1
The role of asset beta within the asset-light business
model
Thomas Winkler, 11893745
Bachelor Economics & Business Economics
Specialisation: Finance
Supervisor: Nancy Edwards
June 2020
Abstract
In the last decades, lodging corporations started to shift their business strategy by reducing properties under ownership to an asset-light business model. With an asset-light strategy, hotels are operating the properties through leases, franchise, and management contracts. This paper researches the behaviour of the asset beta, a proxy for business risk, in the framework of asset-light business strategies. The asset beta is compared against different levels of an asset-asset-light strategy adopted by the corporations, the degree of which is measured by the fixed-asset ratio in this work. Besides, this paper analyses factors of an asset-light and fee-oriented strategy that might be influencing underlying business risk. Using the OLS regression analyses this paper finds that there is a significant correlation between, a decrease in business risk (asset beta) and the adoption of an asset-light strategy. In conclusion, this paper provides new insights in the ongoing discourse in literature of the effectiveness of the asset-light business models.
2 Statement of Originality
This document is written by Student Thomas Winkler who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are 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
Table of Contents
1. Introduction ... 4
2. Literature Review ... 6
2.1 Classification of asset-light corporations ... 7
2.2 Performance evaluation ... 8
2.3 Opposing evidence ... 9
2.4 Advantages of the Fee-business model ... 10
2.5 C-corporations and REITs ... 11
2.6 Asset beta ... 11
3. Data and Methodology ... 12
3.1 Data ... 12
3.2 Methodology ... 13
3.3 Hypothesis and Expectations ... 16
4. Results ... 16
6. Discussion and Conclusion ... 21
References ... 23
4
1. Introduction
In the 1980th lodging firms started whittling down properties to transform their asset-heavy balance sheets to asset-light ones, this process involved shifting their focus on the more profitable and less capital intensive side of operating and franchising properties for fees (Yuan & Singal, 2019). Many hotel C-corporations started the process of shifting to the asset-light strategy in the last decades with the goal to reduce the risk of holding real estate and freeing capital which can then be returned to shareholders (for listed companies) and improve return on capital employed (Page, 2007).
Research on the asset-light fee-oriented strategy has shown that decreasing the fixed-asset ratio elevates firm value and mitigates operating risk (Sohn & Tang, 2013). On the other hand, there has been research done which determined that the required return on intangible fixed assets (an increasing factor when adopting the asset-light strategy) is significantly higher than the levered cost of equity. Overall, there is still discordance in the literature if the decision of shifting to an asset-light fee-based model is the right one in terms of risk and reward (Low & Prashant, 2015).
A blind spot so far was how the underlying business risk, a proxy, therefore, is the asset beta is affected when firms are changing to an asset-light firm strategy. This paper will pursue this unknown aspect and therefore this yields the following research question.
Are there differences in asset betas between firms in the hospitality industry that differ chiefly in the degree to which they have adopted the asset-light business model?
This paper will approach this question by analysing hospitality C-corporations listed on the NYSE or NASDAQ in terms of to what degree they have adopted the asset-light business model and then use this gained information to find out how the asset beta changes. The underlying concept of the asset beta is to isolate the risk which is solely due to the company’s assets by removing the financial effects of leverage and capital structure for a firm (Lesseig & Payne, 2017). Besides, this paper does not only look at the differences of the asset betas but also what are the factors influencing it when hospitality C-corporations have adopted the asset-light strategy. This brings us to the sub-question of this research paper?
To what extent are factors of the asset-light business model explaining a change in the asset beta of hospitality C-corporations?
5
This research will look at factors of asset heaviness, fee revenue weight and return on assets to figure what factors of the asset-light strategy are influencing underlying business risk and therefore the paper can offer companies some research on how they can influence their asset beta. The research of this paper was carried out by using 19 publicly listed C-corporations of the hotel sector available on the FactSet Database. The paper collected data in a period from 2000-2019, however, the analysis does not always include the full-time span because the companies got listed at different times and differed when and if they adopted the asset-light approach.
The hypothesis for this research is, “H0: There is no significant correlation between the
asset beta and the degree of which the asset-light business model is adopted”. The research
first ordered the investigated lodging firms on their asset heaviness, unlevered betas were calculated and were compared in a table to conclude if asset betas and the level of the adoption of the asset-light business model show some correlation.
The sub-question is carried out by applying an OLS regression against the average asset beta returns of the sample to determine the factors influencing it significantly. The hypothesis tested is, “H0: βi= 0”. Where 𝛽𝑖 represents the coefficient of the independent variable impacting
the asset beta. The independent and control variables tested against the asset beta are the fixed asset ratio, fee-income ratio, return on assets and the degree of operating leverage.
The structure of this paper begins with an analysis of literature supporting this study. The literature is ranging from the advantages and disadvantages of the asset-light business model to investigations into the assumptions and models used in this work. This is continued by the explanation of the data, and why it was chosen. The research methods are described in the methodology section followed by the outline of the hypotheses. In the next section the results will be discussed answering if there are differences in asset betas between firms in the hospitality industry that differ chiefly in the degree to which they have adopted the asset-light business model, will be made. Lastly, a conclusion will be reached on what extent factors of the asset-light business model are explaining change in the asset beta of hospitality C-corporation.
