Identification of Merger and Acquisition Waves and Their Macroeconomic Determinants in the Hospitality Industry

Abstract

This study investigated the occurrence of merger and acquisition (M&A) waves in restaurant and lodging industries. By comparing actual frequency of M&A deals with simulated randomly generated deal frequency distributions, this study proved the presence of restaurant and lodging M&A waves. Macroeconomic determinants of the waves were then identified using factor analysis to extract underlying latent factors from 16 macroeconomic variables and employing a distributed lag model to investigate the effect of the extracted macroeconomic factors on the waves. Results showed that all factors (overall activity, market value, cost of debt, and inflation) significantly affected deal frequency in the hospitality industries studied. Theoretical and practical implications of the findings are presented.

Keywords

merger and acquisition waves macroeconomic determinants restaurant industry lodging industry simulation

Merger and acquisition (M&A) provide firms with various benefits, including cost efficiency through economies of scale, increase in market share, sales growth, financial stability, and profitability improvement (Gugler, Mueller, & Weichselbaumer, 2012; Hsu & Jang, 2007; Kiymaz, 2004; Tse & Crawford-Welch, 1989), thereby ultimately increasing shareholder value (Ma, Zhang, & Chowdhury, 2011). The significance of M&A in the hospitality industry has been growing along with the substantial increase of M&A activity for firm expansion and growth. According to the Thomson One Banker database, only two M&A deals valued at $248.53 million were announced in the hospitality industry in 1980. In contrast, the frequency and total value of hospitality M&A deals increased to 265 deals with a value of $17.50 billion in 2016.

The drastic growth in hospitality M&A activity over the past three decades has undergone constant fluctuations. This uneven distribution of the aggregate level of M&A deals suggests the presence of an M&A wave, described as a remarkable increase in deal frequency and deal value during a certain period of time. M&A waves have been observed consistently at the industry, national, and global levels (Martynova & Renneboog, 2008), thus indicating that certain circumstances or conditions are likely to facilitate M&A activity. In this respect, considerable scholarly efforts, including the application of behavioral and neoclassical approaches, have been made to identify determinants of M&A waves (Gugler et al., 2012).

One approach to identifying the determinants of M&A waves is to investigate the relationship between macroeconomic conditions and aggregate M&A activity. This macroeconomic approach can reveal what and how macroeconomic conditions encourage M&A deals. Also, significant macroeconomic determinants of M&A waves can be used to identify the period of an M&A wave. According to Maksimovic, Phillips, and Yang (2013), M&A deals within the waves are value-added, while those outside the waves are less value-added or even value-destructive. This suggests that M&A deals cluster when economic conditions are favorable for business and that clustered M&A deals create more value than nonclustered ones that occur under less favorable conditions. Therefore, identifying macroeconomic determinants allows firms to maximize M&A benefits by determining the appropriateness of economic conditions and by predicting optimal timing for an M&A deal. However, significant macroeconomic conditions have been inconsistent, which makes it difficult to apply the findings of extant M&A wave studies to the hospitality industry.

Recent finance studies have highlighted the fact that general M&A waves are initiated by the simultaneous clustering of M&A deals in several industries and that the industries creating general M&A waves are different over time (Ahern & Harford, 2014). For example, the electricity industry in the 1960s and production and manufacturing industries such as oil/gas in the 1980s led general waves before the 1990s. Since then, service industries such as banking and media/telecom have been the leading ones (Yaghoubi, Yaghoubi, Locke, & Gibb, 2016). This implies that inconsistent macroeconomic determinants might be due to different leading industries that drive M&A waves. In other words, some industries are more affected by a certain macroeconomic condition than others, and macroeconomic conditions continue to change, resulting in inconsistent outcomes. Thus, it is essential for enhancing our understanding of the determinants of M&A waves to re-identify significant macroeconomic conditions at the industry level.

This study therefore empirically investigates the macroeconomic determinants of M&A waves in the hospitality industry. To this end, this study ponders two research questions: (1) Do hospitality M&A waves exist? (2) Do macroeconomic conditions drive the hospitality M&A waves? And if so, what macroeconomic conditions drive the M&A waves? To answer these questions, this study first examines if restaurant and lodging M&A waves occur by accident using simulation analysis. Then, the effect of macroeconomic conditions on the M&A waves in each industry is investigated using factor analysis and time series techniques.

Literature Review

Identification of M&A Waves

Finance scholars (e.g., Ahern & Harford, 2014; Rhodes-Kropf, Robinson, & Viswanathan, 2005) proposed and empirically proved that general M&A waves occur as a result of industry-level waves that simultaneously emerge in several industries. Accordingly, some studies (e.g., Harford, 2005; Mitchell & Mulherin, 1996) examined whether industry-level M&A waves exist and, then after confirming their existence, further identified the period of the waves. Mitchell and Mulherin (1996) studied M&A waves of 51 industries during the 1980s. The authors attempted to prove the occurrence of industry-level M&A waves by assessing whether the deal frequency between 1982 and 1989 varied across 51 industries. The chi-square test results showed that the variation of M&A activity in each industry was different from that of all industries. They also found that more than 50% of all M&A deals were clustered within the peak period of 24 consecutive months.

Harford (2005) used simulation technique for the identification of industry-level M&A waves to investigate the effect of industry circumstances on M&A waves. His study divided the study period into two 10-year periods, the 1980s and 1990s, and focused on the 24-month peak period where M&A deals were concentrated in each decade based on the finding of Mitchell and Mulherin (1996). A total of 1,000 simulated distributions of actual deal frequency in each decade were generated in a given industry. The frequencies of M&A deals in the highest 24-month period from each of 1,000 simulated draws were counted. These 1,000 highest deal frequencies formed a simulated distribution for the 24-month peak period. The actual deal frequency in the highest 24-month period of each industry was compared with the simulated distribution for the 24-month peak period. When the actual highest frequency was higher than the 95th percentile of this simulated distribution, the period was identified as an M&A wave. The author found 35 M&A waves in 28 industries, including lodging and retail industries. The average deal frequency in a 24-month period outside the M&A waves was 8.5 deals, while it was 34.3 deals within the waves.

