What is the effect of LEED and Energy Star certifications on rents and occupancy rates of multifamily homes? : a spatial regression analysis focusing on the U.S. market

(1)

Amsterdam Business School

Master of Science in Business Economics

Dual degree in Real estate finance and Finance

Master’s Thesis

What is the effect of LEED and Energy Star certifications on

rents & occupancy rates of multifamily homes? A spatial

regression analysis focusing on the U.S. market.

By Collin Julien

(10603077)

January 2015

Supervisor: Marcel Theebe

(2)

Statement of originality

This document is written by Student Julien Collin (10603077) who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3 Acknowledgements

I would like to express all my thanks to Norm Miller from the San Diego

University, without you my thesis wouldn’t have been possible. Rare are the sources for multifamily databases, and I am very fortunate that you could provide me with one of the very first ones that cover the whole United States.

A very special appreciation and thanks to my supervisor Marcel Theebe. You have been guiding my work, encouraging me to always think further and giving me in-‐depth advices.

Thank you to my family for their support during all these years of studying and to all my classmates following this enriching study program.

(4)

Abstract

This paper investigates whether obtaining sustainable building certification, either LEED or Energy Star, entails a rental premium and occupancy premium for multifamily buildings. Additionally, the average rent concessions is studied in order to find out whether an owner accords fewer concessions if the property is green certified. To this aim, a hedonic model controls for building size and characteristics. A spatial regression is performed to control for location effects, quality and other missing variables. Moreover, the spatial lag model (SLM) accounts for asking rent

autocorrelation between neighborhoods that are spatially close to each other. This approach is applied to a final sample of 1,362 multifamily buildings in the United States in the 2001–2014 period. The dataset is cross-‐sectional and thus dummy variables are created for each year in which the rent level of a property is observed. The results indicate a significant rental premium for both ENERGY STAR and LEED certified

buildings. The results also show a significant positive relationship between the Ask rent and the different green labels. A green multifamily building (LEED or Energy Star) generates a rent premium of 8.9%, while LEED obtained 9% and Energy Star 8.7%. The different LEED sub-‐certifications were found to be 2% for LEED certified, 5.5% for LEED silver, 12.1% for LEED gold, and 21% for LEED platinum. Moreover, an occupancy premium of 0.79% was found for green buildings. Finally, green buildings in this sample appear to have 50% less concessions on average.

(5)

5 1. Introduction

Technological progress in the real estate industry has captivated investor’s attention since several years. They have observed the opportunity to develop buildings with decreasing their impact on the environment. This is an important contribution to the planet while knowing that buildings account for 39% of the world’s pollution (USGBC, 2011) with residential buildings alone accounting for 22.5%(U.S. Department of Energy, 2012). Projections for the U.S. show that in 2016 green buildings are expected to be worth $230 billion, compared to only $10 billion in 2005 (McGraw Hill Construction, 2014). Certainly, there is an interest in both developers and tenants for eco-‐friendly properties. Moreover, future real estate developments in the U.S. are now subject to new regulations, which are requiring higher environmental certification standards (McGraw Hill Construction, 2009). The regulations are applied to the main types of real estate: The housing and commercial real estate markets. Several researches exist on the impact of LEED or Energy Star certification for commercial real estate. Fuerst, F., McAllister, P., (2009), Miller, N., J. Spivey, and A. Florance (2008), Reichardt, A., Fuerst, F., Rottke, N. B., & Zietz, J. (2012) have discussed the effect of green certification on rent and transaction price levels. However, all existing literature for green buildings focuses on the commercial real estate market. Hence today very few articles exist on the green housing market. Brounen & Kok (2011), Deng et al. (2012) and Kahn & Kok (2013) studied the effect of green labels on the residential market, respectively for the Dutch, Singaporean and Californian markets. However the researches are mainly based on single-‐family homes and their transaction prices. Therefore, rare are the studies on multifamily homes and their rent premiums and/or occupancy rate premiums. This shortcoming is principally due to a lack of data availability, which is harder to get in the residential sector compared to the non-‐ residential sector.

