The Sharing Economy Checks In: An Analysis of Airbnb in the United States Implications on Traditional Hotel Development and Market Performance Going Forward By Jamie Lane, Senior Economist & R. Mark Woodworth, Senior Managing Director EXECUTIVE SUMMARY The sharing economy has become a prominent though not well understood economic phenomenon over the past several years. Airbnb is the market leader as it relates to the temporary accommodations industry. CBRE Hotels’ Americas Research compiled select information from STR, Inc. and Airdna, a company that provides data on Airbnb, for hundreds of U.S. markets to assess the relevancy of this sharing platform to the traditional hotel industry. Airbnb’s presence in key markets throughout the U.S. is growing at a rapid pace, with users spending $2.4 billion on lodging in the U.S. over the past year, according to analysis from CBRE Hotels. Over the study period of October 2014 – September 2015, more than 55 percent of the $2.4 billion generated was captured in only five U.S. cities (New York, Los Angeles, San Francisco, Miami and Boston), represents a significant portion of the lodging revenues in these markets. CBRE Hotels compiled select information for hundreds of U.S. markets to assess the relevancy of this sharing platform to the traditional hotel industry. From this data, the firm has developed an Airbnb Competition Index. This measure incorporates a comparison of Airbnb’s Average Daily Room rates (ADR) to traditional hotel ADR’s; the scale of the active Airbnb inventory in a market to the supply of traditional hotels, and the overall growth of active Airbnb supply in that market, into a measure of potential risk. New York was identified as the number one domestic market at risk from the growth of Airbnb, with an Airbnb Risk Index of 81.4, followed by San Francisco, Miami, Oakland and Oahu. Reports detailing the estimated performance of Airbnb facilities covering 59 U.S. cities, encompassing 229 submarkets, may be found at https://store.pkfc.com/airbnb-insights. January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 1 BACKGROUND The sharing economy has become a prominent though not well understood economic phenomenon over the past several years. The motivation for this study is to determine if and how short-term rentals booked through Airbnb are affecting hotel performance in the U.S. While Airbnb is not the only site for short-term rentals, it has become a leader in the industry. Exhibit 1 shows an indexed number of Google searches for each of the most popular short-term rental sites. The chart clearly shows how quickly Airbnb has taken the lead in terms of number of worldwide searches. Exhibit 1: Google Trends Search Index of Seaches for Major Short Term Rrental Sites Google Search Index Airbnb VRBO HomeAway FlipKey 100 90 80 70 60 50 40 30 20 10 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: Google Trends, Q4 2015. As the leading company within this new and expanding system of sharing, Airbnb operates a platform that facilitates the use of residential real estate as transient lodging by enabling people to rent a bedroom, a couch, or an entire home to guests on a short-term basis. The company was founded in August 2008 in San Francisco as a community marketplace for people to list and book unique accommodations around the world, either online or via their mobile application. As of December 2015, Airbnb had sixty million users providing access to locations in 34,000 cities in over 190 countries, averaging half a million stays per night. UNDERSTANDING THE RELEVANCY OF AIRBNB TO THE TRADITIONAL HOTEL INDUSTRY To gain insight into the economics of Airbnb, CBRE Hotels’ Americas Research compiled select information from STR, Inc. (STR) and Airdna for hundreds of U.S. markets to assess the relevancy of this sharing platform to the traditional hotel industry. Reports detailing the estimated performance of Airbnb facilities covering 59 U.S. cities, encompassing 229 submarkets, may be found at https://store.pkfc.com. January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 2 The degree to which a unit offered for rent through Airbnb is competitive with a traditional hotel is a function of many factors including location, availability, type of sleeping accommodation, number