Carow, Heron, and Saxton (2004) employed another method of identifying industry-level M&A waves to analyze the early-mover advantages in M&A deals. To investigate whether industries experienced M&A waves between 1979 and 1998, they searched for the peak year in each industry, then found the start and the last years surrounding the peak year. When a deal frequency became less than one third of the peak frequency, the year was determined as the start or the last year. The period from the start year to the peak year to the last year was defined as an M&A wave. This identification process discovered 14 M&A waves in 9 industries over two decades. This wave identification method is relatively simple to execute, but it may fail to find the waves because of a nonexistent start and/or last year; thereby providing a weak foundation on which to judge the presence of M&A waves. On the other hand, Harford’s method identifies M&A waves through statistical verification processes based on simulation. Thus, this study used Harford’s rigorous method to test for the presence of M&A waves in restaurant and lodging industries.

Theoretical Approaches to M&A Waves

Several theoretical frameworks have been adopted to explain why industry-level M&A waves occur. Neoclassical models are developed based on the assumptions that maximizing shareholders’ wealth is a top priority for managers and that capital markets are efficient so that all information is reflected in stock prices. The industry shock theory indicates that M&A waves occur in a restructuring process of an industry that is experiencing drastic changes in regulatory, economic, and technological environments (Gugler et al., 2012). In this process, M&A help reallocate industrial resources to productive and efficient firms (Martynova & Renneboog, 2008). Another neoclassical model, the q-theory, posits that industrial restructuring is an ongoing activity regardless of industrial shocks. In the market, inefficient management (low Tobin’s q) is constantly removed and replaced by more efficient management (high Tobin’s q) via M&A (Gugler et al., 2012). During an economic boom, more firms become efficient and expand their efficient management by acquiring inefficient firms, which triggers M&A waves. Contrary to neoclassical models, behavioral models do not assume that capital markets are efficient (Shleifer & Vishny, 2003). According to the market timing theory, firms’ stocks can be incorrectly valued, and rational managers take advantage of this market inefficiency by using their overvalued stocks to acquire less overvalued or undervalued firms (Rhodes-Kropf et al., 2005). In this regard, high stock market overvaluation leads to M&A waves (see Table 1).

Table 1

Theoretical Approaches of Studies on Determinants of M&A Waves

Theory	Drivers of M&A Waves	Study	Method	Market	Major Finding
Market timing theory	Stock market overvaluation of acquirers’ shares	Rhodes-Kropf et al. (2005)	Panel probit regression, Fama-French	11 industries and all other industries	Acquirers using stock offer are more overvalued than those using cash offer. Acquirers are more overvalued than targets.
Market timing theory	Stock market overvaluation of acquirers’ shares	Yan (2011)	Regression, univariate quantile analysis	U.S. horizontal M&As	M&As during wave periods have poor postmerger performance because the waves occur due to market overvaluation.
Q-theory	Replacement of inefficient management by efficient management	Dong, Hirshleifer, Richardson, and Teoh (2006)	Logistic regression, OLS, t-test	United States	Acquirers’ investment returns are higher than those of their targets. Acquirers with high returns use more stocks for M&A deals.
Industry shock theory	Drastic changes in economic, regulatory, and technological environments	Harford (2005)	Simulation, logistic regression	48 industries in the United States	Economic and regulatory shocks trigger M&A waves on the condition that macro-level liquidity is abundant.
Economic disturbance theory^a	Discrepancy in expected return on investment between shareholders and potential acquirers	Gort (1969)	OLS	U.S. manufacturing M&As	In a fast growing economy, firms’ expected future earnings increase, while investors’ expected returns from stocks vary. This valuation discrepancy facilitates M&A activities, especially when potential acquirers’ returns are higher than those of shareholders.
Credit market theory^a	Credit market conditions (e.g., cost of debt, credit availability)	Maksimovic et al. (2013)	Logistic regression, OLS	U.S. manufacturing M&As	Credit market liquidity is related to M&A intensity. Public firms’ M&As are more sensitive to credit liquidity as compared with those of private firms.
		Benzing (1991)	Autoregression	U.S. manufacturing and mining M&As	As debt financing cost decreases, firms participate in M&A waves.
Merger activity-economic prosperity theory^a	Economic growth	Komlenovic, Mamun, and Mishra (2011)	Panel random effect model	43 industries in the United States	The positive effect of business cycle on M&A waves holds for related and unrelated M&As.

Note: M&A = merger and acquisition; OLS = ordinary least squares.

Indicates the theories tested in this study.

Relationship Between Macroeconomic Conditions and M&A Waves

Changes at the economy-wide level have also been suggested as main determinants of M&A waves. Seminal work by Nelson (1959) first proposed the relationship between macroeconomic conditions and M&A waves. His findings showed that wave movements in aggregate M&A activity were related to the stock market. Following Nelson (1959), many finance and economics studies have empirically supported the systematic effects of macroeconomic conditions on M&A waves (Finn & Hodgson, 2005).

The effect of stock prices on M&A waves has been extensively investigated in empirical studies (Clarke & Ioannidis, 1996; Kamaly, 2007). The economic disturbance theory (Gort, 1969) describes the effect of stock prices on M&A waves. According to the theory, M&A waves are attributable to the different valuation of a business between its existing shareholders and potential buyers. The difference in the expected return between two parties increases when a stock market rises. When potential buyers’ expected returns from a business are higher than those of existing shareholders, the likelihood of M&A deals is greater. Komlenovic et al. (2011) and Resende (2008) have empirically proven that stock prices are related to M&A waves, but others (e.g., Choi & Jeon, 2011; Finn & Hodgson, 2005) have shown inconclusive results about this relationship.

The credit market theory claims that there is a relationship between M&A waves and interest rates (Benzing, 1991). The theory indicates that most firms tend to use borrowed capital for their investments, and therefore, reduction in external financing cost may encourage sizable corporate investments including M&A investment. Low interest rates, which reduce cost of debt, highly motivate M&A activity. Interest rates also affect return on investment (ROI). Firms estimate expected ROI to decide whether to implement their investment plans. When the estimated ROI is higher than the ROI expected by managers or investors, the plan is implemented. Interest rates are used to calculate the plan’s net present value by discounting future cash inflows from the planned investment. The lower the interest rates, the higher the expected ROI. In sum, low interest rates lead to reduced financing costs and increased ROI, which facilitate M&A deals. Many M&A studies (e.g., Choi & Jeon, 2011; Yagil, 1996) have found that interest rate has a significant effect. However, the effect has not been supported by all studies (e.g., Benzing, 1993; Melicher, Ledolter, & D’Antonio, 1983).