Commercial real estate seems to offer a higher rent level and price if they are green certified. Therefore, looking at sustainable benefits on the housing market is interesting by the fact that the results can be compared to the ones of the commercial real estate sector. If the findings appear to be similar, then green buildings would have a strong relevancy in terms of revenues for investors on the real estate market in general. However, if a rent premium is found for residential buildings, it is expected to be smaller than the green commercial one. Indeed, the residential market is safer in terms of risks

(7)

7

LEED or Energy Star certified buildings, which are the two main green labels in the U.S., should have an increasing interest for investors if the cost premiums are inferior to the rent premiums or price premiums, ceteris paribus. However, the housing market is composed of single and multifamily homes. A new database is available to study the multifamily market across the whole United Sates. This is a very interesting opportunity to study the green impact on apartments and moreover due to the fact that there are very few existing related topics yet.

The research topic of this thesis is “What is the effect of LEED and Energy Star certifications on rents & occupancy rates of multifamily homes? A spatial regression analysis focusing on the U.S. market.” The database contains observations throughout the whole United States, which is a very large country and attractive market for real estate investors. However, the rent premium in this thesis only refers to the asking rent per building, and not the realized rent, as it is not observable in the database used in this thesis. As mentioned previously, this is one of the rare topics on the effect of LEED and Energy Star on multifamily homes. This is mainly what will be added to the area of research on sustainable real estate. Moreover, the geographical location of each building is given by the longitude and latitude. Therefore the methodology will allow for a spatial regression in order to overcome the problem of missing variables. If results show a significant rent premium for green certified apartments, which is the case for the commercial real estate sector (Fuerst & McAllister (2009), Miller et al. (2008), Reichardt et al. (2012)), then investors should study with more attention the extra costs implied for green residential properties. Therefore they can forecast whether or not a return premium is feasible. Hence the effect of this premium should stimulate developers to build greener properties in order to satisfy the investor’s increasing demand for green buildings. Likewise these greener developments would contribute towards a more sustainable environment, which is very important nowadays according to the growing population and expected number of 9.6 million inhabitants in 2050 (UN Report, 2013). Additionally, the rent premium differences between LEED and Energy Star will be performed, as well as the four different LEED certifications (Certified, Silver, Gold and Platinum). Moreover, The vacancy rates within the database will allow to research if eco-‐friendly apartments provide an occupancy rate premium. Last but not least, the average concessions are available and will be compared between sustainable and standard buildings.

(8)

Similarly to the previous cited papers focusing on commercial real estate, the methodology used in this paper will be a hedonic regression. The dependent variable will be the natural log of the rent per square foot, which will be determined by several independent variables. Each building contains the percentage of studios, and one, two or three bedroom apartments. The most relevant variables are the building class, the sub market, the location, amenities, the secondary type, and whether the multifamily home is certified LEED (certified, silver, gold or platinum), Energy Star or nothing. Excepted for the green labels, all the variables will be used in order to adjust for comparability. Indeed, to observe if an apartment has a higher rent due to its LEED or Energy Star certification, it has to be compared to the same secondary type, size, location and as more characteristics as possible, which can obviously have an influence on the rent level and hence be adjusted to compare the effect of LEED or Energy Star on a multifamily home. Moreover, the control for location will be accomplished by doing a spatial regression analysis. It will correct for missing variables like location and quality. Thus, the defined methodology used is a spatial hedonic regression analysis. Externalities for transportations will be corrected using a variable (Closest Public Transport) directly available from the database that observes the distance from CPT in miles. The regression will be cross-‐sectional as the data given is over one-‐year only. That is to say, all units are not observed in the same year. Dummy variables will be made for all different years. This is fundamental as rent levels can change a lot a year to another, especially those last years during the economic downturn. The database contains about 3,600 multifamily homes.

In chapter 2 this paper will begin with an analysis of the previous researches done on eco-‐friendly real estate. Those will cover both the few green residential studies and the consequent sustainable commercial researches and their findings. In chapter 3 the methodology used will be described: the statistical relationship between the independent variables and dependent variable will give the final equation in order to solve the research topic question. Hence this will point out the causal effect of LEED certification on the rent levels and its significance. Additionally, all the hypotheses from this research paper will be settled. In chapter 5, a full description of the data will be given in a new section with summary statistics. Finally, chapter 6 & 7 will respectively present the results of the hypotheses and robustness checks. Concluding remarks will follow to finalize this research paper in chapter 8.