of guest rooms available within a particular unit, the character of the structure in which the unit is located (i.e. single family home vs. a Bed and Breakfast operation), minimum length of stay hurdles and price. While estimating the performance of Airbnb units largely takes the form of traditional hotel metrics (i.e. supply, demand, occupancy, average daily rate), there are several critical differences that must be addressed. As such, we introduce herein terms such as Active and Inactive Units, Supply Fluidity, and Revenue by Bed Type, among others. Individuals or entities that choose to rent their unit on Airbnb for rent are referred to as Hosts. Our initial research on the profiles of Hosts reveals that many have only one unit listed on Airbnb. It may be that these types of Hosts employ a more opportunistic approach to using the services of this online platform. Conversely, and in some markets, many Hosts list multiple units on Airbnb, thus the estimated demand captured and revenue realized by these multi-unit Hosts is disproportionately large. It seems reasonable that these multi-unit Hosts take a more business-like approach to using the services of Airbnb and thus potentially may be considered to be more competitive with traditional hotels in that market. As such, we provide specific estimates and profile the performance of multi-unit Hosts at the market and submarket levels. AIRBNB HOSTS RESPOND TO MARKET INCENTIVES According to STR, the U.S. hotel industry realized its highest annual occupancy level ever in 2015 and average daily rates increased at 2.5 times their long run average. These data alone suggest that the expansion of the sharing economy into the realm of the traditional lodging space is not handicapping hotel performance at the macro level – at least not yet. When examining the markets that have realized the largest growth in Airbnb unit supply, it is clear that a correlation exists with traditional hotel performance. Revenue per Available Room (RevPAR) is the hotel industry’s standard metric for evaluating the health of a market. This measure incorporates the average occupancy level in the market as well as the average daily rate (ADR) paid for each hotel room. By comparing hotel RevPAR to the number of Active Airbnb units 1(see Exhibit 2), it appears that Hosts respond to incentives, such as a higher rate and increased demand, causing more Airbnb units to appear in the market. This hold true at the macro level; where markets with higher ADRs and occupancy have the highest number of Active Airbnb units, and on the micro level; where we see a spike in the number of Active Airbnb units during major events such as the Super Bowl and New Year’s Eve. 1 Active vs. Inactive Unit – Units are considered Inactive if they have not been rented in the previous month, the schedule of days the unit is available for rent has not been updated in the previous two months, the host response rate to a potential booking is low, or all the days in the p r e v i o u s m o n t h a r e “ b lo c k e d ” ( i. e . n o t a v a il a b l e for rent for reasons other than paid occupancy). All other units are considered Active. January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 3 Exhibit 2: Hotel RevPAR and Airbnb Units in a Market. Source: Airdna, STR, Inc., CBRE Hotels’ Americas Research, Q4 2015. Note: New York and Los Angeles have been excluded as outliers because of the high number of active Airbnb units located therein. THE LEVEL OF SUPPLY IS CONSTANTLY CHANGING During the 12 month period ending September 2015, the number of Airbnb units that were posted on the Airbnb site for at least one day or more exceeded 420,000. We restricted our analysis of these units to those that we considered to be Active Units1. This limitation was imposed to achieve a more realistic view of the Airbnb supply in the U.S that is most relevant to traditional hotels. With this restriction to Active Units only, approximately 100,000 Inactive Units are eliminated from consideration. Viewed differently, roughly 24% of the units listed on Airbnb at some point during the year never received a booking. It is also important to recognize that the number of Active Units available from month-to-month and even day-to-day varies greatly. For example, there were