Additionally, overall economic activity has frequently been examined as a potential determinant of M&A waves. The relationship between overall economic activity and M&A waves can be explained based on the merger activity-economic prosperity theory (Melicher et al., 1983). According to this theory, strong economic activity serves as a signal of a growing economy in the near future. Such an optimistic economy outlook is likely to encourage firms to participate in M&A activity. Because economic growth has been driven primarily by aggregate demand, firms expect markets in an optimistic economy to experience a lack of supply resultant from growing demand (Dutt, 2006). M&A is considered a viable way for firms to secure additional capacity if they wish to penetrate undersupplied markets. A primary measure of overall economic activity is gross domestic product (GDP). Resende (2008) found determinants of the U.K. M&A waves from 1969 through 2004 and reported the significant effect of GDP. Choi and Jeon (2011) examined the effect of GDP on M&A waves between 1980 and 2004 with regard to deal frequency and deal value. The authors found that GDP had a significant, positive effect on deal frequency but no effect on deal value.

Industrial production is also a measure of overall economic activity. The relationship between industrial production and M&A waves has been inconclusive. Finn and Hodgson (2005) tested the relationship in Australia from 1972 to 1996 and found a reciprocal relationship. Industrial production had a negative effect on M&A waves, and, in turn, the waves positively affected industrial production. On the contrary, some researchers (Benzing, 1991; Corrao, 2012) found the effect of industrial production on M&A waves to be statistically insignificant. Benzing (1991) investigated M&A waves from 1919 through 1979, dividing the study period into two subperiods. Industrial production was not significant on M&A waves for all periods: the entire period and the two subperiods.

Capacity utilization, another indicator of economic activity, has often been explored in relation to M&A waves. There are two contradictory explanations regarding the relationship between capacity utilization and M&A waves. Neoclassical researchers argue that M&A deals occur to modify excess capacity led by rapid technological, economic, and regulation changes (Harford, 2005). Others view that firms use M&A to immediately gain operating capacity in preparation for an optimistic future economy (Komlenovic et al., 2011).

In addition to the aforementioned macroeconomic conditions, whose effects on M&A waves are supported by theories, other macroeconomic conditions were also examined to identify determinants of M&A waves. Those variables include unemployment rate (Benzing, 1991), liquidity (Resende, 2008), inflation (Golbe & White, 1988), export/import (Kamaly, 2007), and oil price (Polonchek & Sushka, 1987).

Use of Predetermined Macroeconomic Variables

Previous studies on determinants of M&A waves predominantly used predetermined macroeconomic variables. Although several variables such as stock prices, interest rates, and GDP have a clear theoretical background, many other macroeconomic variables examined were subjectively selected based on various untested assumptions. This subjective variable selection process may make it difficult to model the true effect of macroeconomic variables on M&A waves because important variables may be omitted from the model. The omitted variables cause endogeneity problems resulting in over- or underestimated coefficients (Wooldridge, 2012). To handle this potential issue, this study investigated a comprehensive set of macroeconomic variables to accurately capture their effects on restaurant and lodging M&A waves.

Method

Data

The sample includes M&A deals between 1981 and 2010 in which U.S. restaurant and lodging firms falling under the Standard Industry Classification (SIC) code 5812 and 7011 were involved. The study period (1981-2010) was determined based on Harford (2005) who attempted to identify industry-level M&A waves corresponding to general M&A waves in each decade (e.g., the 1980s, 1990s). Using the Thomson One Banker database, this study identified 1,332 completed restaurant deals and 1,942 lodging deals. The quarterly frequency of the M&A deals was calculated for data analysis.

This study derived the financial data of hospitality firms from the Compustat database to create industry-level control variables. The quarterly data on 16 macroeconomic variables were collected from the Center for Research in Security Prices, Federal Reserve Board, U.S. Department of Commerce, and U.S. Department of Labor (DOL). Table 2 shows the macroeconomic and industry-level variables and their definitions.

Table 2

Description of Variables

	Symbol	Description	Database
Macroeconomic variables	CPI	Consumer price index, 1982-1984 = 100	DOL
	CU	Capacity utilization	FRB
	DJI	Dow Jones Index	CRSP
	EMP	Total number of employees in millions	DOL
	EX/IM	Sum of export and import in billion dollars	DOC
	FEDS	Federal funds rate	CRSP
	GAS	Gas price	DOL
	GDP	Gross domestic production in billion dollars	DOC
	IP	Industrial production index, 2007 = 100	FRB
	LTI	10-year treasury bond rate	CRSP
	M1	Sum of currency, traveler’s checks, demand deposits, and other checkable deposits in billion dollars	FRB
	M2	Sum of savings deposits, small-denomination time deposits, retail money market mutual funds, and M1 in billion dollars	FRB
	PPI	Producer price index, 1982 = 100	DOL
	S&P	Standard & Poor’s 500 Index	CRSP
	STI	3-Month treasury bill rate	CRSP
	UEMP	Unemployment rate in all industries	DOL
Industry-level variables	CFM	Average operating cash flow margin on sales of restaurant firms	Compustat
	AT	Average asset turnover of restaurant firms calculated as the ratio of sales to assets	Compustat
	CE	Average capital expenditures of restaurant firms calculated as the ratio capital expenditures to assets	Compustat
	ROA	Average return on assets of restaurant firms	Compustat
	SG	Average sales growth of restaurant firms	Compustat

Note: DOL = U.S. Department of Labor; FRB = Federal Reserve Board; CRSP = Center for Research in Security Prices; DOC = U.S. Department of Commerce.

Identification of M&A Waves in the Restaurant Industry

Based on Harford’s method, this study tested whether restaurant and lodging M&A deals occur by chance or in waves. The study period was divided into three subperiods: the 1980s, 1990s, and 2000s. As shown in Figure 1, each decade appeared to have distinct M&A waves in each industry, at least one of which has occurred every decade since the 1980s. The restaurant deal frequency for each decade was, in chronological order, 323, 654, and 355, while the lodging deal frequency was 290, 952, and 700. This study simulated 1,000 distributions of the actual deal frequency for each decade. In the simulation, it was assumed that the probability of assigning each M&A deal to a quarter in a decade was the same (1/40). Accordingly, the simulated distributions had a random process.

Figure 1

Restaurant and Lodging M&A Deals, 1981-2016

Next, this study calculated the deal frequency in the highest 2-year period from each simulated draw to develop a distribution of the 2-year peak period. Then, to investigate the presence of M&A waves in each decade, this uniform 2-year peak distribution was compared with the actual deal frequency. When the actual frequency in any consecutive 2-year period of a decade (e.g., first quarter through eighth quarter or second quarter through ninth quarter) exceeded the 95th percentile of the simulated 2-year peak distribution, the 2-year period was considered to be an M&A wave. This simulation wave identification was conducted for restaurant M&A deals, lodging deals, and hospitality deals that combine restaurant and lodging deals. The results showed that there was at least one M&A wave every decade in each industry. Additionally, this study used the Kolmogorov-Smirnov (K-S) test to confirm the presence of these M&A waves identified using Harford’s method. The results showed significant difference between the distributions of actual 2-year M&A frequency and the simulated 2-year frequency. Both Harford’s simulation method and the K-S test provided empirical evidence of M&A waves, and therefore, the rationale for further study on macroeconomic determinants of restaurant, lodging, and hospitality waves was supported.