(9)

9 2. Literature review

The following section discusses the existing literature on green buildings. The main objectives of the authors were to find out whether green pays off. In some cases, other interesting findings were discussed (i.e. higher occupancy rate, lower risk premium). Few existing articles on green residences were written as December 2014. The difficult availability and access to green residence databases is a constraint for authors with an interest in the eco-‐friendly residential market. The researches focused on the residential market as a whole or only on the single-‐home market. The articles on commercial real estate all used the Costar database, as it is one of the rare ones to be available for eco-‐friendly buildings as it contains ENERGY STAR and LEED properties. However, the methodologies and data used will be similar to the one that will be performed in this paper. Therefore, the following section will review both data and methodology of different existing studies, and obviously, on top of its hypotheses and findings.

At present, it is essential to define what is a green building and how it is certified LEED or Energy Star. But first of all, an overview of the Costar Group, the largest real estate database in the U.S. and the most used by authors that discussed green benefits.

2.1 Energy Star and LEED certification levels: General benefits and

definitions

The largest real estate data collector in the United States and the U.K is the Costar Group. They have over 4.2 million commercial buildings in their dataset, which makes them the number one in commercial real estate information. Therefore, the reliability of their information is trustful and large real estate companies such as CBRE global investors, Jones Lang LaSalle and Cushman and Wakefield are strong partners of the Costar Group. Large banks are also interested to follow Costar Group’s researches, as for example UBS, Deutsche Bank or ING (Costar Group, 2014).

The Costar Group has created late 2013 the largest multifamily database (Costar Group, 2014). Before this, the company only focused on commercial real estate, which was more attractive for its clients as this asset class can provide higher returns (De Wit et al., 2003). Since the credit crunch in 2008, the housing market has gain in importance as prices dropped severely (Brunnermeier, 2008). Besides the fact a large number of people lost their home due to the underwater phenomenon, large real estate companies have increased their interest on how much house prices can fluctuate. In fact, the

(10)

residential market, and especially more multifamily homes were perceived as a very safe asset class before the crisis.

In this new database developed by the Costar Group through different acquisitions, like for example Apartments.com in 2014, there are multifamily homes that obtained the label Energy Star or LEED. Those are two eco-‐friendly certifications that will be described later on.

An article of the Costar Group from March 2008 underlines the advantages of green buildings, either ES or even LEED (which requires a higher energy efficiency than ES). Those advantages are higher occupancy rates, stronger rents and sale prices, lower operating costs, lower cap rates and hence are more valuable (Burr, 2008). Obviously the “building” term used in the article refers to commercial real estate only. The rent premiums found by Burr (2008) and the Costar Group were: “LEED buildings command rent premiums of $11.33 per square foot over their non-‐LEED peers and have 4.1 percent higher occupancy rate. Rental rates in Energy Star buildings represent a $2.40 per square foot premium over comparable non-‐Energy Star buildings and have 3.6 percent higher occupancy rate”. Additionally, the company also studied the selling prices, “and, in a trend that could signal greater attention from institutional investors, Energy Star buildings are selling for an average of $61 per square foot more than their peers, while LEED buildings command a remarkable $171 more per square foot” (Burr, 2008). In fact, those are remarkable numbers for investors. Moreover, the group added “If you’re building today without LEED, you’re building in obsolescence”, which points out the importance of building greener and stepping towards a sustainable future.

ENERGY STAR (ES) is a federal program established by the U.S. environmental protection agency (EPA) in 1992. To obtain its certification, the energy efficiency of the building has to score within the first 25% of EPA. In 2005, 2,000 multifamily buildings and 350,000 houses received the Energy Star certification. This certification can be applied to any type of real estate. The number of ES certified buildings quintupled from 2007 to 2012, which represents a significant increase, especially during five years crisis period time (figure 2.1). This increase is related to both new developments and renovated buildings in order to obtain the ES certification.

(11)

11

Figure 2.1: Increase in ES certified properties in the U.S.