6
2. Literature Review
Opening and running a hotel has always been associated with large capital investments for land and for real estate. However, since the end of the 20th century, there has been a growing trend for lodging firms to move from owning most of their real estate to an asset-light business model or better known to an asset-light fee-oriented strategy (ALFO). An asset-light business model indicates a business model where the focus lies on managing and franchising hotels rather than controlling and owning real estate as well as selling and leasing back properties (Seo & Soh, 2019). The trend has been driven by the factors that hotels have recognized as an asset class for property investor, an increase in prices for traditional commercial property investments, an opportunity to grow without the need of large capital investments and an increasing demand of shareholders to increase the capital returned to shareholders and to improve return on capital employed (Page, 2007).
The transition to an asset-light business model includes selling properties and investing capital in technology and loyalty-based assets. Besides not having to undergo large capital investments, Yuan & Singal (2019) argue that operational risk is reduced substantially as well. Operational leverage and earnings volatility are lower because asset-light translates into a lower fixed-asset ratio that decreases operating leverage. Furthermore, Yuan & Singal (2019) argue that the asset-light business model also reduces systematic risk. Traditionally lodging firms have had high fixed costs due to capital investments in property, as well as from depreciation and maintenance of fixed assets. As a result of the inflexibility of fixed costs, firms with a higher fixed-asset ratio are subject to higher earnings volatility, especially in times of uncertain economic conditions.
Franchising was a revolution to business expansion as companies could grow at a faster rate than before and did not need to invest their capital. Moreover, because long term contracts with the franchisee lead to more stable earnings stability and motivated franchisees who want to reach desired returns (Combs, Ketchen, & Short, 2011).
Franchise contracts consist of two main fees paid by the franchise, an upfront initial fee which is due at the beginning of the contract and continuing fees, mostly based on sales revenue. Similarly, management contracts consist of two components, a base management fee which is dependent on the revenue gained by that property and when the revenue of that property exceeds a predetermined target the operators also receive an incentive fee on top. Often management contract agreements operate in a way that a hotel corporation sells a property to a buyer and
7
signs a long-term contract with that buyer to continue to operate the hotel in exchange for the above-mentioned fees (Deroos, 2010).
Even though scaling with franchising and management contracts does not make hospitality firms immune from economic conditions, but it provides more stable earnings in periods of economic downturn and disproportionately higher earnings in periods of economic expansion. Moreover, a management agreement with that associated income stream is an asset, whereas a lease and its obligation to pay rent, is a liability. Therefore, the management agreement can have some tax and accounting advantages (Seo & Soh, 2019).
Leases on the other hand, lost popularity over time due to management contracts and franchising, especially when some larger hotels could not meet their rent payments in the economic downturn of 9/11 and the financial crisis in 2008 (Page, 2007). Nevertheless, some hotel operators still prefer the sell and leaseback method and if it is structured to avoid risk, it has its advantages. The main risk with a lease is that the operator will not be able to pay rent. Two ways of mitigating that risk would be to set rents at a manageable rate or to use turnover rents that reduce if business drops. The other advantages of a lease over a management contract are that the operator can retain all the operational turnovers and profits instead of only getting a small share of it. Secondly, leasing gives an operator much more freedom and control to run his operations in a way that he thinks is best without having conflicting interests with the property owner (Page, 2007).
2.1 Classification of asset-light corporations
To compare hospitality firms with different degrees of how they have adopted an asset-light strategy different ratios will be looked at to get an idea to what extent the asset-asset-light model is adapted. Two ratios to look at are the capital intensity ratio and the fixed asset ratio which help to figure out how asset-heavy the firms are. The capital intensity ratio measures capital expenditures as a percentage of total assets and the fixed-asset ratio, also known as asset tangibility, is the ratio of the firm’s new property, plant, and equipment relative to its assets. While capital intensity reflects a firm’s spending on fixed assets in the time of a year, the fixed asset ratio shows the increasing effect of the adoption of an asset-light strategy on the stock of tangible assets (Sohn & Tang, 2013).
Fee-business orientation is another component of the asset-light strategy. This is measured by the fee-income ratio, and an increase in this ratio implies that firms are generating
8
more income from management contracts and franchising than from operating their properties. Furthermore, the degree of franchise measure (DOF), which is closely related to the fee-income ratio, shows the proportion of a firm’s total units operated under management contracts and franchising (Combs, Ketchen, & Short, 2011).
Firms which have adopted the ALFO should both have had an increase in their fee-income ratio and DOF. Simultaneously the capital intensity ratio and the asset tangibility must have decreased due to the asset-light strategy (Sohn & Tang, 2013). Prior research by Sohn & Tang (2013) indicated that the asset-light strategy started to spread rapidly from 2002 onwards with a 75% increase in the industry-wide fee-income ratio till 2010 (7,5% to 13,3%). The research of this paper will implement the fixed-asset ratio to classify the various degrees of asset-light corporations used in the data set and the income ratio as a factor of how fee-orientation affects the asset beta of a company.
2.2 Performance evaluation
In 2014 InterContinental Hotel group’s annual report (2014) has shown that around 85% of their portfolio was franchised and over 90% of their annual operating profit came from franchising or management contracts. The main advantages of the asset-light approach were higher returns and reduced volatility stated by InterContinental Hotel group’s annual report (2014). To figure out if the asset-light strategy is superior, Low & Prashant (2015) further examined if the risk-adjusted return is higher when hotels follow an asset-light business model or an asset-heavy business model. To perform this analysis Low & Prashant (2015) made use of the modern portfolio theory by Markowitz. Markowitz (1952) showed that risk-averse investors are seeking to maximise their return for a given level of risk.
