by professionally licensed drivers. Most Uber requests now are for the UberX or UberPool services.

Like its competitor Lyft, Uber also provides rides for various businesses and their employees or clients, and refers to the service as Uber for Business. Uber Health provides nonemergency medical transportation for patients and caregivers.

While Uber’s network effect moat source has helped generate more riders, drivers, and requests, the company has also gone the acquisition route. The latest one was the $3.1 billion (funded by cash and convertible notes) acquisition of one of the Middle East’s largest ride-sharing providers, Careem, which has been providing such a service for more than seven years. Uber expects this acquisition to close at the beginning of 2020, after which we think it will drive further growth in the firm’s ride-sharing market share, gross bookings, and net revenue. Based on various estimates of Careem’s net revenue, we estimate Uber’s purchase of the firm was equivalent to around 10 times trailing net revenue.

In the long run, like other ride-sharing firms, Uber is investing to replace some or most drivers with autonomous vehicles. Those investments are taking place in the firm’s advanced technology group. While we expect further toward fully autonomous vehicles may help increase Uber’s take rate, it will probably keep drivers even in autonomous vehicles over the long run for safety reasons. The advanced technology group has received investments from various parties, including Toyota, which bodes well for both as the two may more quickly and smoothly commercialize autonomous vehicle technology and cars.

Uber Has a Narrow Moat as Riders and Drivers Continue to Hop On In our view, Uber's core business, the ride-sharing platform, benefits from network effects and valuable intangible assets in the form of user data. We think these sustainable competitive advantages will help Uber to become profitable and generate excess returns on invested capital. For this reason, we assign Uber a narrow moat rating.

Uber’s network effects benefit drivers and riders, creating a continuous virtuous cycle. Drivers and riders make up the supply and demand in ride-sharing, respectively. As a first mover in this market, where requests for rides from anywhere could be made in real time via a simple-to-use mobile app, Uber began to attract riders mainly via word-of-mouth. Growth in demand and further word-of-mouth marketing attracted more drivers, increasing the supply of Uber vehicles. In turn, as the number of drivers increased, the timeliness and reliability of the service improved, which drove the number of users or riders higher, which in turn attracted more drivers, all of which indicates a network effect. Uber was able to accelerate this network effect by focusing on smaller areas, such as San Francisco, before expanding into more cities. A comparable tech leader that profits from network effects is Facebook, which started at Harvard before expanding to all colleges and then opened up globally.

Growth in demand is driven not only by more users, but also likely by more rides or trips per user. Globally, Uber’s monthly active platform consumers and number of trips support this and our network effect assumption. During the last three years, total monthly active platform consumers and trips, along with trips per monthly active platform consumer, have been growing.

Plus, increasing supply is based on more drivers and further capacity utilization of each driver and the vehicle. Network effects, by definition, not only allow new network participants to benefit, but also existing ones. Thus, the riders on the platform benefit as more drivers are added, and existing drivers benefit from more riders, making the driver’s utilization even more efficient. A figure that we believe supports this and demonstrates the increase in vehicle capacity utilization is growth in average number of rides dispatched per unique Uber vehicle, which has been increasing gradually from 2015 through 2018.

While Uber has benefited nicely from network effects in recent years, we don’t believe it benefits from customer switching costs. In our view, the ride-sharing industry lacks barriers to entry or exit for customers and drivers. Both customers and drivers can easily switch to Lyft, while customers have other transportation options like taxis and public transit. In general, just as firms with network effects benefit from the positive flywheel effect when a network expands, they also run the risk of a negative flywheel if customers, drivers, or both start to depart, especially if the network lacks meaningful switching costs. Networks in tech are strong when faring well but can unwind quickly (think MySpace).