approximately 152,000 Active Units on an average day in September 2015. In January of 2015, the number of Active Units available averaged 79,000. The availability of units generally varies because of the seasonality of leisure demand in the market which peaks during the summer and fall and hits a trough during the winter months. Exhibit 3 shows the estimated average daily Active Units available on Airbnb per month as well as the average daily supply of units. The difference between the number of Active Units and the available supply are blocked nights (defined as the number of nights an Active Unit is not available for rent). There are a variety of reasons why a Host would periodically not make their unit available for rent. Examples include owner occupancy, friends and family visits, etc. These blocked nights are not included when computing Airbnb unit occupancy. January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 4 Exhibit 3: U.S. Airbnb Average Daily Active Units and Average Daily Available Supply (Oct. 2014 – Sept. 2015) Airbnb Units ACTIVE UNITS AVAILABLE SUPPLY 180,000 160,000 140,000 120,000 100,000 80,000 60,000 40,000 20,000 0 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep 2014 2015 Sources: Airdna, CBRE Hotels’ Americas Research, Q4 2015. Exhibit 4 shows the breakdown of the top 10 U.S. markets for Active Airbnb Units by city and compares it to the number of Airbnb bedrooms available and the number of traditional hotel rooms in that city. New York is the top market in terms of the number of Active Units with roughly 23,000, followed by Los Angeles and San Francisco. The top 10 markets average 9.9% of Airbnb Units per traditional hotel room compared with 3.4% for the entire U.S. Exhibit 4: Top 10 Markets with Active Airbnb Units, September 2015 AIRBNB ACTIVE AIRBNB ACTIVE AIRBNB BEDROOMS PER MARKET HOTEL ROOMS UNITS/HOTEL UNITS BEDROOMS UNIT ROOMS New York 22,876 27,965 1.2 117,367 19.5% Los Angeles 13,023 17,967 1.4 98,166 13.3% San Francisco 6,428 8,790 1.4 51,561 12.5% Miami 5,199 7,368 1.4 51,498 10.1% Chicago 4,626 6,153 1.3 111,408 4.2% Washington DC 4,443 5,784 1.3 107,776 4.1% Boston 4,147 5,566 1.3 52,119 8.0% Seattle 4,044 5,601 1.4 42,455 9.5% San Diego 4,016 6,290 1.6 60,754 6.6% Austin 3,357 6,024 1.8 33,877 9.9% Top 10 U.S 72,159 97,508 1.4 726,981 9.9% Overall U.S. 173,057 277,256 1.6 5,031,645 3.4% Sources: Airdna, STR, Inc., CBRE Hotels’ Americas Research, Q4 2015 January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 5 We consider the number of Airbnb units in the market as the most important factor when assessing the relevancy of Airbnb to the performance of the traditional lodging supply in that market. Close behind in importance is the amount of growth that is taking place in the market for Active Units. Exhibit 5 shows the Q3 year-over-year growth rates for the top 10 and bottom 10 markets and how they compare with the national average. The markets with the highest levels of supply (i.e. New York, San Francisco, Los Angeles, etc.) generally have the lowest growth rates. The markets with the highest growth rates can be indicative of those that are becoming more popular on Airbnb; however, this can also be misleading if they are capturing the growth because of a special event. On average there was a 107% increase in the number of Active Airbnb units in Q3 2015. Exhibit 5: Top 10 and Bottom 10 Markets for Active Unit Growth in Q3 2015 Y-o-Y Growth 300% 250% 200% 150% 107% 100% 50% 0% Richmond Norfolk Cincinnati Indianapolis Philadelphia Denver Phoenix Omaha Houston National Average Charlotte Raleigh-Durham Kansas City Los Angeles New Orleans Washington DC Albuquerque San Francisco New York Savannah Sources: Airdna, CBRE Hotels’ Americas Research, Q4 2015 THE SHARE OF DEMAND ACCOMMODATED ON AIRBNB GROWS Using the data provided by Airdna, we estimate the number of bookings on Airbnb. In the year studied, there were roughly 16 million bookings on Airbnb. The volume of room-night demand generated by these bookings during this period, as compared to the number of traditional hotel rooms occupied during the same period as determined by STR, was 1.4%. This ratio increased from 1.0% at the beginning of the period to 1.7% by the end of the year. Thus, the volume of Airbnb demand is now