The Modern portfolio theory approach considers three main components on how to allocate capital within different asset classes: 1. It integrates risk and return considerations simultaneously: 2. it quantifies the investment decision making process and lastly 3. It combines their contributions at the portfolio level taking into consideration and investors overall wealth (Elton & Gruber, 1987). Peterson et.al. (2004) found in his research that by taking risk-adjusted performance measures into account, the hotel sub-sector is the strongest performer compared to other real estate sub-sector in terms of total return and it produces the second-best performance of risk-adjusted returns (the period taking into account was from 1992-2001).
9
However, research by Corgel & deRoos (1997) has shown a different picture. They concluded that hotels are more volatile than other real estate asset classes, therefore are one of the riskiest commercial real estate classes as well as that the hotel industry exhibits higher risk and leverage compared to other industries within the US economy. With the research done so far, there is no consensus reached on the risk and reward trade-off of the hotel sector in a wider portfolio context and it requires further investigation into the industry to conclude.
2.3 Opposing evidence
Along with benefits, there are also some criticisms for the asset-light business model. One of the main difficulties pursuing the asset-light model is to align the interest of the property owner with the operator interests. The division between hotel ownership and hotel operations in management contracts indicates that the involved parties might have some conflict of interest when it comes to investment choices but also ongoing management decisions (Gannon, Roper, & Doherty, 2010).
Additionally, when hotel chains expand their business abroad owners and operators typically originate from different cultural and institutional backgrounds, for hotel chains, this is an advantage to gain access to the local market through the host and the host acquires expertise and the name and expertise from a global hotel brand. Giving away control over the property makes the hospitality firms dependent on an outstanding party, even though it has several advantages, it is extremely important to find a trustworthy partner and to set clear guidelines and rules to avoid conflict of interest or inflexibility to act in certain situations (Gannon, Roper, & Doherty, 2010).
After the global financial crisis in 2008, the real estate prices dropped dramatically. This inclined many hotel companies to go asset-light out of nervousness from the falling real estate values and not because they believed that the asset-light model will be the right choice to grow their business (Low & Prashant, 2015). Furthermore, Schauten, Stegink and Graff (2010) have researched intangible assets of firms and their required return. In their paper, they concluded that intangible assets carry a greater risk than the company. One of their other findings has shown that the required return of intangible assets is significantly higher than the levered cost of capital.
10
2.4 Advantages of the Fee-business model
Even though the literature of Schauten, Stegink and Graff (2010) has shown that intangible assets have a higher risk then the overall business risk and the required return is higher than the levered cost of capital. Intangible assets such as brand name, management expertise and established distribution channels along with a network of operating contracts added a lot of value for firms in the lodging industry. These intangible assets are accumulated over a long period and cannot be copied easily by competitors. Resource-based literature has shown that firms who can draw and share common resources had better performances in the long run. Especially because management contracts and franchising businesses were able to transfer core competencies (for example property management skills or human resources) across properties globally (Sohn & Tang, 2013).
An additional positive factor that can be considered is that fee businesses have stable earnings. This is due to low volatility of operating cash flows from franchising firms which is the case because of a low variance of royalties, received from the franchisee, in comparison to the variance of revenue and profit from the hotel owned operations. Furthermore, low volatility in earnings leads to lower financial costs such as underinvestment, financial distress, taxes and eventually even to higher firm value (Roh, 2002).
Another value-creating factor of the asset-light strategy is the effect it has on the asset structure and risk exposure of hospitality firms. This strategy not only provides additional liquidity for the company to finance profitable projects but also reduces firm-specific risk by stabilizing cash flows and diversification. It also eliminates the need for high capital expenditure while increasing a firm’s operation areas with low levels of opportunity costs (Sohn, Tang, & Jang, 2014).
Not having to worry about financing real estate acquisitions increases operation, which is highly dependent on the firm’s respective cost of capital, increases operational flexibility for firms. Since real estate purchases are generally founded by raising equity or issuing new debt, the acquisition may stimulate financial distress costs. Along with the issues of a high sum investment, there are multiple disadvantages of owning real estate such as high transaction costs, low liquidity, high irreversibility and low depreciation of real estate (Petersen, Singh, & Sheel, 2004).
11
2.5 C-corporations and REITs
The hospitality industry generally consists of two organization types REITs and C-corporations (Hanson, 1997). C-C-corporations generate their profits by selling products and services directly to their customers by operating leased, owned, and contracted properties. Since the implementation of the asset-light business model, the ownership of properties by C-corporations has decreased significantly. A REIT, on the other hand, is an entity that mainly purchases real estate assets and rents or leases them out to third parties. REITs do not provide products or services directly to customers as it is bound by REIT specific regulations. Thus, there is a clear difference in organizational structure between REITs and lodging C-corporations (Kim, Jackson, & Zhong, 2011).
However, because REITs are asset-heavy by design and heavily rely on debt and equity offering and this paper is focusing on asset-light firms and the unlevered beta, they will not be taken into consideration for further research of this paper and the focus will be on C-corporations and their asset-light approach. Nowadays C-C-corporations are mostly focusing on the sale of their core intangible assets, therefore trying to keep asset risk as low as possible as well as improving risk-adjusted profitability. Furthermore, C-corporations have a very flexible organization structure as they are under no obligation to distribute any of their taxable parts to shareholders as dividends (Kim, Noh, & Lee, 2019).
2.6 Asset beta
For understanding the asset beta, it is important to first know what the ‘normal’ beta (levered beta) tells us. Beta is the slope of the coefficient for a stock regressed against a benchmark market index, for example, the Dow Jones Index. The levered beta considers the capital structure of a firm including debt, equity and measures the risk of it to the volatility of the market. Therefore, leverage is a key determinant of the beta because it evaluates the level of a company’s debt and its equity (Lesseig & Payne, 2017). The risk which is measured with the beta is systematic. Systematic risk is the type of risk that cannot be diversified away, it is caused by factors that are beyond a company’s control, for example, unexpected political events, wars or natural disasters (Bali, Brown, & Caglayan, 2012).