Exploratory Factor Analysis

This study employed exploratory factor analysis (EFA) to discover the underlying structures of 16 macroeconomic variables, which were selected through extensive review of previous studies on macroeconomic determinants of general M&A waves. When a large number of macroeconomic variables are included in a model, it may not be easy to extract the unique effect of individual variables. Moreover, economic activities are complicatedly interrelated, and thus, as the number of macroeconomic variables examined increases, more correlation among those variables is likely to occur, causing multicollinearity (Cheng, 1995). EFA can address the multicollinearity problem by identifying latent factors from a large set of explanatory variables. Also, EFA can ease interpretation of the analytical results by reducing dimensions, but it cannot change the nature and character of the original variables.

Because the macroeconomic variables are all time series data, they may be nonstationary. This nonstationarity may lead to violation of the independent and identically distributed assumption required in factor analysis (Gilbert & Meijer, 2005). This study used Augmented Dickey-Fuller test to check the presence of a unit root in each macroeconomic variable indicating nonstationarity. The statistical results showed that most of the macroeconomic variables (except the interest rate variables of federal funds rate, 10-year treasury bond rate, and 3-month treasury bill rate) had a unit root. Therefore, the macroeconomic variables were taken as the first difference to avoid the violation of independent and identically distributed assumption. Also, logarithm transformation was applied to all macroeconomic variables to make them more normally distributed and equally variant.

After transforming the macroeconomic variables, EFA was conducted using principal axis factoring method and direct oblimin method, a nonorthogonal rotation. The number of factors was determined by Kaiser’s criterion and scree plot, and four factors were extracted. Considering the characteristics of macroeconomic variables with high factor loadings in each latent factor, this study named those factors as follows: Overall Activity (OA), Cost of Debt (CD), Market Value (MV), and Inflation (INF).

Distributed Lag Model

Three distributed lag models were developed to explore the relationship between macroeconomic factors and M&A waves in the restaurant, lodging, and hospitality industries. To determine the length of lag, this study compared Akaike Information Criterion of different lag models following the suggestion of Pagano and Hartley (1981). The results indicated that four-lag models were the best fit to the sample. The dependent variable was the quarterly deal frequency. This study also incorporated industry-level variables into the models to control for industry-specific effects. The Durbin-Watson test was used to detect whether there was serial correlation in the models. The Durbin-Watson statistic indicated that the residuals in all models were serially correlated. Therefore, to control for the effect of past aggregate M&A activity on present M&A activity, Prais-Winsten estimation was employed for the distributed lag models based on the approach used by Hansen and Huang (1997). In addition, dummy variables were used to control for seasonality. The estimated models are the following:

{DF}_{Restaurant, t} = β_{0} + \sum_{i = 1}^{4} \sum_{j = 0}^{4} β_{i} {MACRO}_{i, t - j} + {IND}_{k, t} + \sum_{l = 1}^{3} {QUARTER}_{l}

{DF}_{Lodging, t} = β_{0} + \sum_{i = 1}^{4} \sum_{j = 0}^{4} β_{i} {MACRO}_{i, t - j} + {IND}_{k, t} + \sum_{l = 1}^{3} {QUARTER}_{l}

{DF}_{Hospitality, t} = β_{0} + \sum_{i = 1}^{4} \sum_{j = 0}^{4} β_{i} {MACRO}_{i, t - j} + {IND}_{k, t} + \sum_{l = 1}^{3} {QUARTER}_{l}

where DF_t = deal frequency of restaurant, lodging, and hospitality deals; MACRO_i,t = four macroeconomic factors including OA, CD, MV, and INF; IND_k,t = industry-level variables, and QUARTER_l = dummy variables for the first, second, and third quarter of a year (the fourth quarter is the reference season).

Results

Identification of Restaurant M&A Wave

Table 3 provides the results from the investigation of the presence of restaurant, lodging, and hospitality M&A waves. According to the simulation analysis, one restaurant wave occurred for each decade while the lodging and hospitality industries had two M&A waves in the 1980s and one wave in both the 1990s and the 2000s. The deal frequency within each wave accounted for 40% to 80% of the total deals in each decade. Average deal frequency within the waves was at least 52% more than that in nonwave periods. Compared to the restaurant M&A waves, lodging waves were longer and M&A deals were more concentrated in the waves. However, their wave periods were mostly overlapped and corresponded to fourth (1981-1989), fifth (1993-2000), and sixth (2003-2007) general M&A waves.

Table 3

Results of M&A Wave Identification

	Restaurant			Lodging			Hospitality
	1980s	1990s	2000s	1980s	1990s	2000s	1980s	1990s	2000s
Wave period	1987Q3-1990Q2	1994Q4-1999Q1	2006Q1-2008Q4	1983Q1-1985Q1, 1987Q4-1989Q3	1994Q4-1999Q4	2003Q3-2009Q1	1983Q1-1985Q3, 1986Q4-1990Q2	1994Q4-1999Q4	2004Q1-2008Q4
M&A within wave	134/323 (41.40%)	392/654 (59.94%)	161/355 (45.35%)	184/290 (63.40%)	765/952 (80.36%)	501/700 (71.57%)	481/613 (78.47%)	1,206/1606 (75.09%)	664/1055 (62.94%)
Average M&A within wave	11.17/Q	21.78/Q	11.67/Q	9.68/Q	36.43/Q	21.78/Q	18.50/Q	57.43/Q	33.20/Q
Average M&A out of wave	6.75/Q	11.91/Q	7.68/Q	5.04/Q	9.84/Q	11.71/Q	9.43/Q	21.05/Q	19.55/Q
Kolmogorov-Smirnov statistic	0.315***	0.355***	0.258**	0.346***	0.440***	0.483**	0.417***	0.440***	0.452**

Note: M&A = merger and acquisition; Q = quarter.

0.1, **0.05, ***0.01.