approaching the STR, Inc. reported long run annual average growth of traditional hotel demand. January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 6 Exhibit 6: Airbnb Accommodated Demand as a Percent of Traditional Hotel Demand Demand: Airbnb/Hotel 2.0% 1.8% 1.6% 1.4% 1.2% 1.0% 0.8% 0.6% 0.4% 0.2% 0% Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep 2014 2015 Sources: Airdna, STR, Inc., CBRE Hotels’ Americas Research, Q4 2015 The data shown in Exhibit 6 clearly illustrates that Airbnb’s share of accommodated demand has grown significantly and that this share becomes more or less relevant to hotels depending upon the time of the year. December looks to be a strong month for Airbnb while in many markets hotels struggle to fill rooms. As an aside, and based on data from STR, the long run average level of demand growth in the traditional lodging industry is approximately 2.05. Thus, the volume of Airbnb demand nationally is rapidly approaching the level of organic growth for hotels. The implications on the volume of traditional hotel new development going forward are clear. REVENUE GENERATED IN THE U.S. ON AIRBNB IS CONCENTRATED IN A FEW MARKETS The results shown below in Exhibit 7 represent only a small portion of the estimated data available. We limit our analysis to the largest 59 domestic lodging markets that comprise our current Hotel Horizons® forecast universe. Collectively, these markets contain in excess of 48% of the traditional hotel stock in the U.S. Over a period of twelve months, ending September 2015, we estimate that there were roughly $2.4 billion in domestic Airbnb bookings and that more than 55% of the dollars generated were captured in only five U.S. cities. January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 7 Exhibit 7: Top 10 U.S. Markets for Airbnb Revenue from Oct. 2014 – Sept. 2015. $500,000,000 $450,000,000 $400,000,000 $350,000,000 $300,000,000 $250,000,000 $200,000,000 $150,000,000 $100,000,000 $50,000,000 $0 New York Los Angeles San Francisco Miami Boston Washington DC Austin San Diego Chicago Seattle Source: Airdna, CBRE Hotels’ Americas Research, Q4 2015. New York is dominating the U.S. market in terms of supply of Airbnb units as well as the number of units booked and revenue generated. While New York leads in terms of revenue, the year-over-year change in revenue in this market from the Q3 2015 over Q3 2014 (a gain of 45%), was one of the lowest of all major U.S. cities. This suggests that either the market is maturing or the New York Attorney General’s (NYAG) increased scrutiny over Airbnb is having an impact on Airbnb’s ability to keep growing. The fact that other top markets also have growth rates lower than the national average of 76% suggests that the relatively lower levels of revenue increase may be a function of maturity: Los Angeles (65%), San Francisco (44%), Washington DC (47%), Austin (54%), and San Diego (64%). Of top 10 ten markets shown in Exhibit 7, Seattle had the highest growth rate in revenue of 118%. While total revenue is important, it is also appropriate to look at Airbnb’s revenue as compared to traditional hotel room revenue at the market level. As illustrated in Exhibit 8, Los Angeles is the top market with roughly $250 million in revenue generated by Airbnb properties. This represents 5.7% of the $4.4 billion in room revenue generated by traditional hotels in Los Angeles over the same period, according to STR. January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 8 Exhibit 8: Top 10 U.S. Markets for Airbnb Revenue as a percent of Hotel Rooms Revenue from Oct. 2014 – Sept. 2015. 0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% Los Angeles New York Austin Oakland San Francisco Long Island Portland Miami Sacramento Seattle Sources: Airdna, STR, Inc., CBRE Hotels’ Americas Research, Q4 2015. THE PRESENCE OF MULTI-UNIT HOSTS A review of Airdna data reveals that the vast majority of Hosts using the Airbnb system have only one unit listed therein. It seems reasonable to expect that these Hosts generally take a more opportunistic approach in attempting to benefit from Airbnb. As such, the units controlled by these Hosts represent a diminished potential threat to traditional hotels. Conversely, multi-unit Hosts are more likely to take a proactive