Asset beta or also known as unlevered beta measures the market risk of a company without the impact of debt. The asset beta isolates the risk which is solely due to the company’s assets by removing the financial effects of leverage. Meaning it shows how much the
12
company’s equity contributes to its risk profile. Hamada (1972) introduced a decomposition of the market beta (equity beta) into operating (asset beta) and financial risk (financial leverage) components. His method of removing financial leverage from the market beta is still the most used technique to determine an asset beta.
Unlevered beta is used in academia as a control variable to measure operating risk in various analyses. Investors use the asset beta to gain clarity on the composition of risk assumed when comparing companies. The unlevered beta is almost always equal to or lower than the levered beta (because debt is either 0 or positive) and only in rare cases when companies’ debt is negative then the unlevered beta can be higher. A positive asset beta is an indicator to invest into the company when prices are expected to rise, and a negative unlevered beta suggests investors to short if prices are expected to decline (Lesseig & Payne, 2017).
Bernando, Chowdhry and Goyal (2007) proved in their research that a firm’s project with relatively high growth opportunities has a higher asset beta than a firm’s project with relatively low growth opportunities resulting in a 2-3% higher cost of capitals for projects with higher growth opportunities. In this paper, the results are expected to build up on their research, meaning that firms who have adopted an asset-light business strategy have lower asset betas and therefore lower cost of capital.
3. Data and Methodology
3.1 Data
The data used in this paper was selected from three main sources which includes: 1. FactSet, which was founded in 1978 and is a publicly listed company providing accurate data and analytics for the investment community throughout the world (Hicks, 2017). FactSet was used to obtain data from the financial statements of the companies included in the research. 2. Wharton Research Data Services (WRDS) is providing leading business intelligence, data analytics and research platforms to educational institutions and companies alike (Pennsylvania, 2020). From the platform WRDS, data was gathered for the equity betas and corporate tax rates, both further used to calculate the asset betas needed. 3. Financial times is an international newspaper that focuses on current business and economic affairs (Times, 2020). From their database, missing values were added to complete the dataset and to obtain further information about the company’s part of this research.
13
For the analysis in this paper, first, all the hotel stocks listed at the NYSE and NASDAQ were looked into and 30 lodging stocks are listed. After excluding all REITs and companies with a fixed-asset ratio of higher than 0,75, 19 firms were left for the research. The data collection period spanned from 2000-2019, however, the data available was not the same for all firms throughout the full period because the listings of the companies took place at different times and it also differed when companies started to report management and franchise income separately from their other revenue sources. Therefore, the time interval was adjusted to have consistent data. For tests in the period of 2015-2019 data was collected for all 19 companies and therefore 95 observations were available. For the regression to answer the second hypothesis most data were available from 2006-2019 and therefore 241 observations were used. The type of data used for this study is panel data. Panel data (also known as cross-sectional or longitudinal time-series data) is a dataset in which the behaviour of entities are observed over time (Hsiao, 2014).
3.2 Methodology
The purpose of this work is to research the change of the asset beta when the degrees of the asset-light business model is adopted in various degrees. Firstly, the asset betas for firms included in this research were calculated using equity beta data found on the WRDS database. For this step, the method by Hamada (1972) was used, his process decomposes the observed levered beta into its operating (asset beta) and financial risk (financial leverage) components. He isolates the firm’s unlevered beta by removing the effects of financial leverage from the firm’s levered equity beta with the following formula:
𝛽𝑖,𝑡 = 𝛽𝑖,𝑡
𝐸
1 + (1 − 𝜏) ∗𝐷𝐸𝑖,𝑡
𝑖,𝑡
Equation 1: Hamada Formula
Where βi,t is the firm’s unlevered beta, βi,tE is the equity beta of firm i at time t, τ is the
tax rate, and Di,t /Ei,t is the ratio of total long-term debt to total shareholders’ equity value for
firm i at time t.
To understand better to what extent the companies in the sample were asset-light, the fixed-asset ratio was calculated. The fixed asset ratio is the ratio of the amount of Net Property, plant, and equipment to total assets.
14
𝐹𝑖𝑥𝑒𝑑 − 𝑎𝑠𝑠𝑒𝑡 𝑟𝑎𝑡𝑖𝑜 =𝑁𝑒𝑡 𝑃𝑟𝑜𝑝𝑒𝑟𝑡𝑦, 𝑃𝑙𝑎𝑛𝑡 𝑎𝑛𝑑 𝐸𝑞𝑢𝑖𝑝𝑚𝑒𝑛𝑡 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠
Equation 2: Fixed-asset ratio
Furthermore, to see the implementation of the ALFO strategy, fee income ratio was used (a measure of the degree of fee business), which is the sum of fee income from franchising and management contracts against total sales revenue (Sohn & Tang, 2013).