K-S test results rejected the null hypothesis for all three periods so that the cumulative distribution function of the actual 2-year M&A frequency is equal to that of the simulated 2-year deal frequency. Figure 2 presents the distribution functions of the actual 2-year, simulated 2-year, and simulated 2-year peak restaurant deal frequencies in the 1990s. The cumulative distribution functions of the actual and simulated 2-year restaurant deal frequencies are shown in Figure 3.

Figure 2

Distribution Functions of 2-Year Restaurant M&A Frequency in the 1990s

Figure 3

Cumulative Distribution Functions of 2-year Restaurant M&A Frequency in the 1990s

Determination of Macroeconomic Factors

Table 4 presents descriptive results of the variables examined. During the 30 years of the study period, stock markets prices have increased more than nine times and GDP more than four times, while interest rates such as 10-year treasury bond rate and federal funds rate have plunged from 12% to 16% to below 0.2%.

Table 4

Descriptive Statistics of Macroeconomic Variables

Variable	M	SD	First Quarter	Last Quarter
CPI	153.91	38.54	87.93	219.63
CU	79.54	3.88	80.32	75.38
DJI	6111.83	3778.20	1005.76	11573.49
EMP	116.87	15.46	91.11	130.28
EX/IM	489.97	290.05	152.37	1096.00
FEDS	5.58	3.62	16.57	0.19
GAS	1.56	0.66	1.34	2.92
GDP	8270.77	8288.84	3051.40	14755.00
IM	277.77	172.21	78.36	607.98
LTI	6.96	2.92	12.96	2.83
M1	1027.53	342.46	415.03	1813.13
M2	4380.00	1974.59	1620.73	8727.67
PPI	129.08	26.04	96.1	188
S&P	717.49	457.70	136.00	1257.64
STI	5.03	3.13	13.36	0.14
UEMP	6.27	1.65	7.43	9.57

This study used EFA to determine the underlying dimensions of 16 macroeconomic variables. Bartlett’s test of sphericity and the Kaiser–Meyer–Olkin measure of sampling adequacy were employed to examine the appropriateness of the macroeconomic data. These test results confirmed that the data were adequate for EFA. Principal axis factoring with direct oblimin rotation was applied to the data. According to Hair, Tatham, Anderson, and Black (1998), oblique rotation is better than orthogonal rotation under the assumption that factors are correlated with each other.

In the factor analysis, each macroeconomic variable was evaluated according to statistical criteria of Hatcher (1994), Hair et al. (1998), and Tabachnick, Fidell, and Osterlind (2001). Both measures of liquidity, M1 and M2, were deleted because of low communalities below .4. This study also discarded EX/IM due to factor loadings greater than .4 on more than one factor and the small difference (less than .3) between high cross-loadings. Without these deleted variables, this study returned to Bartlett’s test and a Kaiser–Meyer–Olkin measure, and these results were still robust. A four-factor model was determined based on Kaiser eigenvalue criterion and the scree plot test. The factor loadings of macroeconomic variables were all above .65. The four factors and factor loadings are shown in Table 5.

Table 5

Results of EFA

Variable	Factor
Variable	Overall Activity	Cost of Debt	Market Value	Inflation
CU	.923
EM	.804
GDP	.699
IP	.964
UNEM	−.910
FEDS		.991
LTI		.817
STI		.941
DJI			.956
S&P			.969
CPI				.770
GAS				.838
PPI				.907
Eigenvalue	5.199	2.540	1.936	1.859
Variance explained	39.99%	19.54%	14.89%	14.30%

Note: EFA = exploratory factor analysis.

Relationship Between Economic Conditions and Restaurant M&A Wave

This study established three distributed lag models to identify the macroeconomic determinants of restaurant, lodging, and hospitality M&A waves. Table 6 summarizes the results from the time-series analysis. In Model 1 (Restaurant), there was no significant factor in the short term as shown in the coefficients at time t. In terms of lagged effect, OA_t_-4 was significantly positive at the 1% level and MV_t_-1, MV_t_-2, and MV_t_-4, were positively significant. These findings indicate previous economic activity, and previous stock prices have a positive effect on current restaurant M&A deals. CD and MV had a significant long-term effect on deal frequency that combined the short-term and all lagged effects. This suggests that although interest rate and stock prices do not affect restaurant M&A activity immediately, the total effect of these factors over time is significant.

Table 6

Coefficients in Distributed Lag Models

		Model 1 (Restaurant)	Model 2 (Lodging)	Model 3 (Hospitality)
Overall activity	t	−2.44 (−1.33)	−3.43 (−1.50)	−3.57 (−1.07)
	t − 1	0.83 (0.38)	1.82 (0.78)	2.12 (0.52)
	t − 2	−1.44 (−0.62)	3.02 (1.31)	2.38 (0.61)
	t − 3	1.31 (0.56)	−0.56 (−0.21)	2.11 (0.52)
	t − 4	3.88*** (2.42)	2.03 (0.98)	5.92** (2.00)
	Short-term	−2.44 (−1.33)	−3.43 (−1.50)	−3.57 (−1.07)
	Long-term	2.15 (0.98)	0.51 (1.52)	8.97* (1.92)
Cost of debt	t (short-term)	4.68 (0.52)	−5.54 (−0.51)	−2.08 (−0.13)
	t − 1	−12.30 (−1.19)	2.13 (0.16)	−9.31 (−0.53)
	t − 2	8.60 (0.84)	14.04 (1.31)	17.42 (1.09)
	t − 3	−3.80 (−0.32)	−10.10 (−0.86)	−10.77 (−0.63)
	t − 4	−4.47 (−0.52)	−6.87 (−0.71)	−8.67 (−0.65)
	Short-term	4.68 (0.52)	−5.54 (−0.51)	−2.08 (−0.13)
	Long-term	−7.29*** (−3.07)	−6.14** (−2.25)	−13.42** (−2.61)
Market value	t (short-term)	1.03 (0.99)	−0.11 (−0.09)	1.11 (0.64)
	t − 1	3.19*** (3.22)	3.23** (2.36)	6.74*** (3.24)
	t − 2	3.18*** (3.13)	4.90*** (3.45)	7.81*** (3.47)
	t − 3	1.65 (1.50)	1.76 (1.59)	3.17* (1.91)
	t − 4	2.08* (1.90)	2.32*** (2.73)	3.58** (2.17)
	Short-term	1.03 (0.99)	−0.11 (−0.09)	1.11 (0.64)
	Long-term	11.14*** (3.22)	12.11*** (3.77)	22.42*** (3.48)
Inflation	t (short-term)	−1.04 (−0.96)	−1.40 (−1.00)	−2.14 (−1.22)
	t − 1	−0.21 (−0.20)	−1.96 (−1.54)	−3.25 (−1.42)
	t − 2	−1.19 (−1.02)	−3.65*** (−2.79)	−4.28** (−2.01)
	t − 3	−1.19 (−0.95)	−1.52 (−1.03)	−4.01* (−1.77)
	t − 4	−0.03 (−0.03)	0.36 (0.26)	0.35 (0.17)
	Short-term	−1.04 (−0.96)	−1.40 (−1.00)	−2.14 (−1.22)
	Long-term	−3.67 (−1.25)	−8.18** (−2.32)	−13.33** (−2.13)
Cash flow margin		5.96*** (3.18)	33.21 (1.27)	9.63*** (2.64)
Asset turnover		16.51 (1.22)	2.20 (1.55)	−13.59 (−0.68)
Capital expenditure		25.77 (0.29)	257.39* (1.69)	437.81** (2.05)
Return on asset		−160.28 (−0.59)	401.45 (1.09)	−16.088 (−0.16)
Sales growth		5.36 (0.07)	60.62** (2.05)	−109.20 (−1.00)
Seasonality		5.63 (0.48)	19.13 (0.97)	38.95 (1.93)
F-statistics		3.30**	3.46***	2.58***
DW statistic		1.82	1.83	1.85
R ²		.41	.54	.41