LEED certification (Leader in Energy and Environmental Design) was developed by the U.S. green building council (USGBC) in 2001 and has different levels of energy efficiency certification: standard certification, silver, gold and platinum (see figure 2.2).

Figure 2.2: The 4 different LEED labels

The criteria used for the rating are the water consumption, heater efficiency and usage of materials (origin and reuse). LEED certified properties have increased significantly since its creation, about 450 million additional square feet in 2010 (see figure 2.3). The number of LEED gold certified buildings also have increased. They require a higher standard of construction and thus extra costs for developers (see figure 2.4). This could indicate the profitable side of higher green certification levels.

(12)

Figure 2.3: Growth in the number of LEED buildings in the U.S.

Figure 2.4: Growth of LEED Gold certified offices in the U.S.

Sustainable labels for green buildings around the world can be named differently, like for example BREEAM for the U.K. or Greenstar for Australia.

Thus, eco-‐friendly buildings contribute to a large reduction of energy use and provide costs savings for tenants. Obviously, the larger the gap in energy costs between a green property and standard property the more attractive it should be for tenants. However, a higher rent is observed for those benefits in costs reduction. Eco-‐friendly buildings are important towards the environment as it represents a large reduction in CO2 emissions. Moreover, green buildings strengthen the brand image and reputation of a tenant (Frombrun and Schanley, 1990). Additionally, Turban et al. (2009) found an

(13)

13

“increase in worker productivity and retention rates of employees, but also reduced staff turnover and reduced employee absenteeism”.

Now that the different green certifications and benefits of eco-‐friendly properties have been accentuated, an in-‐depth analysis of the findings/results and methodology about different authors that used the Costar database (hence U.S. market like in this paper) will be accomplished in order to inspect whether green buildings outperforms standard ones.

2.2 Green residential market literature review

Kahn & Kok (2013) studies the effect of different green labels on the transaction price of single homes in the Californian market. Their hypothesis is to find a transaction price premium for green certified homes towards comparable unrated houses. The authors start by stating the advantages of sustainable properties and the definitions of LEED, GreenPoint and Energy Star.

The study uses the United States Green Building Council (USGBC) to obtain the GreenPoint and LEED certified properties. GreenPoint properties amounted for 4,900 houses and 10% them were labeled LEED. Energy Star rated properties where obtained through inside information by green rating agencies and amounted for 700 properties. The control group (unrated green single homes) amounted for 1.6 million homes. All variables were observed annually through the period 2007-‐2012.

The methodology used was a hedonic regression analysis. The authors therefore controlled for location and quality (using zip codes), size, age and other characteristics (e.g.: swimming pool, air-‐conditioning, balcony). An important fact of this paper is that green homes were on average 1.7 years old and houses of the control group were 32 years old on average. The authors controlled for age by creating different categories and creating a dummy for each age category. However, the more “old” the age category the fewer the number of sustainable houses, and therefore the less significant the regression results. Hence green buildings could not be compared to unrated houses older than 25 years. This is an important fact and should be taken into consideration in this thesis. The results obtained showed a significant average transaction price premium for green rated homes of 5%. This amounted on average for $45,000.

Another study from Deng et al. (2011) analyzes the residential sector in Singapore. This was one of the first studies about the implication of sustainability on the residential market sector, and the very first in the Asian market. Their hypothesis was to

(14)

find a premium in asset values for residences that obtained the Green Mark label (sustainable certification for real estate in Singapore).

The Green Mark (GM) label is similar to the LEED certification described in the previous section of this chapter. Indeed, GM involves water efficiency, heater and cooling performances and other characteristics. Also, the GM label has 4 levels of certification: certified plaques, Gold +, Gold and Platinum. Here again this implies 4 different certifications exactly like LEED.

The database was obtained from the BCA, a building registration company. The database involved 250 buildings with a Green Mark label, in which 86 were residential properties. The authors could observe 37,000 transaction prices through the period 2000-‐2010.