𝐹𝑒𝑒 − 𝑖𝑛𝑐𝑜𝑚𝑒 𝑟𝑎𝑡𝑖𝑜 =𝑀𝑎𝑛𝑎𝑔𝑒𝑚𝑒𝑛𝑡 𝑓𝑒𝑒 + 𝐹𝑟𝑎𝑛𝑐ℎ𝑖𝑠𝑒 𝑓𝑒𝑒 𝑆𝑎𝑙𝑒𝑠 𝑟𝑒𝑣𝑒𝑛𝑢𝑒
Equation 3: Fee-income ratio
The expectation of this research is to see the operating risk (asset beta) to be lower for the firms with a low fixed asset ratio and higher for firms with a higher fixed-asset ratio. Regarding the fee-income ratio, it is expected that there will be an opposite behaviour, therefore it can be expected that the fixed-asset ratio and the fee-income ratio have a negative correlation. To test the strength and direction of the asset beta and the adoption of the asset-light strategy (fixed asset ratio), an analysis will be performed with the help of the Pearson correlation measure. To calculate the correlation coefficient the following formula will be used:
𝜌𝑥𝑦 = 𝑛(∑ 𝑥𝑦) − ∑ 𝑥 ∙ ∑ 𝑦
√[𝑛 ∑ 𝑥2− (∑ 𝑥)2][𝑛 ∑ 𝑦2 − (∑ 𝑦)2]
Equation 4: Correlation coefficient
A coefficient value between 0.1 and 0.3 indicates a small correlation, between 0.3 and 0.5 a moderate correlation and larger than 0.5 suggests a strong correlation. The level of significance used is 5% to either reject or accept the null hypothesis.
The next step is to understand how different factors are influencing the asset beta in the asset-light hospitality sector. For that, an OLS regression will be run to identify which variables are having an impact on the calculated asset beta. With the asset beta (ab) as the dependent variable, for independent variables, the fixed-asset ratio (far), the fee income ratio(fir), and the return on assets (roa) will be included. Return on assets is a measure of how profitable a company is relative to its total assets (Leeds, et al., 2002). Therefore, it is included in the regression to observe the effect of profitability relative to a company’s assets on the asset beta.
15
Furthermore, a control variable, the degree of operating leverage (DOL) is included in the regression because it is one of the main determinants of the asset beta (Mandelker & Rhee, 1984). Data for the ROA was available on the WRDS database and the degree of operating leverage was calculated with the following formula:
𝐷𝑒𝑔𝑟𝑒𝑒 𝑜𝑓 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑛𝑔 𝑙𝑒𝑣𝑒𝑟𝑎𝑔𝑒 (𝐷𝑂𝐿) = % 𝑐ℎ𝑎𝑛𝑔𝑒 𝑖𝑛 𝐸𝐵𝐼𝑇 % 𝑐ℎ𝑎𝑛𝑔𝑒 𝑖𝑛 𝑆𝑎𝑙𝑒𝑠
Equation 5: Degree of operating leverage
For easier interpretation, the natural logarithm for the fixed asset ratio (lnfar) was calculated. By taking the natural logarithm the semi-log functional form was used, meaning that if the fixed asset ratio changes by one per cent, asset beta will change by 𝛽1/100 units. The following regression will therefore be tested:
𝑎𝑠𝑠𝑒𝑡 𝑏𝑒𝑡𝑎(𝑎𝑏) = 𝛽0+ 𝛽1𝑙𝑛𝑓𝑎𝑟 + 𝛽2𝑓𝑖𝑟 + 𝛽3𝑟𝑜𝑎 + 𝛽5𝑑𝑜𝑙 + 𝜀
Equation 6: OLS-regression
As there is a gap of information about the structure of heteroskedasticity, robust standard errors will be used. Even if there is no heteroskedasticity, because of the large same size with 241 observations, the robust standard errors will become conventional OLS standard errors. Thus, it is even appropriate to use them under homoskedasticity (Hayes & Cai, 2007). In addition, a Variance inflation factor test is also done to check for multicollinearity. VIF values of the independent variables that are greater than 10 may merit further investigation (Mansfield & Helms, 1982). In this research the VIF results were lnfar (1,23), fir (1,2), roa (1,16) and dol (1), therefore there was no multicollinearity.
16
3.3 Hypothesis and Expectations
By combining the various analyses mentioned in the previous section, the aim is to find out how the asset beta differs in the same industry (hotel sector in this case) when lodging C-corporations have adopted various degrees of the asset-light business model. The goal is also to identify which factors are influencing the asset beta when a firm is starting to implement an asset-light business strategy. To carry out the research, this paper will test the following hypothesis:
H0: There is no significant correlation between the asset beta and the degree to which the
asset-light business model is adopted
H1: There is a significant correlation between the asset beta and the degree to which the
asset-light business model is adopted
And to assess whether there are factors of the asset-light strategy that are influencing the asset beta, the following hypothesis will be evaluated:
H0: 𝛽𝑖= 0
H1: 𝛽𝑖 ≠0
𝛽𝑖 represents the coefficient of the independent variable impacting the asset beta. This
will be tested for the factors selected in our regression. If the hypothesis is rejected at a 10% significance level, then there is evidence that there are factors that are explaining the change in asset beta.
4. Results
In the first step asset betas of the lodging firms in the sample were calculated which can be seen in the Appendix Table A. An interesting observation is that Choice Hotels group and Intercontinental Hotels group both had a negative asset beta. Their asset beta is negative due to a negative debt to equity ratio. A negative debt to equity ratio can occur due to several reasons which include when a company has interest on its debts that are greater than the return on investments when a firm takes on additional debt to cover losses instead of issuing shareholder equity or experiences financial loss in periods following large dividend payments. In general, a negative debt to equity ratio is a red flag for investors, shareholders and creditors alike (Anuar & Chin, 2016).