approach to the management of their units and, as a result, encroach on demand sources that historically have used traditional hotels. A review of Airbnb activity in New York yields insights to this issue. In December 2015, Airbnb made data available on roughly 60,000 units in New York2. In the summary accompanying these data, Airbnb noted that of the Hosts renting their entire home, 95% have one listing. Airbnb also stated that 75% percent of the Airbnb revenue in New York is earned by entire home Hosts who have one or two listings. We believe that these are important metrics as they provide insight as to how competitive Airbnb might be with traditional hotels in a particular market. Our analysis of New York employs a slightly different geographic definition and a different study period3. We also expanded the Host type considered to include private and shared rooms. Relative to the report issued by Airbnb, this decreases the percentage of Hosts with only one unit from 95% to approximately 80%. It is important to note that the number of Hosts with 3 or more units is only 7%. In addition, the revenue generated by Hosts with multiple units increased from the 25% as reported by Airbnb, to 29%. 2 https://www.airbnbaction.com/blog/data-on-the-airbnb-community-in-nyc 3 New York geography & Study Period January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 9 Hosts with multi-private and shared rooms do not appear to be generating much revenue when compared to Hosts with entire home units. While 29% is a large share, it is lower than the national average of 34% and pales in comparison to the percentages we see in many other U.S. cities. In Miami, 19% of Hosts have 3 or more units and they generate 62% of revenue in the market. Other markets that have a high revenues generated by hosts with 3 or more units include Savannah (60%), Las Vegas (59%), Orlando (58%) and Oahu (52%). It should be noted that Airbnb appears to be actively limiting the number of multi-unit Hosts using their system. Should they be successful at this effort, the number of multi-unit Hosts may diminish. AIRBNB RATES ARE HIGHER THAN YOU MIGHT EXPECT Contrary to popular belief, Airbnb is not always the lower priced option for those seeking temporary accommodation. The average rate paid for an Airbnb unit over the 12 months ended in September 2015 was $148.42. This is 25% higher than the average hotel rate of $119.11 paid over the same period as reported by STR. This disparity holds true for most Airbnb Unit types as illustrated in Exhibit 9. Exhibit 9: U.S. Airbnb and Hotel ADR by Unit and Property Type - Trailing 12 Month Average (Oct. 2014 – Sept. 2015) APT, AIRBNB HOTEL ADR BED & BED & UNIT TYPE ROOMS HOME CONDO, OTHER AVERAGE AVERAGE PREMIUM BREAKFAST BREAKFA LOFT ADR ADR (DISC) 1 $130.29 $148.94 $119.03 $129.78 $144.51 $119.11 21% Entire 2 $188.71 $214.78 $179.19 $157.68 $204.03 $119.11 71% home/apt 3 $280.50 $302.05 $255.43 $194.07 $285.03 $119.11 139% 4+ $448.27 $476.65 $349.83 $400.71 $450.02 $119.11 278% Private room $74.95 $86.54 $77.10 $95.25 $80.67 $119.11 (32%) Shared room $44.20 $56.42 $45.45 $47.48 $51.10 $119.11 (57%) Average $153.64 $147.21 $113.49 $103.25 $148.42 $119.11 25% Sources: Airdna, STR, Inc., CBRE Hotels’ Americas Research, Q4 2015. Part of the reason for higher rates can be attributed to the amenities found at some Airbnb units: 90% offer access to a kitchen, 68% offer access to a washing machine, 67% have free parking, and 12% offer breakfast. Data provided by the American Hotel & Lodging Association reveals that the availability of amenities offered by hotels varies by chain and property type. On average, only 60% of hotels even offer a microwave while 82% provide complimentary breakfast. Seventy two percent of hotels offer free parking which is slightly more than the 67% of Airbnb properties. In addition, 34% of the Airbnb units have 2 or more bedrooms and the rates for units with 2 or more rooms are offered at a considerable premium to single room units. Exhibit 10 contains a comparison of the ADRs achieved by traditional hotels and Airbnb Hosts at the local market level. Markets that fall along the 45 degree line are those in which hotel ADR’s are equal to those January 2016 CBRE Hotels’ Americas Research © 2016 CBRE, Inc. | 10