***

p < .01. **p < .05. *p < .1.

In Model 2 (Lodging), this study did not find any significant short-term effect. However, some significant lagged relationships between macroeconomic factors and lodging deal frequency was found. INF_t_-2 was significantly negative, indicating that previous inflation is closely related to the current deal frequency. MV_t_-1, MV_t_-2, and MV_t_-4 had a significantly positive coefficient. This implies that the current deal frequency is likely to increase one quarter after stock prices increase and that the positive lagged effect lasts up to four quarters after the increase. MV and INF were significant in the long term. The results from Model 3 (Hospitality) showed that all four factors had a significant long-term effect on deal frequency and OA, MV, and INF had a significant lagged effect. For industry-level variables, cash flow margin was significant on restaurant and hospitality deal frequencies, and capital expenditures was significantly related to lodging and hospitality frequencies. Sales growth was also related to lodging deal frequency. Finally, there was no seasonality effect.

Discussion and Conclusions

This study investigated the macroeconomic determinants of restaurant, lodging, and hospitality M&A waves in the United States. To provide rationale for this investigation, this study attempted to verify the presence of M&A waves in those three areas and identified 11 waves between 1981 and 2010. After confirming the presence of M&A waves, this study proceeded to seek macroeconomic fundamentals of the waves. First, this study used EFA to identify the underlying factors among 16 macroeconomic variables and determined a four-factor solution for 13 variables. Then, deal frequency was estimated as a function of these four macroeconomic factors.

There were several distinctive findings from the analysis. First of all, market value had a significant effect on all three industries’ M&A waves supporting the economic disturbance theory. Previous literature has shown the same positive relationship between M&A waves and stock prices (Clarke & Ioannidis, 1996; Resende, 2008). Specifically, the positive lagged effect of market value was significant on deal frequency. This implies that an increase in stock prices precedes the occurrence of M&A waves, and therefore, stock prices can be a significant predictor of the waves. However, it is noted that the positive lagged effect of stock prices became insignificant for all models in the current term. This change in significance can be explained in the context of deal price. The value of target firms is more likely to increase in strong stock markets (Rhodes-Kropf et al., 2005). Higher values of target firms can force acquirers to resile from M&A deals or turn to more affordable targets, which in turn reduces the aggregate frequency of M&A activity. Moreover, increased target value reduces ROI, making M&A deals less attractive or even harmful to firms. Given that hospitality businesses are highly vulnerable to external factors, such as economic swings and rapid changes in consumer preference (Parsa, Self, Njite, & King, 2005; Tang & Jang, 2009), inappropriate or unaffordable investments may not only significantly deteriorate profitability but may also put their sustainability at risk. Because the final price offer is made in the later stage before an M&A agreement, the negative effect of stock prices tends to be contemporaneous and thus may be able to offset the significantly positive long-term effect of stock prices in the short term.

Cost of debt had a significant effect on restaurant, lodging, and hospitality M&A waves. The significant effect was negative in the long term. This finding is consistent with previous studies confirming the credit market theory (Benzing, 1991; Choi & Jeon, 2011). The results can be attributed to the preferred payment method in the hospitality industry. According to Chatfield, Chatfield, and Dalbor (2012), cash was the primary payment method in restaurant and lodging M&A deals. A cash offer was used for 41.2% of all restaurant M&A deals and 71.4% of lodging deals. In this respect, our empirical finding suggests that hospitality acquirers predominantly used debt financing directly associated with interest rates rather than internal cash funds to pay for their M&A deals. Indeed, restaurant firms having higher growth opportunities tend to have more and longer debt using low interest rates effectively (Kim & Gu, 2003; Upneja & Dalbor, 2001), and long-term mortgages are one of primary financing options for lodging firms (Corgel & Gibson, 2005).

Lagged macroeconomic factors were influential on three industries’ M&A waves. The findings indicate that in general, significant changes in macroeconomic conditions can be detected one to four quarters prior to M&A waves. According to Boone and Mulherin (2007), the M&A process prior to the public announcement of an M&A deal is called the Private Takeover Process (PTP), and it begins with top managers’ internal review of various strategic alternatives including M&A. Their study showed that PTP took 8 to 12 months regardless of the method of selecting a merging partner (auction or exclusive negotiation of two parties). It is noticeable that this duration of PTP corresponds to the time lags it takes for macroeconomic conditions to have an effect on M&A waves. This implies that from the initial stages of the M&A process, restaurant and lodging firms constantly consider changes in critical macroeconomic conditions to make more informed M&A decisions. Combined with the findings on the lagged significant factors, it also suggests that changes in economic conditions lead hospitality firms to initiate internal reviews of M&A possibilities as their main strategic plan.

In the short term, no significant macroeconomic factor influenced deal frequency, while all factors had significant long-term and/or lagged effects. From these findings, it is noted that hospitality firms can plan M&A from a long-term perspective. The insignificant short-term relationship indicates that temporary changes in the macroeconomy do not trigger clustering of M&A activity. Along with the significant long-term relationship, this suggests that restaurant and lodging firms are less likely to perform large-scale investment projects such as M&A to take advantage of or address the ever-changing temporary trend in economic conditions but that when obvious trends are observed, firms actively respond to those changes. The significant long-term and lagged relationships also indicate that long-term economic trends that cause restaurant and lodging waves are traceable and therefore can be used to predict M&A waves.