17
Figure 1 shows the historical trends for the average sample fixed asset ratio and the average income ratio. There is an inverse relationship between fixed-asset ratio and fee-income ratio, implying that firms who have decreased their fixed-asset ratio, have increased their fee-income ratio. This demonstrates that firms who have adopted a fee-based model within the asset-light strategy have decreased their fixed assets. This supports the approach of this work to take the fixed-asset ratio as a measurement of asset-light firms, which have or have not adopted an asset-light or fee-based strategy.
Figure 1: Historical trends
Table A in the Appendix reports the average of the yearly asset betas and the average of the yearly fixed-asset ratios of the 19 C-corporations in this sample, for the time between 2015-2019. Figure 2 reports the mean asset beta as well as the average asset beta for firms within the 25th percentile of fixed-asset ratio (Q1) and the average asset beta for firms in the 75th percentile
fixed asset ratio (Q3) for the sample in research. For the time period 2004-2019, the same empirical analysis was done, however for this time period the data was only available for10 corporations and 9 were included in the sample as Choice hotels had a high negative average asset beta of -8,07 due to high negative shareholders equity in the years 2004-2009 and thus, it was excluded from the sample for this observation. The average values are reported in Table B in the Appendix and the results in Figure 3.
18
In both the periods and both sample sizes firms with a below-average fixed-asset ratio have on average a lower asset beta than firms with an above-average fixed-asset ratio. The data strongly supports our hypothesis that there is a correlation between the extent to which lodging C-corporations have adopted an asset-light strategy and the asset beta. The difference between both samples tested is relatively small, the mean asset beta for the periods 2015-2019 and 2004-2019 are 0,37 and 0,38 respectively. However, there is a large difference between a firm at the 25th percentile in fixed asset ratio with an asset beta of -0,09 while a firm at the 75th percentile in fixed asset ratio has an asset beta of 0,70. This difference can represent a significantly higher cost of capital for a firm’s project with a high fixed asset level relative to a firm’s project with a low fixed asset level (using the Capital Asset Pricing Model for cost of capital calculations). These findings are consistent with the empirical results by Bernando, Chowdhry and Goyal (2007) but in disagreement with the results in the paper of Schauten, Stegink and Graff (2010) who proved that in the hospitality sector the required return on intangible assets is significantly higher than the unlevered cost of equity. This is inconsistent with the research done in this paper as a higher number of intangible assets (mainly franchise and management contracts in the hospitality sector) implies a lower fixed-asset ratio. A reason for this difference in findings would be that the firms in the sample this paper used have all not fully adopted an asset-light fee oriented approach (only have intangible assets) and therefore they still own, have leased and operate only some properties through franchise and management contracts and only the latter is classified as an intangible asset.
Figure 2: Asset beta 2015-2019
Figure 3: Asset beta 2004-2019
Asset betas Q1 Mean Q3 Sample of 19 C-corporations -0,09 0,37 0,7 2015-2019 Asset betas Q1 Mean Q3 Sample of 9 C-corporations* -0,42 0,38 0,75 *adjusted for outliers (we excluded choice hotels)
19
Table 1 reports pairwise correlation coefficients. Asset beta and the fixed asset ratio are positively correlated, with a correlation coefficient of 0,352 and a significance at ∝= 0.01, indicated by the triple * sign. Therefore, the null hypothesis: “H0: There is no significant
correlation between the asset beta and the degree of which the asset-light business model is adopted” is rejected and it can be concluded that there is a moderate positive correlation between the asset beta and the degree of which the asset-light business model is adopted in the hospitality sector.
Table 1: Pairwise correlation
Furthermore, there is a moderate negative correlation between the fee-income ratio and fixed asset ratio, which is also significant at∝= 0.01. This confirms the opposite movement evidence of the fixed asset ratio and fee income ratio in Figure 1. A significant weak negative correlation exists among the asset beta and the fee-income ratio.
To test the identifying factors of the asset beta for the lodging C-corporations the next step was to run the regression:
𝑎𝑠𝑠𝑒𝑡 𝑏𝑒𝑡𝑎(𝑎𝑏) = 𝛽0+ 𝛽1𝑙𝑛𝑓𝑎𝑟 + 𝛽2𝑓𝑖𝑟 + 𝛽3𝑟𝑜𝑎 + 𝛽5𝑑𝑜𝑙 + 𝜀
The regression was applied to the dataset of all 19 C-corporations from years 2006-2019 where data was readily available. After adjusting the dataset for outliers at the 1st and 99th
percentile, 241 observations were tested. This resulted in the following regression output in Table 2. Observation from the output is that the independent variables natural logarithm of the fixed-asset ratio(lnfar), return on assets and the degree of operating leverage are statistically significant at an alpha level of 10%. The return on assets ratio even shows significant results at an alpha level of 1%.
Variables ab far fir
ab 1.000
far 0.325*** 1.000
fir -0.210*** -0.420*** 1.000
20 Table 2: Regression output Factors influencing Asset beta
ab Coef. Robust St. Err. t-value p-value [95% CI] Sig lnfar 0.157 0.085 1.85 0.066 -0.010 0.325 * fir -0.503 0.917 -0.55 0.583 -2.309 1.302 roa -0.105 0.034 -3.10 0.002 -0.171 -0.038 *** dol -0.002 0.001 -1.66 0.098 -0.004 0.000 * Constant 0.823 0.092 8.92 0.000 0.641 1.005 ***
Mean dependent var 0.169 SD dependent var 1.754
R-squared 0.263 Number of obs 241.000
F-test 6.647 Prob > F 0.000
*** p<0.01, ** p<0.05, * p<0.1
With the results in Table 2, the null hypothesis (H0: βi= 0), where 𝛽𝑖 represents the
coefficient of the independent variable impacting the asset beta, can be rejected in the following cases: The natural logarithm of the fixed-asset ratio(lnfar) has a coefficient value of 0,157 and a p-value of 0,066, therefore with a significance level of 10% the null hypothesis can be rejected and it can be concluded that there is a positive relation between the fixed asset ratio and the asset beta for lodging C-corporations which means that if the fixed asset ratio changes by one per cent the asset beta will change by 0,00157 units.