In early 2017, Uber faced much criticism for appearing to support an immigration order signed by President Donald Trump and attempting to profit from protests related to that event in New York City, both of which led to the #deleteUber campaign launched on Twitter and eventually tainted the Uber brand and reputation. Then CEO Kalanick was appointed to Trump’s economic advisory council, to which some employees at Uber objected. The lack of exit barriers and switching costs for riders and drivers was on display during this period, as other ride-sharing providers, such as Lyft, made headway in New York City and experienced faster growth in trips as riders easily downloaded apps for other services.

While we may have witnessed a slight pause in Uber’s network effect in 2017, we think the firm’s return to faster growth in trips serviced supports our assumption that the platform still benefits from the moat source characteristic. After February 2017, growth in Uber trips reaccelerated and hit the triple-digit rate again in April. At the same time, Lyft’s growth slowed a bit. In addition, Uber continues to dominate the New York City ride-sharing market, with a 2018 average daily trips run rate more than 5 times that of Lyft, based on data provided by the Taxi and Limousine Commission. In 2017, Uber received 4 times more ride requests than Lyft. However, there remains a risk that what currently appears to be a strong network effect for Uber could reverse.

In our view, Uber’s ride-sharing network effect can also help the firm tap into other markets and generate additional revenue streams. An example is the meal takeout and delivery market, in which Uber has gained traction with its Uber Eats service. According to Recode, Uber Eats has grabbed share from Grubhub GRUB and has more than a fifth of the U.S. market. In Uber’s S-1, the firm mentioned a lower take rate, which implies that Uber Eats is willing to compete on price in the short term.

Regarding the network effect, the same can be said about Uber extending its reach into the bike-sharing and freight brokerage markets. Perhaps a good comparison for adjacent network effects, where one strong network enables a firm to expand elsewhere, is Microsoft: Its dominant network around Windows allowed it to leverage its strength into the productivity software market with Microsoft Office. Similarly, Uber’s strength in ride-sharing might be leverageable elsewhere.

There are concerns about whether Uber’s network effects can remain an economic moat source if the firm is forced to incur additional costs imposed by municipal, state, or federal regulations. For example, given the firm’s tarnished reputation and negative feedback from riders, Uber may be forced to conduct more thorough background checks on driver applicants, such as adding fingerprinting to the process. Plus, like Lyft, the company is implementing an annual background check on all of its drivers, which we estimate will add an average of $450 million annually to the firm’s operating expenses and lower our projected 20%-plus 2028 operating margin by 500-1,000 basis points. However, we still expect the firm to hit profitability by 2024 and generate excess returns on invested capital for at least 10 years beyond.

Other concerns include whether Uber will have to meet minimum wage or minimum pay per trip requirements for its drivers. If so, we think the company will implement those policies as a trade-off and probably take a higher percentage from the gross revenue generated per ride. With this, we think the firm’s take rate will remain in the 20%-30% range, likely at the lower end. However, while paying drivers a minimum wage may lower vehicles’ capacity utilization, we think it may help attract them.

Additional worries surrounding Uber’s network effect moat source include the potential impact of autonomous vehicle adoption on car ownership. Some believe autonomous vehicles could attract more car buyers and increase car ownership, thereby possibly lowering demand for ride-sharing services, as self-driving cars may diminish annoyances and other costs currently associated with driving in traffic or long distances. We disagree, as we think the availability of autonomous vehicles on ride-sharing platforms will significantly reduce the necessity of car ownership, which also lowers the return on the higher car ownership costs.

More Data Can Get Riders There Faster and Safer As Uber benefits from its network effect, we think it gains access to valuable intangible assets in the form of user data, which we suspect helps the firm improve its services and increase its vehicles' capacity utilization. In turn, Uber's service may become more effective as it further monetizes its riders via real-time supply and demand-driven pricing. Uber may also use this extensive data and knowledge to tap into other markets.