Unlike restaurant M&A waves, lodging waves were affected by inflation. This finding implies that M&A decisions of lodging firms are more responsive to changes in prices than those of restaurant firms. The link between inflation and tourism demand can help explain the sensitivity of lodging waves to inflation. As travel activities are accompanied by a series of product and service purchases, inflation has been one important determinant of tourism demand (Cortés-Jiménez & Blake, 2011). In addition, high inflation can increase price differentials between countries (relative prices) and make the destination less competitive in the international tourism market (Mangion, Durbarry, & Sinclair, 2005). In sum, inflation is negatively related to domestic and international tourism demand that directly affects lodging businesses. It is reasonable that inflation becomes an important consideration in business decisions of lodging firms. On the other hand, given local customers, restaurant businesses may be less dependent on tourists than lodging businesses are, and thus, the importance of inflation could be relatively low in making restaurant M&A decisions. Our finding supports this contention.

Theoretical Implications

The findings from this study have several contributions to academia and the hospitality industry. Academically, this study is an initial attempt to prove the presence of restaurant and lodging M&A waves and investigate macroeconomic determinants of the M&A waves. Growing importance of M&A as an effective growth strategy has led to a number of hospitality M&A studies. However, little hospitality research has focused on M&A waves, and consequently, M&A waves remain a fruitful area of research in the hospitality industry. With regard to the method, this study extends Harford’s M&A wave identification method by supplementally comparing the actual with simulated deal frequencies using the K-S test. While Harford’s method focuses on the comparison of peak points in the actual and the simulated distributions, the K-S test investigates the overall equality of the two distributions based their cumulative distribution function. This study also extends the effect of macroeconomic conditions on industry-level M&A waves by applying the theories for general M&A waves to the hospitality industry. M&A wave studies on macroeconomic determinants have been conducted focusing primarily on general M&A waves. Their results have, however, been inconsistent. Moreover, studies on industry-level waves put great emphasis on industrial shocks and overvaluation of businesses as the main causes of the waves. Therefore, this study identified macroeconomic determinants of M&A waves in the restaurant and lodging contexts demonstrating which general M&A wave theories were applicable to M&A waves in the hospitality industry.

Practical Implications

Another contribution of this study is to provide practical implications for hospitality owners or executives who seek M&A. They should take our findings into account when formulating and implementing their M&A strategy to improve operating performance and shareholder value. The findings show that macroeconomic conditions encourage hospitality M&A deals and the shifts in the important macroeconomic conditions over M&A processes from internal review of strategic options to deal announcement. These macroeconomic conditions should be checked continuously to determine whether economic conditions are appropriate for conducting M&A deals. Hospitality practitioners can also adjust the details of their M&A deals by looking into different macroeconomic conditions at each step of the M&A process. Furthermore, practitioners are able to use the conditions to predict a proper time for M&A.

The findings would also be beneficial for financial analysts and investors. The combination of the significant effect of four macroeconomic factors can help identify the period of hospitality M&A waves. Based on the unique impact of macroeconomy in the hospitality industry, financial analysts can develop a creative industrial forecast to show feasible projections of M&A deals; and, investors and fund managers can find out the optimal time for investing in the hospitality industry.

Limitations and Suggestions

Although this study identified important macroeconomic conditions driving M&A waves, it is not free from limitations. First, this study did not consider firm-level variables such as cost structure and financing capability and did not control firm-specific factors such as firm size and market share. Second, global macroeconomic factors were not considered. As the global market is integrated, more restaurant firms become multinational companies. Consequently, global macroeconomic conditions can be influential on the M&A activity of hospitality firms. Third, this study focused on the aggregate M&A activity without considering the profitability of the M&A deals. Fourth, this study focused only on deal frequency. Future studies can expand understanding of M&A waves in the hospitality industry by addressing the aforementioned limitations. Finally, it would be worthwhile to identify a specific change or event that triggers M&A waves.

Concluding Summary

This study investigated whether M&A waves existed in the hospitality industry and what macroeconomic conditions could significantly lead to hospitality M&A deals creating the M&A waves. For these investigations, this study employed simulation analysis and time-series estimation. The results from the analyses showed that for last three decades, M&A waves constantly occurred in the restaurant and lodging industries and that market value and cost of debt were the significant determinants of the M&A waves. These findings contribute to the hospitality M&A literature by proving the existence of hospitality M&A waves and their significant relationship with macroeconomic conditions and help practitioners and investors more effectively use macroeconomic conditions for their own business activities.

Footnotes

Acknowledgements

We thank the College of Business, Iowa State University, for making the hospitality M&A and financial data available to us.

Jewoo Kim, PhD (e-mail: jjawoo@iastate.edu), is an assistant professor in the Department of Apparel, Events, and Hospitality, Iowa State University, Ames, Iowa. Tianshu Zheng, PhD (e-mail: tzheng@iastate.edu), is an associate professor in the Department of Apparel, Events, and Hospitality, Iowa State University, Ames, Iowa. Susan W. Arendt, PhD (e-mail: sarendt@iastate.edu), is a professor in the Department of Apparel, Events, and Hospitality, Iowa State University, Ames, Iowa.

References

Ahern

K. R.

Harford

(2014). The importance of industry links in merger waves. Journal of Finance, 69, 527-576.

Benzing

(1991). The determinants of aggregate-merger activity: Before and after Celler-Kefauver. Review of Industrial Organization, 6, 61-72.

Benzing

(1993). Mergers: What the “Grimm” data tell us. Review of Industrial Organization, 8, 747-753.

Boone

A. L.

Mulherin

J. H.

(2007). Do termination provisions truncate the takeover bidding process? Review of Financial Studies, 20, 461-489.

Carow

Heron

Saxton

(2004). Do early birds get the returns? An empirical investigation of early-mover advantages in acquisitions. Strategic Management Journal, 25, 563-585.

Chatfield

H. K.

Chatfield

Dalbor

(2012). Returns to hospitality acquisitions by method of payment. Journal of Hospitality Financial Management, 20(1), Article 2.

Cheng

A. C. S.

(1995). The UK stock market and economic factors: A new approach. Journal of Business Finance & Accounting, 22, 129-142.

Choi

S. H.

Jeon

B. N.

(2011). The impact of the macroeconomic environment on merger activity: Evidence from US time-series data. Applied Financial Economics, 21, 233-249.