The methodology used was a two stage hedonic regression analyzes. The first stage involved a simple hedonic model, not controlling for location and amenities. The second stage model was obviously more significant with the inclusion of location fixed effects (parks, rivers) and amenities such as heater and cooling systems. Dummy variables were created for the amenities. Also, the authors controlled for different categories of residences, such as apartments, corner dwellings, detached or semi-‐ detached homes, for example. However, the effect of the Green mark was not studied for each house category but for the residential market in a whole.

The results found by the authors indicated a 21% transaction price premium for platinum, 15% for both Gold + and Gold and finally 10% for certified plaques. Thus, GM rated residential properties yielded on average a 15% premium.

This paper strengthens the findings of the previous article, with a much higher transaction price premium.

Finally, Brounen & Kok (2011) studied the effect of different energy consumption ratings on the Dutch residential market. Like Deng et al. this included different residential units. However, the study did not focus on green rated properties but on which energy consumption level they figure (with A the most eco-‐friendly property and G the worst). Their hypothesis was to find a higher selling price for a lower energy consumption rated property. Thus A-‐rated energy properties should sell at the highest price ceteris paribus.

The data used in this research paper involved 100,000 certified homes. The data was collected through the Dutch Ministry of Economic Affairs (Agentschap), NVM and CBS. The sample period was from January 2008 till August 2009.

(15)

15

The methodology used was an ordinary least squares regression. It involved dummies for housing type and control for location with zip codes (comparing buildings within the same 1km radius).

The results found a premium (above the OLS regression) for properties with an energy consumption rating A, B or C. One property rated D would be considered as the benchmark as it had no effect on the transaction price. Hence properties rated E and lower had a negative impact on the sales price of the dwelling. Therefore residences with an A, B or C rating would generate a price premium of 10.2%, 5.6% and 2.2%, respectively. The average premium for the sample was 3.7%, which underlines the minor number of A-‐rated properties.

As interesting fact, a property rated C in energy consumption is no longer considered “green” nowadays. This implies the technology improvements involved in new constructed green buildings and its increasing supply since this paper was wrote (2010).

More literature about sustainable buildings is available in the next section and involves the commercial real estate sector. It is important to do an in-‐depth review on the existing studies in this domain. In fact, the methodology used is similar or the same as the one that was performed for residences and that will be used in this thesis.

2.3 Green commercial real estate market literature review

The article of Reichardt et al. (2012) studies the impact of LEED and ENERGY STAR certification on rent levels and occupancy rates for commercial real estate in the United States. Their hypothesis is to find a higher rent level on green certified commercial properties, and further, investigate whether green buildings are more attractive for tenants, and thus have a higher occupancy rate. The authors begin by defining what are the advantages for a building to be certified green, and what differentiates Energy Star and LEED.

The study used the largest database for commercial real estate in the United States, the Costar database that had 2.8 million buildings. Certification for LEED and Energy Star is mostly available in the 10 largest cities of the U.S. To achieve comparability the authors controlled for submarkets, location and characteristics of the buildings. The final sample data resulted in 7,140 buildings. From this amount, 1,584 are certified ENERGY STAR and 337 LEED. Rent levels and vacancy rates were collected quarterly from 2001:Q1 to 2009:Q4.

(16)

The methodology used is a difference-‐in-‐differences estimator (DiD). It allows comparing certified and controlled buildings in the same submarket in terms of outcome changes over time, relatively to the pre-‐certification period. Hence, two observations are used for both building and control group: One before certification and one after certification. Thus the study allows for 4 different groups and gives a DID approach. The panel data regression was then performed including age of the building, rentable size, lot size, time since last refurbishment, control variable for building class, submarkets and an error term.

The hedonic model was performed for each group. The results found were a significant rent premium for both labels. The Energy Star obtained a rent premium of 2.5% and LEED 2.9%. However this rent premium fluctuates over time and has reached 7% for LEED before crisis time. This gap between “green” rent and standard rent was increasing before 2008. Lastly, the authors found a higher occupancy rate for Energy Star label.

This article strengthens the research question of this thesis. Moreover, additional articles found similar results and have a solid hypothesis that green properties are more valuable (at least in the commercial real estate sector). Furthermore, it is important to look at different methodologies used and contrast them with the findings.