Uber gathers data about riders and drivers. As it compiles data from the rider app about the locations users request rides to and from and at what times of day, Uber can get a clearer picture of its users’ tendencies. When combined with the user-generated driver ratings, we think such information helps Uber improve the timeliness of matching riders with drivers. Such overall enhancement in service could help the firm strengthen its network effect by increasing users and ride requests per user, which helps Uber gather additional data, possibly further increasing the overall value of the data. Simply put, data can also be considered an indirect network effect moat source. Google’s search engine is an example of the benefits from indirect network effects associated with data, as more searches lead to better algorithms, better results, and in turn, more searches. Similarly, more rides with Uber may lead to better algorithms and supply/demand balance, thus reducing wait times for riders and idle capacity for drivers, leading to more ride requests.

Uber also collects telematics via its driver app, which can monitor car movements and location, along with drivers’ tendencies. The firm continues to analyze such data to minimize chances of crashes by addressing or preventing unsafe driving. For example, Uber’s driving safety team has written code that analyzes driving tendencies, such as harsh braking or unnecessary speed acceleration, which the firm views as indications of bad driving that could increase the probability of crashing. We believe this may also help reduce insurance costs over time.

While Uber also collects its own mapping data, the firm still uses Google Maps for its drivers and has the product integrated into its platform. We think that in the future, Uber may be able to use its own mapping data for its drivers or to enhance the autonomous vehicle technologies it is working on currently.

Like other firms with intangible assets such as data as moat sources, Uber faces pressure regarding data security. Its largest debacle concerning privacy occurred in 2016, when data from 57 million users was stolen and Uber paid the hackers $100,000 to avoid having to report the theft. The breach ended up being reported in November 2017, more than a year after it happened. The data did not include Social Security or credit card information or location details, but it did include contact information and license plate numbers. While the data compromised wasn’t as bad as it could’ve been, Uber’s cover-up attempt led to a $148 million settlement with the U.S. Federal Trade Commission.

We do not view Uber’s brand as an intangible asset moat source. While many no longer say they “called a cab” and instead say they “called an Uber,” we think the firm’s brand has been tarnished by issues that surrounded the former CEO. While demand for Uber’s services has rebounded, whether such recovery also applies to Uber’s brand remains to be seen.

We do not believe that Uber benefits from moat sources such as switching costs or cost advantages, either. Uber users can easily download other apps, such as Lyft, and begin requesting rides. In addition, drivers can provide service for both Uber and Lyft, as they are not exclusive contractors. Last, while the development of Uber’s platform included very high fixed costs initially (according to PitchBook, the firm raised $19.9 billion from 2009 until before the IPO filing), we do not view it as a cost advantage for Uber, as other well-capitalized tech firms, such as Alphabet, can replicate what Uber has created. In addition, we view Uber as an asset-light business.

Eating While Riding or Sitting With Uber Eats While Uber helps riders get to another location, it also helps them order food and sit tight and wait for the food to be delivered with a service provided by Uber Eats. Uber Eats allows diners to easily search for meals and select the foods from many different restaurants to be delivered. The orders can also be submitted while the diner is getting a ride from an Uber driver.

The company is currently servicing more than 220,000 restaurants in more than 500 cities globally in what we now estimate will be a $191 billion market by 2023. We believe Uber Eats has 11% share in the global food delivery market today and that its share can rise to 25% by 2023.

Uber Eats is an example of how Uber can continue to cross-sell various services provided on its on-demand platform. According to the firm, around 16% of its total monthly active platform consumers, or around 15 million, ordered and received a meal in the fourth of quarter of 2018. Uber Eats generates net revenue based on its total gross booking, which is the total amount paid by consumers (excluding tips) who order the food delivery. From that, Uber Eats takes a portion, which the last two years has averaged at about a 17%-18% take rate.

Gross booking of Uber Eats grew 168% because of more aggressive marketing and pricing in 2018 and now represents 16% of the firm’s total bookings, up from 9% in 2017. In July 2018, we had estimated Uber Eats growth of only 25%. While we see Uber Eats growth decelerating during the next five years as it has more quickly grabbed a sizable market share, we continue to expect the business to grow at a 39% compound annual growth rate during that period. While globally this market is very fragmented, in the U.S., the main players are Uber Eats, Grubhub, and DoorDash.