Clarke

Ioannidis

(1996). On the relationship between aggregate merger activity and the stock market: Some further empirical evidence. Economics Letters, 53, 349-356.

10.

Corgel

J. B.

Gibson

(2005). The use of fixed-rate and floating-rate debt for hotels. Cornell Hospitality Quarterly, 46, 413-430.

11.

Corrao

(2012). United States mergers & acquisitions market dynamics: What drives deal flow in each industry? (Unpublished Working Paper). Evanston, IL: Northwestern University.

12.

Cortés-Jiménez

Blake

(2011). Tourism demand modeling by purpose of visit and nationality. Journal of Travel Research, 50, 408-416.

13.

Dong

Hirshleifer

Richardson

Teoh

S. H.

(2006). Does investor misevaluation derive the takeover market? Journal of Finance, 6, 725-762.

14.

Dutt

A. K.

(2006). Aggregate demand, aggregate supply and economic growth. International Review of Applied Economics, 20, 319-336.

15.

Finn

Hodgson

(2005). Takeover activity in Australia: Endogenous and exogenous influences. Accounting & Finance, 45, 375-394.

16.

Gilbert

P. D.

Meijer

(2005). Time series factor analysis with an application to measuring money (Research Report 05F10). Groningen, Netherlands: Research Institute SOM (Systems, Organizations and Management), University of Groningen.

17.

Golbe

D. L.

White

L. J.

(1988). A time-series analysis of mergers and acquisitions in the US economy. In Auerbach

A. J.

(Eds.), Corporate takeovers: Causes and consequences (pp. 265-310). Chicago, IL: University of Chicago Press.

18.

Gort

(1969). An economic disturbance theory of mergers. Quarterly Journal of Economics, 83, 624-642.

19.

Gugler

Mueller

Weichselbaumer

B. B.

(2012). The determinants of merger waves: An international perspective. International Journal of Industrial Organization, 30, 1-15.

20.

Hair

J. F.

Tatham

Anderson

R. E.

Black

W. C.

(1998). Multivariate data analysis (7th ed.). Englewood, NJ: Prentice Hall.

21.

Hansen

Huang

(1997). Road supply and traffic in California urban areas. Transportation Research Part A: Policy and Practice, 31, 205-218.

22.

Harford

(2005). What drives merger waves? Journal of Financial Economics, 77, 529-560.

23.

Hatcher

(1994). A step-by-step approach to using the SAS system for factor analysis and structural equation modeling. Cary, NC: SAS Institute.

24.

Hsu

L. T. J.

Jang

S. C. S.

(2007). The postmerger financial performance of hotel companies. Journal of Hospitality & Tourism Research, 31, 471-485.

25.

Kamaly

(2007). Trends and determinants of mergers and acquisitions in developing countries in the 1990s. International Research Journal of Finance and Economics, 8, 16-30.

26.

Kim

(2003). Risk-adjusted performance: A sector analysis of restaurant firms. Journal of Hospitality & Tourism Research, 27, 200-216.

27.

Kiymaz

(2004). Cross-border acquisitions of US financial institutions: Impact of macroeconomic factors. Journal of Banking & Finance, 28, 1413-1439.

28.

Komlenovic

Mamun

Mishra

(2011). Business cycle and aggregate industry mergers. Journal of Economics and Finance, 35, 239-259.

29.

Zhang

Chowdhury

(2011). Stock performance of firms acquiring listed and unlisted lodging assets. Cornell Hospitality Quarterly, 52, 291-301.

30.

Maksimovic

Phillips

Yang

(2013). Private and public merger waves. Journal of Finance, 68, 2177-2217.

31.

Mangion

M. L.

Durbarry

Sinclair

M. T.

(2005). Tourism competitiveness: Price and quality. Tourism Economics, 11, 45-68.

32.

Martynova

M. V.

Renneboog

(2008). A century of corporate takeovers: What have we learned and where do we stand? Journal of Banking & Finance, 32, 2148-2177.

33.

Melicher

R. W.

Ledolter

D’Antonio

L. J.

(1983). A time series analysis of aggregate merger activity. Review of Economics and Statistics, 65, 423-430.

34.

Mitchell

M. L.

Mulherin

J. H.

(1996). The impact of industry shocks on takeover and restructuring activity. Journal of Financial Economics, 41, 193-229.

35.

Nelson

R. L.

(1959). Merger movements in American industry, 1895-1956. Princeton, NJ: Princeton University Press.

36.

Pagano

Hartley

M. J.

(1981). On fitting distributed lag models subject to polynomial restrictions. Journal of Econometrics, 16, 171-198.

37.

Parsa

H. G.

Self

J. T.

Njite

King

(2005). Why restaurants fail. Cornell Hospitality Quarterly, 46, 304-322.

38.

Polonchek

J. A.

Sushka

M. E.

(1987). The impact of financial and economic conditions on aggregate merger activity. Managerial and Decision Economics, 8, 113-119.

39.

Resende

(2008). Mergers and acquisitions waves in the UK: A Markov-switching approach. Applied Financial Economics, 18, 1067-1074.

40.

Rhodes-Kropf

Robinson

D. T.

Viswanathan

(2005). Valuation waves and merger activity: The empirical evidence. Journal of Financial Economics, 77, 561-603.

41.

Shleifer

Vishny

R. W.

(2003). Stock market driven acquisitions. Journal of Financial Economics, 70, 295-311.

42.

Tabachnick

B. G.

Fidell

L. S.

Osterlind

S. J.

(2001). Using multivariate statistics. New York, NY: Harper & Row.

43.

Tang

C. H. H.

Jang

S. S.

(2009). The tourism–economy causality in the United States: A sub-industry level examination. Tourism Management, 30, 553-558.

44.

Tse

E. C. Y.

Crawford-Welch

(1989). An analysis of merger and acquisition activity in the hospitality industry for the period 1970-1988. Journal of Hospitality & Tourism Research, 13(3), 1-13.

45.

Upneja

Dalbor

M. C.

(2001). An examination of capital structure in the restaurant industry. International Journal of Contemporary Hospitality Management, 13, 54-59.

46.

Wooldridge

J. M.

(2012). Introductory econometrics: A modern approach. Mason, OH: South-Western.

47.

Yaghoubi

Locke

Gibb

(2016). Mergers and acquisitions: A review. Part 1. Studies in Economics and Finance, 33, 147-188.

48.

Yagil

(1996). Mergers and macro-economic factors. Review of Financial Economics, 5, 181-190.

49.

Yan

(2011). Merger waves: Theory and evidence. Quarterly Journal of Finance, 1, 551-606.