Fuerst & McAllister (2011) also analyzed the effect of green certification in the United States using the Costar database just like Reichardt et al. (2012). The goal was also to find out if the rent level would differ with a green label, but moreover if its price were significantly higher.

The authors first used microeconomic theory to state their hypothesis. They found while shifting demand and supply that the green buildings hypothetically have a higher rent that is driven by the willingness to pay from tenants. This last variable is influenced from the benefits of certified buildings discussed previously in the article of Reichardt et al. (2012). The authors argue (but do not test) that there are likely to be three main drivers of positive price differences between certified (i.e. LEED and Energy Star) and non-‐certified buildings: Additional occupiers benefits, lower holding costs (due to lower vacancy rate) and lower risk premium (due to lower regulatory risk and income risk). The data used by the authors selected only metropolitan cities that have a consequent number of green buildings, which allowed for a large treatment group. In their data, there are 326 LEED and 1027 ENERGY STAR properties. The number of metropolitan cities is 60 and non-‐certified properties have been randomly selected in

(17)

17

those same cities. Finally, the sample data used contains 110 LEED certified and 433 ENERGY STAR.

The methodology used was a hedonic regression. The authors performed two regressions: One hedonic rent regression model and one hedonic transaction price regression model, both dependent variable expressed in logarithm. The regression included different variables to control for characteristics and location, similarly from the ones used by Reichardt et al. (2012). On the other hand, the effect of green certification was expressed as dummy variables, taking the value of 1 if certified and 0 otherwise. In the rent model there was simply a “green” dummy and in the transaction price model both LEED and ENERGY STAR dummies.

Fuerst & McAllister (2009) found a rent premium of 5% in average and a price premium of 20% in average for green office buildings in the U.S. The research confirms a significant higher value in green buildings. Furthermore, Fuerst & McAllister discussed the lower risk engaged when constructing green. Indeed, it appears that sustainable buildings have a lower market risk as it represents the future of real estate development. Hence, there is a lower risk premium for green buildings, therefore a lower discount rate and thus a lower capitalization rate, which leads to a higher value. Finally, the article’s finding’s observes lower holding costs for investors and additional occupier benefits for offices. Holding costs for investors will be important as it involves one of the main issues about building green: higher costs. Indeed, according to the author there are two main additional costs involved: the price for being certified green by experts and production costs to meet the standards. However, Kats (2003), Hershfield (2005), and Berry (2007) found a cost premium of 2% only for green buildings, which could be for the very lowest requirements to be green certified. Miller (2008) found the costs for each different LEED levels of certification (see figure 2.1).

Figure 2.5: Extra costs for each LEED sub-‐certifications

11 Exhibit 8: Extra Costs to Become LEED Certified as of 2007 Excluding Certification Fees

0 1 2 3 4 5 6 7

Certified Silver Gold Platinum

Extra Costs in Percentage to Build Green

(18)

The results found by the authors are interesting due to the fact that tenants are willing to pay a 5% rent premium. Moreover buyers are ready to bid about 20% more than a comparable non-‐green dwelling. The gap between price and rental premium expresses the profitable investment that sustainable buildings could represent. Hence, there should be not only a significant rent premium but also a return premium. However this requires costs information and this is hard to obtain.

The article of Eichholtz et al. (2010) studies both transaction and rent premiums of green buildings, but also the effective rent premium. This reflects whether green buildings have a smaller vacancy rate compared to unrated buildings. The authors had the main objective to find out a significant rent premium between green buildings and standard buildings.

The authors used the Costar database, with a period going from 2007 and 2009, using panel data. The methodology used was a panel data, using quarterly observations. The authors controlled for the same factors and including characteristics similarly to Reichardt et al. (2012). Green commercial buildings are the treatment group and nearby standard buildings is the control group. Moreover, they included micro-‐ location control for only comparing properties that are in a radius of 0.2 mile. This led to almost 700 dummy variables, one for each cluster.

The findings revealed a rent premium of 5.2%, 3.3% and 3.5% respectively for LEED, ES and green rating. The effective rent premium appeared to be much higher: 9.4% for LEED, 10% for ES and 10% for green rated. These findings strengthen the hypothesis of this paper whether eco-‐friendly buildings have less vacancy compared to non-‐green buildings. Finally, sales prime premiums were 11.3%, 19.1% and 16.8% respectively for LEED, ES and green rated.