Uber is using its on-demand matching platform to also tap into the logistics market. Uber Freight matches carriers with shippers as it makes the demand and supply visible in real time. As a broker that matches the shippers and carriers, Uber Freight gets a piece of the total amount exchanged between the two parties. Uber is well positioned to take a substantial part of this nearly $20 billion market in the U.S. In only about two years, Uber Freight is already working with 36,000 carriers and over 300 shippers. We think it faces competition from market leaders, including C.H. Robinson CHRW , which are also moving into digital freight matching. Uber’s S-1 says it plans to also tap into the European market, which represents additional growth opportunities. The European market is only a bit more than 50% of the U.S. market.

Uber and Its Peers Cannot Avoid the Laws, Old and New Uber faces many regulatory hurdles that we think it can either overcome or has the flexibility to adapt to via price and take rate adjustments while maintaining its market-leading position. However, legal and regulatory pressure will probably persist.

In many states in the U.S., the firm and its competitors are forced to conduct more in-depth background checks on their drivers. Plus, the safety of the individually owned vehicles become more important, and many states now require vehicle inspections and driver training. In addition, various insurance coverage not only for the drivers and riders, but also for third parties are required. While some of these requirements increase the costs associated with each ride, they may also help Uber or its rivals maintain their drivers and riders, thereby ultimately lowering sales and marketing costs. In other countries, Uber faces more stiff regulations, such as the requirement for each driver to be professionally licensed. In some countries, standards and requirements imposed on transport services are also likely to apply to Uber.

In some areas, the large demand for rides, which has also increased the supply of drivers, has resulted in extreme traffic congestion, which in turn has forced lawmakers to act. In New York City, licenses for ride-sharing service providers are now limited as the city researches the increased congestion and looks for possible solutions.

Uber also faces pressure to pay a minimum amount per trip to its contracted drivers, which we think could create a barrier to entry for smaller players, as the more established companies like Uber, with their large rider and driver networks, can pass the additional cost on to riders.

We Calculate a $740 Billion Addressable Market by 2023 Based on the scalable, successful implementation of an on-demand platform, we think Uber's total addressable market consists of not only ride-sharing, which includes the takeaway from taxi markets, but also what it can grab from public transportation, meal delivery, and digitizing the logistics of transporting cargo. We estimate Uber's total addressable market to be $740 billion by 2023, growing at a 21% five-year CAGR. In 10 years, Uber's TAM might well be worth over $1 trillion, in our view.

Uber values its total addressable market at $12.3 trillion, although it admits that it does not expect to address such a market in its entirety. We also believe that Uber can capture only a portion of this $12.3 billion market and that only parts of this large TAM are truly addressable for Uber and its peers. For example, we think it is unrealistic to view ride-sharing, e-bikes, and e-scooters as complete substitutes for public transportation or car ownership.

Taxi and Ride-Share We estimate Uber's addressable taxi and ride-sharing market growing at a 24% five-year CAGR to $411 billion by 2023. We started by looking at one of the firm's largest markets, New York City, which has substantial amounts of public data available for the taxi market. It is also a robust ride-share and taxi market, making it a reliable base to understand market share potential in the United States for the ride-share and taxi industries. New York City's Taxi and Limousine Commission publicly releases revenue and trip data on total taxi, Uber, and Lyft trips. We used this data as a base for the current taxi and ride-sharing market size in New York.

We then compared New York City’s total revenue with total U.S. taxi revenue estimates from Statista to assume New York City’s 21% composition of the U.S. taxi and ride-sharing market until 2019, after which we believe New York City will represent less share of the overall U.S. taxi and ride-sharing market (down to 10% in 2023) as ride-sharing continues to take revenue from the taxi market.