Furthermore, the null hypothesis can also be rejected for the independent variables, return on assets and degree of operating leverage with coefficient values of -0,105 and -0,002 respectively. In both cases, a significant small negative relation can be observed. This implies that firms with higher profitability on their assets have a lower business risk. In addition, the small negative relation with the control variable (DOL) is consistent with the literature of Mandelker and Rhee (1984).
The only insignificant explanatory variable in the regression is the fee-income ratio with a coefficient of -0,503 and a p-value of 0,583. Therefore, the null hypothesis of “H0:
βfee−income ratio= 0” cannot be rejected. However, the inverse relation between the fee-income ratio and the asset beta would be in line with the previous research done in this paper. Therefore, a reason for the insignificant results for the independent variable “fir” could be attributed to a lack of accuracy in the data gathered for the fee-income ratio. Many of the
21
companies in the sample used differed heavily in the way they reported revenue from franchise and management operations. Furthermore, the time they started reporting those revenues separately varied as well. Those limitations could be a reason for the high p-value and the resulting conclusion. For further research, a more accurate database for revenue from fee-based operations would maybe produce different results.
6. Discussion and Conclusion
The hotel industry has always been regarded as having a high level of business risk in comparison to other industries. Performance is tied closely to the overall state of the economy and a great deal of capital is tied up to fixed assets and therefore leaves hotel firms with less capital flexibility to market changes (Sohn, Tang, & Jang, 2014). Due to these risks there was a turnabout in policies like the decrease of fixed assets while expanding management or franchising business. This restructuring became to be known as the asset-light business model or asset-light fee-oriented strategy.
Within the asset-light model there has been numerous research about whether adopting an asset-light-fee oriented strategy is the best option for lodging firms, if the sale and leaseback model is the right choice to go or if owning the underlying property is the most profitable option. Thus, to create some more clarity in the discourse this paper focused on the behaviour and the factors influencing the underlying business risk (asset beta) for hotel C-corporations. Therefore, by better understanding how the asset beta differs within the asset-business model and the factors which are influencing it, creating value for researchers and investors alike.
The results indicate that reducing the fixed-assets ratio decreases the asset beta on average throughout the sample used. This implies that a lower amount of fixed assets mitigates operational risk. Thus, suggesting that the advantage of being more flexible and leaner with its capital is considered a positive result for firm operators and investors. The findings of this work are aligned with previous research of Sohn & Tang (2013) and it is also consistent with the main motivations of sale and leaseback transactions (Page, 2007).
In the regression output for the factors influencing the asset beta, two different indicator ratios for an asset-light strategy were tested. First the fixed asset ratio over time which helps to conclude if the firms are transitioning to an asset-light strategy or not. However, nothing can be said about what kind of actions were adopted by the firm, either sale and leaseback or franchise and management contracts. Thus, the second ratio included was the fee-income ratio
22
which analyses that to what degree the sales are made up from fee business. Although both are indicators for an asset-light strategy but only the fixed asset ratio had a significant relation with the asset beta. It would be interesting for further research to find out if that implies that only the sale and leaseback model reduces the underlying business risk and not the fee-oriented strategy of franchise and management contracts.
The low R2 value of 0,263 is an indicator of some relevant missing explanatory variables (Rights & Sterba, 2018). Additional research is required to work out which other independent variables have a large effect on a company’s asset beta. Further limitations to this work are that only companies listed at NYSE and NASDAQ were used. Another drawback is associated with the limited amount of data available for the revenue from franchise and management contracts. More data, a larger sample size and a longer period would help to get more significant and clearer results.
The findings of this study provide some meaningful implications for managers in the hotel industry. By knowing the behaviour of the underlying business risk, managers can take this into account when making decisions to implement or further extend an asset-light strategy. Furthermore, knowing how the asset beta changes when selling or buying properties, or entering franchise or management agreements, it can help firms to better predict the cost of capital.
This paper does not suggest that firms should simply sell all their properties to reduce operational risk. There are no clear answers due to the advantages fixed assets provide to corporations. Maintaining ownership of properties gives firm’s an important level of control or can be used as a source for collateral for creditors (Low & Prashant, 2015). Additionally, research from Schauten, Steignik and Graff (2010) have shown that intangible assets (franchise and management contracts) would require a higher return than the levered cost of leverage. Thus, managers should carefully consider the risk and reward of disinvestment and try to find the optimal mix between owning and operating properties.
In conclusion, the critical stance of this paper on the asset light business model by investigating the correlations and the behaviour of the asset beta has hopefully brought somehow more clarity for management and academia. Potential direction for future research, could be to find the optimal mix of operating and owning properties for the hospitality sector to keep the level of business risk at an advantageous level.
23
References
Anuar, H., & Chin, O. (2016). The Development of Debt to Equity Ratio in Capital Structure Model: A Case of Micro Franchising. Procedia Economics and Finance, 35, 274-280. Bali, T., Brown, S., & Caglayan, M. (2012). Systematic risk and the cross section of hedge
fund returns. Journal of Financial Economics, 106(1), 114-131.