Contrary to the previous articles, ES here generates a higher sales premium than LEED. Moreover, the effective rent premium is also higher for ES than LEED, which indicates a higher vacancy for this later. Therefore the higher costs involved for LEED properties forces the investors to set up higher ask rents, but tenants are not willing to rent at this high price and prefer ES rated buildings. Finally, ES have a higher demand while considering the selling process.

Miller et al. (2008) also discussed the real benefits of green buildings. The authors researched whether the technology involved reducing costs gives rise to “green value or green noise”. Their hypothesis was to find a transaction price premium for either LEED

(19)

19

or ES certified commercial real estate. This study was one of the very first to argue if green certified buildings provide a rent premium.

The Costar Group remains the database used in this research. In 2008, the data was containing 1,200 ES certified commercial buildings, whose ¾ were office buildings and the rest mainly retail. 580 buildings were LEED certified, but the data didn’t allow enough observations for each LEED certification (certified, silver, gold and platinum). Therefore, all the different LEED were put into one single variable LEED.

The methodology used was a hedonic regression. The dependent variable was the price per square foot and the independent variables were: Age, Size and dummy variables taking the value of 1 if whether LEED, ES certified, located in a city and 0 otherwise. The authors used different filters in order to control for comparability. Those were using only class A buildings, 200,000m2_{or
more,
5
stories
or
more,
built
since}

1970 and finally multi-‐tenanted buildings only.

The results showed a rent premium of $2.50 per square foot for ES and $14.15 for LEED. The transaction price premium found was 9.3% for LEED and 5.3% for ES. The occupancy rate premium was 3.7% for ES and 4.2% for LEED.

The authors prove their hypothesis and that green do pay off. Moreover, occupancy rates were higher for green buildings, which shows the interests from tenants. On top of it, the authors found a higher absorption and that the high values broadly exceeds the marginal costs for going green. Due to the higher construction quality of sustainable buildings, it is argued that their obsolescence is greatly reduced, and thus begets lower maintenance costs and depreciation (Miller et al., 2008).

Profit and risk are two major elements of a business. In real estate and especially the housing market, comfort and well-‐being are non-‐negligible. Maliene and Malys (2009) discusses the housing market in the U.K. and the day-‐to-‐day real benefits of what green housing should provide to human beings: “Sustainable housing should be well available, high-‐quality, economical, ecological, aesthetical design, comfortable and cozy one, which would better suit the needs of a person”. The authors mention that the government should drive the well fare of the housing market. Osmania and Davis (2013) ensue this idea. They state that there are not enough regulations towards architects to enforce them to design greener. Moreover, governments should encourage refurbishments for greener homes by VAT deductions (Osmania and Davis, 2013).

To summarize, the existing findings concerning green label were studied on the commercial real estate mainly and fewer on the residential market. The findings

(20)

discussed above clearly show a rent premium and a price premium for green-‐labeled commercial real estate buildings. Moreover, some authors (Reichardt et al. (2012) & Miller et al. (2008)) discovered an occupancy rate premium generated by eco-‐friendly certifications, and thus a higher source of revenue for investors. In this chapter, authors always found out a higher rent premium for LEED than ES for the commercial real estate sector, which will also be studied in this thesis. This is also the case for the transaction price premium, excepted for the article of Eichholtz et al (2012). Additional researches were made in the commercial real estate market; showing a lower risk premium for sustainable commercial real estate (Fuerst & McAllister, 2011), lower energy cost and lower obsolescence. On the other hand, higher costs are implied for going green, and Fuerst & McAllister have studied the extra costs for each different LEED labels.

The residential sector only revealed a transaction price premium, and not on LEED and ES distinctly and neither on a particular residential type. The methodologies and researches in this literature review were essential for creating the most accurate model to reveal what is the causal effect of a sustainable certification on multifamily homes in the U.S. The results from all articles are summarized in figure 2.2 below.