Using assumptions from Statista, we then applied our estimated U.S. market to project the size of such market globally. Combining our global estimates for the taxi and ride-sharing submarkets, we discounted this figure by excluding China’s ride-sharing potential, as Uber China conceded to Didi in 2016. We then estimate that Uber’s addressable global taxi and ride-share market will total $411 billion by 2023, growing at a 24% CAGR from 2018. Through 2028, we expect this market to be valued at nearly $600 billion.

Food Delivery We estimate the food delivery/takeout market to be over $190 billion in five years, representing an attractive 20% five-year CAGR. We think this market can grow further to over $270 billion by 2028. We based our estimate on our U.S. market estimate used in our equity analysis of Grubhub and global estimates from Statista.

Digital Freight Matching With most freight brokerage business taking place in the U.S., we believe that's where Uber will focus. We used the midpoint of Morningstar's estimate of the U.S. freight brokerage market at $15 billion-$20 billion and expect it to grow at a 9% CAGR through 2023 to $30 billion. We then assumed Europe's freight brokerage market to be slightly more than half of the U.S.' and combined the two to arrive at our $46 billion estimate in 2023. We expect this market to grow at an 8% CAGR through 2028 to nearly $63 billion.

Public Transport We estimate the global public transportation addressable market for Uber and other ride-sharing service providers to be $82 billion by 2023. Based on municipal commissioned and academic reports, we identified public transportation to be a substitute market from which Uber takes partial share. According to a report by the UC Davis Institute of Transportation Studies, ride-sharing takes approximately 6% of the public bus share and 3% of light-rail share. Considering the weighting of public buses and trains in the U.S., we estimate that ride-sharing took 5% of public transportation market share in 2017. Going forward, we increased this take rate based on our belief that Uber's network effect will strengthen, which will make it relatively easier for people to share rides. We took IBIS' global public transport market estimate of $500 billion in 2017, along with its 2.1% CAGR assumption, to arrive at our 2023 $82 billion valuation for the global transportation market that ride-sharing providers can target.

Looking at 2024 and beyond, we reduced our estimate for what we believe ride-share companies can take from public transportation. We lowered this assumption based on ride-share companies providing roughly 5% more rides to public transport rides, as reported in a recent Journal of Urban Economics study published after our first report on Uber. Compared with the UC Davis study, this study looks at Uber's effect on public transportation over a longer span of time and notes that it takes several years of Uber's presence in a city for the service to contribute to greater public transport rides. This idea of ride-share providers adding rides to public transport isn't incompatible with UC Davis' study, which noted ride-share takes rides from public transport. We ultimately think that ride-share companies achieve a net gain from taking away from public transportation rides. However, we think that, percentage-wise, this gain is less than what it could have been because public transportation ridership is increased by ride-share providers' situation-dependent complementary nature to public transport--like in "last mile" cases.

New Mobility We estimate the U.S., Canada, Latin America, and Europe bike-sharing addressable market for Uber and its peers at $4.3 billion by 2023, which represents a 4% CAGR from our 2018 estimate. We believe Uber's total addressable market through 2023 will consist mainly of the U.S., Canadian, Latin American, and European market potential for bike-share. We first looked at the U.S. via Citi Bike, New York City's collection of public docked bikes, and considered its publicly disclosed total revenue to determine the total addressable market in New York City. We slightly lowered our bike-share total addressable market for New York City after adding 2018 actuals for total trips and revenue per ride--both lower than we expected. We then calculated the revenue per capita that New York City generates from Citi Bike. Knowing that bike-sharing depends on dense populations, we assumed a discounted revenue per capita for the rest of the U.S. market, for which we arrived at a nearly $1.8 billion valuation by 2023.

We used our estimate for the U.S. market to derive a market valuation for Canada, Latin America, and Europe. We assumed a discounted revenue per capita for Canada, as we assumed that that impact of Canada’s similar population density on revenue per capita will be offset by better public transportation when compared with the U.S. We also estimated lower revenue per capita for Europe because while population density might be higher, overall ownership and usage of bicycles, along with availability of public transportation, might be higher. As for Latin America, we also discounted our revenue per capita assumption. While Pew research suggests bicycle ownership to be slightly lower than the U.S., at 49%, (though not all Latin American countries were included in the study), we think lower prices in that region will lower bike-sharing revenue per capita when compared with the U.S.