Bernardo, A., Chowdhry, B., & Goyal, A. (2007). Growth Options, Beta, and the Cost of Capital. Financial Management, 36(2), 1-13. doi:https://doi.org/10.1111/j.1755-053X.2007.tb00084.x
Combs, J., Ketchen, D., & Short, J. (2011). Franchising Research: Major Milestones, New Directions, and its Future within Entrepreneurship. Entrepreneurship Theory and
Practice(35), 413–425.
Corgel, J., & deRoos, J. (1997). Hotel investments in the portfolio: are they part of the core?
Real Estate Finance, 14, 29-37.
Deroos, J. (2010). Hotel Management Contracts—Past and Present. Cornell Hospitality
Quarterly, 68-80.
Elton, E., & Gruber, M. (1987). Modern portfolio theory and investment analysis. Wiley. Gannon, J., Roper, A., & Doherty, L. (2010). The impact of hotel management contracting on
IHRM practices. International Journal of Contemporary Hospitality Management,
22(5), 638-658. doi:https://doi.org/10.1108/09596111011053783
Hamada, R. (1972). The effect of the firm’s capital structure on the systematic risk of common. The Journal of Finance, 27(2), 435-452.
Hanson, B. (1997). C-Corp returns rise, but REITs still dominate. Hotel & Motel
Management, 212(4), 71.
Hayes, A., & Cai, L. (2007). sing heteroskedasticity-consistent standard error estimators in OLS regression: An introduction and software implementation. Behavior Research
Methods, 39, 709-722.
Hicks, R. B. (2017). FactSet : The Advantage of Scored Data. Tuscaloosa,: 3rd International Workshop on Data Science for Macro.
24
Hsiao, C. (2014). Analysis of panel data. Cambridge University Press.
IHG, I. H. (2014). Annual Report and Form 20-F 2014. Buckinghamshire, UK: Addison Group.
Kim, S., Noh, S., & Lee, S. (2019). Asset-light strategy and real estate risk of lodging C−corps and REITs. International Journal of Hospitality Management, 78, 214-222. doi:https://doi.org/10.1016/j.ijhm.2018.09.004
Kim, W., Jackson, L., & Zhong, J. (2011). Performance Comparison of Lodging REITs, Hotel C-Corporations and Resorts and Casinos. Tourism Economics, 17(1), 91-106.
Leeds, P. A., B, M. K., Girish, H., C, B. V., C, S. J., & Kong, C. L. (2002). Management systems and methods for maximizing return on assets. European Patent Office. Lesseig, V., & Payne, J. (2017). The precision of asset beta estimates. International Journal
of Managerial Finance, 13(2), 213-224.
doi:https://doi.org/10.1108/IJMF-05-2016-0091+
Low, W., & Prashant, D. (2015). The role of hotels in mixed asset portfolios:. Int. J. Built
Environment and Asset Management, 273-292.
Mandelker, G., & Rhee, S. (1984). The Impact of the Degrees of Operating and Financial Leverage on Systematic Risk of Common Stock. Journal of Financial and
Quantitative Analysis, 19(1), 45-57.
Mansfield, E. R., & Helms, B. P. (1982). Detecting Multicollinearity. The American
Statistician, 36(3a), 158-160.
Markowitz, H. (1952). ‘Portfolio selection'. Journal of Finance, 7, 77-91.
Page, T. (2007). ‘ Asset-Light ’ — Managing or. Journal of Retail & Leisure Property, 6(2), 97-99.
Pennsylvania, W. U. (2020). Wharton Research Data Services. Retrieved from https://wrds-www.wharton.upenn.edu/pages/support/
Petersen, G., Singh, A., & Sheel, A. (2004). Hotel Real Estate in a Property Investment Portfolio - Analysis of Results From 1992 to 2001. The Journal of Hospitality
Financial Management, 12, 87-87.
25
Rights, J., & Sterba, S. (2018). A Framework of R-Squared Measures for Single-Level and Multilevel Regression Mixture Models. Psychological Methods, 23(2), 434-457. Roh, Y. (2002). Size, growth rate and risk sharing as the determinants of propensity to
franchise in chain restaurants. International Journal of Hospitality Management,
21(1), 43-56. doi:https://doi.org/10.1016/S0278-4319(01)00014-7
Schauten, M., Stegink, R., & Graaff, G. d. (2010). The discount rate for discounted cash flow valuations of intangible assets. Managerial Finance, 36(9), 799-2012.
Seo, K., & Soh, J. (2019). Asset-light business model: An examination of investment-cash flow sensitivities and return on invested capital. International Journal of Hospitality
Management(78), 169–178. Retrieved from https://doi.org/10.1016/j.ijhm.2018.12.003
Sohn, J., & Tang, C. (2013). Does the asset-light and fee-oriented strategy create value?
International Journal of Hospitality Management(32), 270-277. Retrieved from
https://doi.org/10.1016/j.ijhm.2012.07.004
Sohn, J., Tang, C., & Jang, S. (2014). Asymmetric impacts of the asset-light and fee-oriented strategy: The business cycle matters! International Journal Of Hospitality
Management(40), 100-108. doi:https://doi.org/10.1016/j.ijhm.2014.03.007
Times, F. (2020). Financial TImes . Retrieved from https://www.ft.com/
Yuan, L., & Singal, M. (2019). Capital structure in the hospitality industry: The role of the asset-light and. Elsevier ScienceDirect(70), 124-133. Retrieved from
26
Appendix
Appendix Table A: Period 2015-2019
Appendix Table B: Period 2004-2019