Figure 2.6: Summary of all the findings for ES,

LEED and Green labels

Energy star, LEED,

and

Green

benefits

Rent

premium

Price

premium

Occupancy premium

Authors

Residential sector

Kahn & Kok (2013)

5.0%

Deng et al. (2011)

15.0%

Brounen & Kok (2011)

3.7%

Commercial sector

Reichardt et al. (2012)

2.5%,

2.9%

4.5%

Fuerst & McAllister (2011)

5.0%

20.0%

Eichholtz et al. (2010)

3.3%,

5.2%

19.1%,

11.3%

Miller et al. (2008)

5.3%,

9.5%

3.7%,

4.2%

The “Green” label mentioned above is in general composed of a mix of LEED and ES labels used by authors that did not have enough of both for the research. This doesn’t apply to Deng et al. (2011), who studied the Green Mark label used in Asia, and for

(21)

21

Studying the impact of a green label on multifamily homes is very rare and will be performed with a hedonic regression analysis, like most of the authors did in this literature review. This thesis will argue whether LEED and ES generate a rent (asking rent) and occupancy premium, and how both labels differ.

After reviewing several methodologies from different authors, the next chapter will present the hypotheses and the methodology used in this paper.

(22)

3. Hypotheses & methodology

This chapter will set the different hypotheses to be verified and will describe which methodology will be used and how it will be performed in order to optimize the credibility of the findings. As mentioned before, while talking about rent levels in this thesis will only refer to asking rents.

3.1 Hypotheses

The main hypothesis of this paper is to find a rent premium for green (either LEED or ES) multifamily homes in the U.S. A green multifamily building should be a source of higher revenue compared to a similar non-‐green multifamily building. Hence a rent premium is expected to be found for sustainable multifamily homes.

H1: Green multifamily buildings generate a rent premium

Moreover, the difference between LEED and ES will be analyzed specifically. The second hypothesis will verify whether this is also the case for the multifamily sector.

H2: LEED has a higher rent premium than Energy Star

Furthermore, LEED has 4 levels of certifications, which can be acquired through “points”. The more points the higher the rating. The first (lowest) rating is simply called LEED certified. Thereafter, additional points permits to be silver, gold and finally platinum (see figure 2.2). Every time a building has a higher LEED certification, this represents extras costs for the investors in order to fit the requirements of the upper certification level. Therefore a higher rent should be observed for every label upgrade, as investors would then not take the risk to build more expensive if they don’t expect a higher rent level for it. This hypothesis has not been examined in the different literatures as LEED buildings are more rare and its development increased very recently. Some authors, like Miller et al. (2008), simply regrouped the four different certification levels. Thus, the hypothesis here is to find a growing rent premium through the different LEED certifications (from “certified” to “platinum”).

What is the effect of LEED and Energy Star certifications on rents and occupancy rates of multifamily homes? : a spatial regression analysis focusing on the U.S. market

Amsterdam Business School

Master of Science in Business Economics

Dual degree in Real estate finance and Finance

Master’s Thesis

What is the effect of LEED and Energy Star certifications on

rents & occupancy rates of multifamily homes? A spatial

regression analysis focusing on the U.S. market.

By Collin Julien

(10603077)

January 2015

Supervisor: Marcel Theebe

Statement of originality

3

Acknowledgements

Abstract

5

Table of Contents

1. Introduction

7

9

2. Literature review

2.1 Energy Star and LEED certification levels: General benefits and

definitions

11

13

2.2 Green residential market literature review

15

2.3 Green commercial real estate market literature review

17

19

LEED and Green labels

Energy star, LEED,

and

Green

benefits

Rent

premium

Price

premium

Occupancy premium

Authors

Residential sector

Kahn & Kok (2013)

5.0%

Deng et al. (2011)

15.0%

Brounen & Kok (2011)

3.7%

Commercial sector

Reichardt et al. (2012)

2.5%,

2.9%

4.5%

Fuerst & McAllister (2011)

5.0%

20.0%

Eichholtz et al. (2010)

3.3%,

5.2%

19.1%,

11.3%

Miller et al. (2008)

5.3%,

9.5%

3.7%,

4.2%

21

3. Hypotheses & methodology

3.1 Hypotheses