Ultimately, these calculations led us to a nearly $7.7 billion addressable market by 2023, representing a 14% CAGR. However, we then heavily discounted this figure by our projections for what scooter-sharing could take from bike-sharing. Discounts led to a final bike-share total addressable market estimate of $4.3 billion by 2023, representing a 4% CAGR from 2018.

Scooter-Share We estimate the U.S., Canada, Latin America, and Europe scooter-sharing addressable market for Uber and its peers to be $5 billion by 2023, which represents a 21% CAGR. Within these markets, we estimate that most of the scooter-sharing addressable market is a matter of how many bike-share rides scooter-share rides can replace. We believe this substitution effect has the potential to be 45% by 2023. This figure is based on our assumptions from the substitution rate in 2018, which we estimated by looking at San Francisco bike rides during a temporary ban in 2018. San Francisco's Ford GoBike usage increased by 36% within the month the scooter ban was announced. Adjusting for what we estimate the growth in bike rides to have been as well as for cities with bikes but without scooters, we get to a 15% substitution effect. While markets like Charlotte and Austin have already recorded scooter trip miles at over 7 times dockless bike rides or total trips at 4 times docked bikes, which we infer to signal a higher substitution rate, we discount this rate severely after accounting for regulation issues and some popularity to be only temporary. It is important to note that while we see high uncertainty in our substitution rates due to scooter legislation, the substitution rate has little effect on Uber's estimated total addressable market.

While electric scooters are illegal in New York City, we used New York as the base case for calculations to be consistent with our bike-share total addressable market methodology, since we believe a significant portion of scooter-share’s addressable market depends on acting as a bike-share substitute. We first estimated what scooter-sharing could replace of Citi Bike rides and multiplied this by our estimates for Citi Bike’s total revenue based on historical publicly disclosed figures. We then calculated the number of new rides we think scooters can attract because of their differences from bikes (faster and no sweat).

After summing the rides taken from bikes and new rides, we then estimate revenue per capita that New York City would generate from scooter-share rides. We use Bird, Lime-S, and Spin’s average revenue per ride based on an average trip distance of 1.42 miles, as reported by Charlotte city data, and a speed of 6.55 mph. Knowing that scooter-sharing depends on dense populations, we assumed a discounted revenue per capita for the rest of the U.S. market, for which we arrived at a $1.1 billion valuation by 2023.

We used our estimate for the U.S. market to derive a market valuation for Canada, Latin America, and Europe. We assumed the same revenue per capita for Canada because of its comparable population density with the U.S. We estimated lower revenue per capita for Europe because while population density might be higher, overall usage and personal ownership of scooters and their substitutes are expected to be higher than in the U.S. As for Latin America, we did not discount revenue per capita, as we believe bikes to be a substitute for scooters, and a Pew Research study says bicycle ownership is slightly lower than in the U.S., at 49%. Considering all four regions, this led to our $5 billion 2023 valuation for the scooter-share market, which represents a 21% CAGR.

We Assign Uber a Fair Value Estimate of $110 Billion We have updated our valuation model for Uber in light of additional details provided in the company's S-1, but we maintain our $110 billion market capitalization estimate. Based on the estimated fully diluted share count, a $110 billion market cap should equate to a fair value estimate of $58 per share upon Uber's IPO. Our fair value estimate represents enterprise value/net sales multiples of 8, 6, and 5 in 2019, 2020, and 2021, respectively. On an EV/gross billings basis, our valuation represents net revenue multiples of 1.8, 1.4, and 1.1. Our valuation is about 23% above the valuation implied by the midpoint of Uber's initial estimated IPO price range of $44-$50 per share. We project that Uber's net revenue over the next five years could grow at a 21% CAGR, in line with gross bookings growth rate we assume for Uber's $730 billion total addressable market.

We expect strong net revenue growth for Uber at a 19% 10-year CAGR through 2028, resulting in net revenue of $64 billion (representing $312 billion gross revenue or bookings), up from $11.3 billion (equivalent to $50 billion gross revenue) in 2018. Our gross bookings growth rates imply that Uber has 18% of its total addressable market today but will capture 31% of the $1 trillion addressable market we estimate by 2028.

We project that Uber will stay atop the ride-sharing market (including bike and scooter sharing and excluding certain regions, such as China) and that its ride-sharing gross bookings growth will be a bit slower than the overall market. We project Uber’s bookings growth at a 20% five-year CAGR compared with our estimate of 24% for the total ride-sharing market for the same period, as Uber has already solidified its position as the market leader. Plus, we think Lyft is positioned to gain some additional market share from Uber in the U.S., as the impact of bad publicity for Uber in 2017 could persist. We think such growth will be driven by Uber’s continuing expansion in more cities and regions globally, acquisitions, such the purchase of Careem, plus an increasing adoption rate as the company attracts more users. Because ride-sharing also represents a substitute for public transportation, we think it can take revenue from transit, which we account for in our projections. Our 10-year gross booking CAGR for ride-sharing is 18%.

While the meal takeout and delivery services market remains fragmented and Grubhub and DoorDash are fierce competitors, we think Uber can solidify its position, given its success in ride-sharing. Unlike Grubhub, Uber has expanded its Uber Eats services worldwide. In our view, Uber’s network effect moat source can help provide more timely deliveries for restaurants and enhance its position in the market as it attracts more drivers in what we estimate will be a $191 billion space (excluding China) by 2023. We model 39% five-year average revenue growth for Uber Eats, outpacing our estimate of the industry’s 20% average annual growth, as Uber Eats’ global expansion is still in its early stages. We apply a 10-year 28% gross revenue CAGR assumption for Uber Eats compared with our 14% assumption for Grubhub. By 2028, we expect Uber Eats to have grabbed around 35% of this market globally, becoming the number-one player worldwide. In the U.S., various reports indicate that Uber Eats and DoorDash continue to grab market share from the once-leading player, Grubhub.

Regarding the freight brokerage side of the business, Uber Freight, we think the company will face fierce competition from market leaders such as C.H. Robinson. While our model calls for Uber Freight gross revenue to grow 17% annually through 2028, that business will continue to account for only around 1% of Uber’s total gross revenue.

As Uber continues to attract more riders and assign drivers to requests more quickly, we think overall vehicle capacity utilization will increase, possibly widening the gross margin (as a percentage of revenue) to 62% during the next 10 years from 50% in 2018. Besides the driver take rate, which is netted out of Uber’s revenue, we believe a portion of Uber’s cost of goods sold is fixed and that revenue will grow faster than these costs, leading to gross margin expansion. In addition, we are assuming that the firm’s collection and analytics of data related to drivers’ tendencies, routes taken, and traffic could help the firm improve safety and more accurately estimate liabilities, which could lower insurance costs a bit. We are seeing indications of this, as in 2018, insurance accrued represented around 28% of dollar increase in net revenue, lower than our 2017 estimate of 31%.

We also project that Uber will benefit from operating leverage in the years ahead. The firm might be able to increase revenue faster than selling, general, and administrative costs, especially in the sales and marketing lines, while also spending less on operations and support costs. In fact, in 2018, operations and support expense per trip completed declined 18% from the prior year. However, we anticipate that research and development D will remain elevated as Uber invests in new ventures, resulting in only slight declines in R&D as a percentage of net revenue.

We think Uber is likely to begin generating adjusted EBITDA in 2022. Within our 10-year discounted cash flow model, we assume the firm will turn its current operating loss to operating income in 2024, while expanding operating margin to over 20% by 2028.